---
title: "Lenny's Podcast — 2025 Q1 合集"
date: "2025-01-01"
source: "Lenny's Podcast"
url: "https://www.lennysnewsletter.com/"
---

<div v-pre>


# Lenny's Podcast - 2025 Q1 (15 episodes)

This file contains 15 articles/episodes.

---

## [1/15] Scripts for difficult conversations: Giving hard feedback, navigating defensiveness, the three questions you should end every meeting with, more | Alisa Cohn (executive coach)

**Lenny Rachitsky** (00:00:00):
I want to dive right into talking about your advice on having difficult conversations, where like in performance review season, what do you suggest when someone's being told they're not going to get the promotion?

**Alisa Cohn** (00:00:10):
Hope for the future is so important. I know this is going to be challenging for you to hear, not going to promote you, but I want you to know this. It's really important to me that you're able to succeed in your career here, and so I want to continue to help you find opportunities to build your skills and to advance.

**Lenny Rachitsky** (00:00:24):
You're big on helping leaders understand that their job is not to make employees happy.

**Alisa Cohn** (00:00:29):
They're trying now to be the leader who everyone loves, but what really needs to happen very often is, we need to drive towards results. This employee continuing to not really do a great job at their job, you don't want to push them because you don't want to upset them. You don't want to give them difficult feedback, so you're just going to keep hoping it works out. Ultimately, that leads to the demise of your company.

**Lenny Rachitsky** (00:00:50):
You have some cool advice on just how to make meetings more effective and how to especially end the meeting.

**Alisa Cohn** (00:00:54):
My three questions to end the meeting are...

**Lenny Rachitsky** (00:01:00):
Today my guest is Alisa Cohn. Alisa is an executive coach who has worked with C-suite execs at both startups like Etsy, Wirecutter, Venmo, and DraftKings, along with Fortune 500 companies like Microsoft, Google, Pfizer, and the New York Times. She was named one of the top 50 coaches in the world by Thinkers50 and the number one startup coach for the past four years by Global Gurus. What I love about Alisa is that she gives her clients very specific and actionable advice. In her conversation, Alisa shares specific language and phrases that you can use when having a difficult conversation with your reports to make these conversations go much smoother and be less difficult. Also, three questions you should ask at the end of every meeting to make the most possible forward progress after each meeting. Plus, why your job as a leader isn't to make people happy and what you should be focused on instead, and a set of questions that she calls the founder prenup that you should talk through with potential founders to make sure that these are the people that you want to be working with for a long, long time.

**Alisa Cohn** (00:04:54):
Lenny, it's so great to be here and thanks for having me.

**Lenny Rachitsky** (00:04:57):
I want to dive right into talking about your advice on having difficult conversations. I personally dread difficult conversations. I feel like I practice ahead of these things. I'm like, "I'm going to say these things. It's going to go like this," and it never goes as well as I hope. I always say the wrong thing. I feel like this is very relatable. They're called difficult conversations for a reason.

**Alisa Cohn** (00:05:18):
Totally.

**Lenny Rachitsky** (00:05:19):
I know you work with a lot of execs on this specifically, and what I love is you've actually come up with a bunch of scripts that help people make these conversations less difficult. So how about we talk through some of these scripts that people can actually start applying?

**Alisa Cohn** (00:05:32):
Let's do that. I love that idea. And also Lenny, as you just said, very relatable and also, so you're not alone. If I could ask you a question, if you're picturing a difficult conversation that you have had, should have, might have, and you're nervous about, it's hard for you, can you sum it up? What's hard about it? Because it's helpful to clarify what is hard about it?

**Lenny Rachitsky** (00:05:56):
Great question. I just don't want to make people sad and upset, and I worry about their reaction, how to deal with that, and them just getting really upset and mad and just like, "Oh, man. This really made things worse." So I worry about the reaction, I guess.

**Alisa Cohn** (00:06:11):
Okay, about making things worse or about their reaction?

**Lenny Rachitsky** (00:06:13):
The reaction, just making someone upset and sad. I don't want to do that.

**Alisa Cohn** (00:06:17):
Making someone upset. Okay, good. And again, you're not alone about that. Just one more question on that. What's the problem if they're sad and upset? What does that mean to you?

**Lenny Rachitsky** (00:06:26):
Oh, I love this life coaching we're doing. Yeah, so it's like what happens if they get sad and upset?

**Alisa Cohn** (00:06:33):
Yeah.

**Lenny Rachitsky** (00:06:36):
I feel like it's stuff that I'm going to have to deal with. It's like this drama all of a sudden, this new fire I have to think about. And yeah, it's like the additional work it creates and also just, I don't know. Yeah, it's a good question.

**Alisa Cohn** (00:06:48):
You can think about it some more, right? I'm not going to put you on the spot right now, but just to say for all of us, the reason they're difficult, to your point, they're difficult. But we're putting meaning on things all the time, every day, all the time, and I think it's important, it's actually helpful in motivating you to have difficult conversations, but also in helping them go well. If you can get to the bottom of what you're putting on top of it, what you're weighting it with, because I can understand that again, you are not alone. I don't want to make people upset. Totally.

**Alisa Cohn** (00:07:20):
And also, I would just say on the other hand, when you're enlightening someone or you're working out a situation with someone and it's difficult, if you don't give them the opportunity to hear what you have to say, if you don't bring this up, then you're never going to have the opportunity to help them see something differently or help them improve or help you improve the relationship or whatever it is you're trying to do. And so, I can understand it's a natural thing. I don't want to make them upset.

**Alisa Cohn** (00:07:51):
No one wants to make anybody upset, but through that upset on the other side of that, can often be a whole new possibility and a whole new revelation, and actually a lot of joy and freedom. I think that we forget about all the other possibilities that come out of difficult conversations and we just land on these really uncomfortable parts about like, "oh, it's going to be a lot of extra work" or like, "They're going to get uncomfortable or even maybe cry." And I think it's just really helpful to tap into what you make it mean and then also what other possibilities it could mean.

**Lenny Rachitsky** (00:08:23):
I love that. And it's one thing to hear that and say that, it's another to actually feel that deeply and feel like I shouldn't be as worried as I am. I think part of it is doing these enough times where you're like, "Okay, it's actually not so hard." And the other is having some of this support. To make this even more real, let's give some examples of what we say when we say difficult conversations. There's like, "You're not getting a promotion that you thought you would, we're going to let you go." What other examples or common difficult conversations that you run across?

**Alisa Cohn** (00:08:50):
Those are two very common ones. And then of course, the most common one is just difficult performance feedback. Or [inaudible 00:08:56] what we say, quote unquote, "constructive performance feedback," which we never made positive. It only is the sort of things that you're not doing well. I think there are two flavors of that. One is, "You're screwing up" and the other is, "Developmentally, I'd like to see you add something or change something."

**Lenny Rachitsky** (00:09:12):
Yes. And as you say that, one of the other fears I have is them just disagreeing and me feeling like maybe it's not right, maybe I'm wrong and feeling shit maybe. I didn't see something and then just looking worse after the whole thing.

**Alisa Cohn** (00:09:27):
Yeah. And so then I think what's also really helpful to, and part of the process that we can talk about this for sure is getting a difficult conversation is number one, tapping into what's uncomfortable for it, for you, about it. And then number two, also getting your mindset right. So to say the obvious, are you doing this to hurt someone's feelings? No, never, right?

**Lenny Rachitsky** (00:09:49):
The opposite.

**Alisa Cohn** (00:09:50):
That's not the reason that anyone's doing it. Sometimes people are giving the performance feedback or talking about something that's been bothering them in order to express themselves and vent. And actually, that is very helpful to identify for yourself, that's why I'm doing it. And then, maybe not do it then, until you can transform your reasoning. But at the end of the day, the hope is as a manager, the reason that you're giving someone this so-called constructive feedback is because you're helping them get better. You need them to change the behavior. They'll never get promoted if they keep doing that. They'll never be successful if they keep doing that. And so, it's your job as a leader and as a manager, to help them out of that problem and help them do something different.

**Lenny Rachitsky** (00:10:33):
The best story I've heard to make that really real for me, I think it was Kim Scott when she came on the podcast. She told a story of, I think it was Bob, where everyone just knew he was terrible and it was like, everyone's was just like knew he was not good and eventually, the boss had a conversation with him eight months into it and told him, "It's not going to work out. You're just doing a bad job." And he's like, "Why didn't anyone tell me? I didn't realize that. If you told me, I would've changed." And everyone assumed he knew. And so I think to your point, this is to help the person. It's not not to hurt them.

**Alisa Cohn** (00:11:05):
Yeah, a hundred percent. One of my clients, he was running a division and one of his people was not doing it right, not doing it right, not getting the right kind of data, not having to do the right kind of analysis, whatever it was. We were talking about it and I said, "Well, how come you have another feedback with her?" And he said, "You know, she's just going to cry. She's just going to cry. She's older, whatever, she's just going to cry. It's going to be too uncomfortable, whatever." So we worked, we talked and talked and talked. I gave him a script. We really worked it out and he agreed that he would go in and have that conversation with her, which he did.

**Alisa Cohn** (00:11:40):
And he reported back to me and he was shaken. She cried. Of course she did. She cried. That's what he knew she was going to do. And so she was upset and she went home early and the whole thing. The next day she came in and she said, "Thank you so much for telling me that. I wish someone had told me that 15 years ago. I think I could have had a different career." And I think that is so meaningful for all leaders and people who are responsible for other people to understand that you're uncomfortable when they start crying, of course, or they have this difficult reaction or whatever. But honestly, the only way you're going to be able to help someone grow in their career and become the best person they can be is by leaning into these tough conversations.

**Lenny Rachitsky** (00:12:23):
What I love about the scripts we're going to talk about, which we probably should transition to, is it's again, one thing to hear that and be like, "Yes, okay, I need to do this. I need to get better at typical conversations. I need to have that talk without someone that we should let go." It's another when it's like tomorrow is the meeting and you're like, "Oh, my God. I have to have this conversation now." And so, I love that you actually give people a really simple approach to how to lay this stuff out in various different contexts. So let's talk through some of these approaches and scripts you've come up with. What do you think would be a good one to start with?

**Alisa Cohn** (00:12:55):
Well, we can start with performance feedback and we can just sort of take a typical example. So first of all, once you've done your work to get your mindset right to kind of know what you're doing it, and then you just really want to really be able to wrap your mouth around the words. So what that looks like is practicing, and the script could be, "You know, Matilda, I want to chat with you about the way you're interacting with your peers. So what I'm hearing from them is that you're missing deadlines on a regular basis and not letting them know you're missing the deadlines, and that also you're not fully keeping your team up to speed.

**Alisa Cohn** (00:13:27):
And so they're kind of confused running around. Now, we both know that the most important way you can be successful here and also achieve your goals is to make sure that you are working with your peers in a way that's consistent and that they can count on you and you can count on them. So I wanted to let you know about this. I want to certainly hear what you have to say, but the most important thing is that we leave this discussion knowing how you're going to make sure that you're keeping your peers in the loop and also your team in the loop."

**Lenny Rachitsky** (00:13:55):
Yeah, there's so many elements there that are really interesting. Just focusing on what I'm hearing versus just coming from you or something you've done wrong. It's, here's what I'm hearing from multiple sources. I think that helps people. Okay, it's not just you and just like, "Oh, my manager hates me."

**Alisa Cohn** (00:13:56):
Right.

**Lenny Rachitsky** (00:14:14):
It's like, "Okay, other people are saying this." And then I love this phrase of, we both know where it's not just me telling you this. It's like, "You also know this. I know you're smart and you also know that this is, something is wrong here." And then this goal of, here's what we need to [inaudible 00:14:31]. You're like very clear call to action, almost action item, like leave this meeting with, "Let's just be aligned on this thing."

**Alisa Cohn** (00:14:36):
Yeah, thanks for calling those out. I hope, and again, what I'm trying to convey in my tone is also, "You know what? It's Tuesday. We got to have this conversation. I'm sure it's going to end well. I'm not mad." The whole point about my manager hates me, right? "I'm not yelling at you." The more even keeled and even matter of fact you can be about something that's kind of just run-of-the-mill feedback, the better. And I think it's just also what I didn't say before, and I think it's also important is that, as you are recognizing that one of your jobs is to give this feedback, is that you have to build a relationship with people so they can hear you through the lens of, "Oh, Alisa wants to help me." Not, "Oh, Alisa hates me. It's always a problem."

**Lenny Rachitsky** (00:15:21):
How did you start that phrase again? Because the starting is always the hardest part for me. How do you kick off the conversation? What was the couple sentences used?

**Alisa Cohn** (00:15:28):
I wanted to have a conversation with you about some things I've been hearing from your peers about the way that you all are interacting together.

**Lenny Rachitsky** (00:15:35):
Awesome. So there's an element of, don't make it feel like a huge deal. Just like, "I want to have this conversation with you about something." And it's just like, "Let's have this conversation and here's what we want to leave this conversation with."

**Alisa Cohn** (00:15:46):
Yes. And I can't stress enough that it's actually really helpful to also have spent some time with Matilda or whoever saying, "Great job on the way that project landed." Or, "Hey, launches, when they happen on time and they're smooth, sometimes we don't notice anything. I want you to notice, we didn't notice anything. That's fantastic. You did a good job in that launch," or whatever it is. Because then, you've had the conversation with them to give them positive feedback and point out what's working, that builds the relationship so that you have the lens of, "Oh, yeah. When something's working, they tell me. When something's not working, they tell me, too." That's how you build trust as well.

**Lenny Rachitsky** (00:16:23):
They want to be criticizing them [inaudible 00:16:25]. We need to have another conversation what we're hearing about, problems [inaudible 00:16:28].

**Alisa Cohn** (00:16:27):
Yep.

**Lenny Rachitsky** (00:16:29):
Obviously, if you say it the same way every single time, they're going to feel like this is weird. Do you recommend it's this kind of Mad Libs approach or is it make it your own as much as you can? What are kind of the key? Or is it like, here's actually how you want to say it every time?

**Alisa Cohn** (00:16:42):
In my book and when I work with my clients, I give specific scripts and what I will regularly say when I'm working with my clients is, "Okay, so this is how I would do it," and then I'll land it for them. But they have to make it their own. You always have to make it your own and I don't think it's a problem of doing it the same way every time. It's not like people are going to notice you because you're talking about different topics, theoretically. If you have a formula that can work for you, that's going to motivate you to do it, that is what's important. And what's important is that it's neutral, not loading on or not venting on someone and not unloading on someone.

**Lenny Rachitsky** (00:17:19):
I love that we started with this one because it feels like the most common one of just your employee is underperforming and you want to make sure they understand and adjust. What if you're not hearing something from a bunch of people? What if it's just your perception of their writing? You need to work on your writing skills or you're coming in late. Is there another way you phrase it where it's not, "I'm hearing it from other people?"

**Alisa Cohn** (00:17:43):
Oh, absolutely. Absolutely. So I'll talk about writing. I think it would be something like, okay, "Matilda, part of your job is to be able to create these documents and I appreciate that you do them on time. What I've observed is that they can often be not as structured as I'd like them to be and they also lack a conclusion. So what I'd love you to do is look at these three or four examples of some folks who are doing them really well and see if you can model your writing on theirs. If you need to take additional classes or if you need help in any way, let me know. But ultimately, I want to get your writing to the level where everybody is appreciating what you bring to the table because the level of your writing really reflects the level of your thinking."

**Lenny Rachitsky** (00:18:27):
Mm-hmm. Wow, I like that. I'd want to follow your advice if I got that. So the way you started that is what I've observed, which also is not like, "Here's what I think" or "Here's what you just need to do." It's more like, "Here's what I've noticed, here's what I've seen, here's what I've observed about what you're doing." And then it reminds me of, what is it, nonviolent communication, that whole framework of just focus on what you see, not what is wrong with them, not what they've done. I guess, is there anything there you want to say of just the importance of focusing on what you've heard from people or what you've observed versus maybe what people often do instead?

**Alisa Cohn** (00:19:05):
Yeah, I mean you just really said it and I think it's such an important point, observable facts. The idea that this is not a judgment. This is not... Sort of as less judgy as possible is also very helpful. It makes it neutral. It's observable facts and it's also sort of based on expectations, right? So the writing is, we expect it to be at a certain level and it's not that way. And here are the reasons it's not, the specific reasons, it's not.

**Alisa Cohn** (00:19:35):
The way you interact with your peers, it's important to be at a certain standard, and here's why. Because when we all work together, we're going to be able to execute and when we don't, unfortunately we won't be able to. So you staying in sync with them is important and the observation is that they don't feel fully in sync with you.

**Alisa Cohn** (00:19:52):
And so every time we talk about this, it doesn't become this, "Oh, I don't know. I just feel..." By the way, some things you have to give feedback on and they are kind of a feeling and those are more difficult, but so many things if you do the work to really think about what is the observable data, I always ask my clients, what's my evidence that this is happening? And you have to spend some time thinking about it, but it's really worth it because it makes the feedback easier for you to give and easier for them to hear.

**Lenny Rachitsky** (00:20:20):
Is there anything else along the lines of this specific type of feedback that is worth sharing before we move on to a different type of feedback?

**Alisa Cohn** (00:20:29):
Well, I think just that the reason, one of the many reasons that people have gone uncomfortable giving feedback is that somebody might get defensive or they might start crying as we talked about. And so I have a script also, which is if someone gets defensive, which is it's like I'm giving you this feedback and you're getting defensive and I say, "Well, let's pause for a second. First of all, I want you to know that I'm telling you this actually, just to make you better because I know how important your career is to you. I know how important the success is to you and it's important to me too as your leader. The second thing is, my observation is that you're getting a little bit emotional. I want to know if we can continue having this conversation now or if we need to kind of pause it. At the end of the day, we really have to have this conversation and I really want to see you make changes, but I understand you might need a few moments to digest it."

**Alisa Cohn** (00:21:18):
The importance of that for you is not even what you say, but that you have prepared and you are prepared for if someone has that kind of reaction and that you don't have to, yourself, react to it. You know, "No, I'm not doing that. No, no, no, no, whatever." And you can say, "Yes, you are." Now we're in a fight and that is not cool for anybody. It's certainly not cool for you as a leader. So it gives you the opportunity to recognize that you have another tool in your toolkit rather than just react.

**Lenny Rachitsky** (00:21:48):
So if you find yourself feeling defensive or they are just not hearing and just fighting back, the tool is just pause. Let's just pause for a moment and it feels like there's kind of two parts to which you just shared. One is, remind them why this is important to them and why you're talking about this. And then two is, if there's just emotions kind of taking over, give them a chance to like, "Let's just pause and maybe come back to this because maybe you're not in the right state right now to listen."

**Alisa Cohn** (00:22:18):
Yeah, exactly.

**Lenny Rachitsky** (00:22:20):
Sometimes people get upset when you mention like, "You're getting emotional," or I don't know. Is that a thing that you deal with of just like, "How dare you say I'm feeling emotional?" I'm just...

**Alisa Cohn** (00:22:29):
I'm not emotional. Why do you think [inaudible 00:22:31] emotional? Right, exactly. Yes, of course. Now, when someone's crying, they're obviously getting emotional. When they're defensive, it's possible that you might want to use a different word. I can see that this is really upsetting you or this is really triggering you, or I can see that the temperature between us has just changed. You could say something like that. I do think also it's helpful to know your people because sometimes you could realize that actually they can deal with that, but then sometimes you have to really [inaudible 00:23:00] the delicate words that you need to use to pause the conversation.

**Lenny Rachitsky** (00:23:05):
Yeah. And I find, to your point, it's helpful to you too as the person giving it. And I feel like sometimes, you may be feeling like I should just pull back and maybe I'm wrong, maybe they're right, maybe I should stop and instead this gives you a chance to know I'm actually, I can't. I need to stay strong about what I believe because I... You put so much thought and effort into this already, it's unlikely you're just like, oh, totally wrong about what you're saying.

**Alisa Cohn** (00:23:30):
Yeah, exactly. There's something going on. There's something going on. And then also, the whole point about it being a conversation is that actually it's a conversation. Actually, Lenny, if you have a different point of view, I would like to hear it. Let's talk about it, but we can't keep going on like this, where I don't feel I can count on you for whatever it is that we're talking about. So we need to have this conversation and recreate a set of expectations between ourselves. Ultimately, that kind of conversation has the potential to really build the relationship and build trust, and that's another reason I encourage everybody to get over their discomfort and to lean into having these conversations because on the other side of that, is a much better, stronger connection.

**Lenny Rachitsky** (00:24:09):
And especially if you do them well.

**Alisa Cohn** (00:24:11):
Yes.

**Lenny Rachitsky** (00:24:12):
Following this advice. So okay, so again, if somebody's feeling defensive, can you again say how you start that, if you notice that? And then I'll highlight the two elements again of the...

**Alisa Cohn** (00:24:25):
So the way to pause is to actually say, "Let's just pause for a second because I'm feeling the energy has changed and I can see that you're getting a little bit heated by what I'm saying and I want you to know that I have no intention of upsetting you. I just want to be able to talk to you about the things that are going to help you in your career."

**Lenny Rachitsky** (00:24:45):
Awesome. And I love, again, just the reminder of here's why this is important to you, here's the benefit to you and why this will help you. And then it's like, "Okay, let's just maybe take a pause and come back to this conversation if you're feeling like this isn't the best time." Awesome. Anything else along that line before we go to another type of a hard conversation?

**Alisa Cohn** (00:25:02):
I mean, I can talk all day about this [inaudible 00:25:06], but I'm happy to move on.

**Lenny Rachitsky** (00:25:07):
Well, let's pick another topic. I know you have kind of five buckets and types of conversation. Maybe the promotion one. That feels like I think we're in performance review season. It feels like these are happening a bunch. What do you suggest when someone's being told they're not going to get the promotion they expected or wanted?

**Alisa Cohn** (00:25:24):
Of course that's challenging. So again, getting your mindset right, recognizing they're disappointed, they're going to be disappointed, recognizing how you felt, the time that when you didn't get a promotion or whatever. And so kind of coming to it with some compassion. And also, you have to get your reasoning right. So sometimes people think they should get a promotion because they were here for a year or whatever. Sometimes people think they should get a promotion because they're the only internal candidate who's qualified for this or they might have a sense of themselves succeeding or achieving that is more inflated maybe than you see them. So trying to think about where they're coming from.

**Alisa Cohn** (00:25:56):
And then the conversation is just, "Matilda, I know this is going to be challenging for you to hear. I know you were hoping to get that promotion, but I want to let you know that we are going to actually be looking for an external candidate. I want to give you a few thoughts about why. First of all, in discussing this with my peers, I'm realizing that we need someone who has done this role multiple times in the past and has that experience. Number two, I think it's really important that they have an expertise in a specific realm that we've identified as really important. So for those reasons, we're going to bring someone in from the outside, not going to promote you, but I want you to know this. Number one, it's really important to me that you're able to succeed in your career here. And so I want to continue to help you find opportunities to build your skills and to advance. And then number two, when we bring this person in, I'm committed to finding someone who's a great people leader, who is going to help you build those skills."

**Lenny Rachitsky** (00:26:54):
So a few elements there that stood out to me. One is just being very upfront and not bearing the lead. Telling them very early, "Here's what I've decided." As you said it, I could see my heart sinking immediately when I feel that. So at least that's over and then it's, here's why. And that starts to help you feel like, "Okay, I get it. I understand at least how you thought about this." And then there's the hope for the future, your painting of, here's how I can get there eventually.

**Alisa Cohn** (00:27:21):
Yes, that hope for the future is so important and I think sometimes we're such in a rush to kind of deliver the bad news that we forget there's a human being over there who needs hope for the future. And hopefully. If they're a good employee, hopefully they have hope for the future.

**Lenny Rachitsky** (00:27:35):
I love that. Is there anything else to that script that you think is really highlighting or do you think I touched on the key elements?

**Alisa Cohn** (00:27:41):
I think you touched on the key elements.

**Lenny Rachitsky** (00:27:43):
Okay. And again, the way you started is, I have some bad news for you or I have some disappointing news for you.

**Alisa Cohn** (00:27:48):
Yes, because it's just [inaudible 00:27:50].

**Lenny Rachitsky** (00:27:48):
Just get right into it. Yeah.

**Alisa Cohn** (00:27:51):
Yeah, just get right into it. Yeah. By the way, the other piece on that might be, if it's appropriate, I'd love you to digest this information and then let's talk about it again next week to see what you've come up with or see how you feel about it because you want to send, this is not the script, this is for me to you. You want to send the, I care about you message because that's the other thing. In the workplace, people, they're going through all their feelings, all their emotions, disappointments. They're going to go home and tell their spouse, didn't get the promotion or whatever. It's going to loom large. It's going to be demoralizing.

**Alisa Cohn** (00:28:24):
When you, as a leader signal a lot, I care about you, I care about your feelings, I care that you're disappointed, I care about your career, you are always going to be able to help people stay resilient in the face of setbacks and ultimately, do extra work, do the right work for you and be engaged in your company because you've spent the time and energy making sure they know that even when things are not going their way, they have an ally in you.

**Lenny Rachitsky** (00:28:56):
What do you do if they just disagree, if they're just like, "But I do have those skills and I don't think this is fair." Thoughts on responding to that sort of feedback? I guess, that's the defensiveness stuff.

**Alisa Cohn** (00:29:06):
Yeah, that's the defensiveness stuff. And again, I hope you've done your homework to identify that actually that person doesn't have those skills and if there is a [inaudible 00:29:16] for example, but I do have those skills or sometimes people, I think more, even more often, they don't respond to what you just said. They will instead explain to you that they've been here for a year or they're the only internal candidate or their peer got promoted.

**Alisa Cohn** (00:29:31):
Right, they'll sort of explain to you things which are not part of your decision-making process and then it's helpful for you to say something like, "Yeah, listen, Matilda, I really understand that you were thinking that after a year, you'd get promoted around here. And in the past, I do think because of the stage of our company, probably people have been promoted at that period. That's not the place we're at right now. As we scale, we really need to think about not just what we need for today and tomorrow, but for the future. And that's why I want these specialized skills in here. I think it's going to help the entire company."

**Alisa Cohn** (00:30:02):
So that's an example of a discussion that you could have. I do have the skills. That's kind of interesting. I'd love to hear what you see as those skills. And it's not a problem to have the conversation right there then, but if there's a "Yes, I do, no, I don't, yes, I do, no, you don't," that pushback is never productive. And so, that's where you want to probably again take a pause and say, "Listen, I totally hear you. You and I have a different point of view about this. I'm not sure if it's productive to continue to discussing right now. Let's talk about it again in a week. But I also want you to know this is a decision that I've made."

**Lenny Rachitsky** (00:30:39):
I love though, when they come back to you and like, "But here's X, Y, Z." And you're like, "That's not what I was saying necessarily." I love that you basically mirror back. I hear what, I understand you believe, I understand you've been here for a year. I understand you're the only internal candidate," like making them feel very heard. That's a really powerful mechanic there. That is a good tool. Is there another script that you think might be helpful to talk through that is a common hard conversation people have?

**Alisa Cohn** (00:31:05):
Well, the hardest conversation is firing someone.

**Lenny Rachitsky** (00:31:09):
Let's do it. Let's get into it.

**Alisa Cohn** (00:31:13):
[inaudible 00:31:13]. I'm willing to get into it. I just want to say two things about that. First of all, when you're firing someone, the hope is that it's not a surprise to them. You've had multiple conversations with them that they're not living up to your expectations. It's essential because the truth is, you want to create a culture where people are not surprised by being fired. And that's not even true for this one person you're dealing with. That's true for the entire company. So just kind of getting in the mindset of recognizing that if you shied away from those conversations, kind of like, "You're the problem here and you have some catch up to do."

**Alisa Cohn** (00:31:43):
The second thing is that before you fire someone, I think it's helpful to have the conversation before the firing conversation because something you said Lenny is like, "Oh, but maybe I'm wrong. Maybe I'm not sure." And that bleeds into, "Maybe I haven't been clear with this person." Regularly with my clients, I'll say, "Okay, have you been crystal clear about what you need from this person?" And what they always do is the hand motion like well, sort of, but well, maybe. Which means no, which means no. You've not been crystal clear or you don't perceive even crystal clear. The way to make sure that you're crystal clear is by having the conversation before it comes to that.

**Alisa Cohn** (00:32:23):
What that looks like is, "Listen, Matilda, we have to have a difficult conversation right now. I've talked to you multiple times about coordinating with your peers and not having them surprised about missed deadlines, and I've talked to you multiple times about keeping your team in the loop on different things. After six months of these conversations, I want you to know that the peers continue to feel like that you're operating on your own without coordinating with them. And I continue to hear from your team that they're not fully on the same page. I need you to know that this is very important. I need you to fix this within the next 30 days. Otherwise, I'm sorry to say, we're going to have to find a way to part ways because I can't keep this going with you. I know you have it in you to change. I value all you bring to the table, but if you don't fix these things, we're not going to have a future together."

**Lenny Rachitsky** (00:33:19):
That is very crystal clear.

**Alisa Cohn** (00:33:20):
Yes, crystal clear.

**Lenny Rachitsky** (00:33:22):
Yeah. Okay.

**Alisa Cohn** (00:33:22):
What do you think of that?

**Lenny Rachitsky** (00:33:24):
Yeah, that was great. So it starts with being upfront. This is a difficult conversation, just to set expectations. They're like, "Oh, shit." And then it seems like you come back to, again, multiple times this happened, observing here's what's happening. It's happened multiple times. I keep hearing from multiple people, [inaudible 00:33:44] be a problem. And so it's just like, "I need you to know," and you're just very clear. "Here's what will happen if this doesn't change."

**Alisa Cohn** (00:33:52):
Yes.

**Lenny Rachitsky** (00:33:52):
Yeah. And I love that you also give them a little, there's always that hope for who they are and how you see them as. They're not worthless. It's just like, "You are great at a lot of things. You have these skills. You're great at blah, blah, blah, but still this is a big problem." And it's communicating how critical this is. [inaudible 00:34:08].

**Alisa Cohn** (00:34:08):
Yeah, and it's a deal breaker. It's a deal breaker. Right?

**Lenny Rachitsky** (00:34:10):
Yeah.

**Alisa Cohn** (00:34:12):
If you have so many talents, but if you can't do these two things, then it's a deal breaker for all of us.

**Lenny Rachitsky** (00:34:16):
Yeah.

**Alisa Cohn** (00:34:16):
And I think it's important to really sort of see that both. Sometimes people think, "Well, but I'm so talented." Yeah, but your talents are not going to make up for these two deal breakers.

**Lenny Rachitsky** (00:34:25):
Yeah. And I feel like I know we were going to talk about the firing conversation, but I think this is even more important than that because hopefully, this addresses the problem and you don't need to fire them, which is more valuable.

**Alisa Cohn** (00:34:36):
Yes. Yeah, hopefully. But even if you do, it's actually easier because you've already had the conversation. Right? They're not surprised. It's clear. We've had the discussion.

**Lenny Rachitsky** (00:34:46):
Yeah. So basically the script is like, "There's going to be a difficult conversation. I've seen multiple times this thing and we've talked multiple times and it's still not fixed and here's what I just want to be very clear about." Is there also a script you have for just actually doing the firing or is that less scriptable?

**Alisa Cohn** (00:35:05):
Well, the script for doing the firing is again, please everybody, talk to your HR professional. Talk to your lawyer. Okay, I'm not a lawyer, I'm afraid. So you have to make sure that you're all buttoned up on what you're going to do. But the conversation is actually very simple, which is just, "Matilda, we talked about this multiple times. The last time we had this conversation, I told you I needed you to make these changes. You haven't made these changes and we're going to part ways. So I have here, Sarah from HR or whatever, and we're going to talk through the logistics of that. I'm happy to have a longer conversation with you, but I want you to know we've made the decision to terminate you."

**Lenny Rachitsky** (00:35:42):
Feels very reasonable to me. Is there anything else along these lines?

**Alisa Cohn** (00:35:46):
I think what I want to say is that the conversations you need to have at work are not just difficult conversations. What I call them is sort of delicate conversations because what I think people also shy away from is just simple praise, specific praise. And I think it's really important to get in the habit of pointing out what your people are doing well as carefully as you need to prepare for pointing out what they need to improve. And sometimes leaders feel like, "Yeah, it's all working. It's all working. I don't have to tell you." Or if I do tell you, it's kind of like "Good job." Right? One time a leader or a manager I was doing in a training program, she said, "I don't like getting positive feedback. I only like getting negative feedback." And I said, "How come?" And she said, "Oh, positive feedback is just like, oh, good job. Negative feedback, you can learn something. You get something from it."

**Alisa Cohn** (00:36:40):
So the positive feedback should have the same standard, which is, "I saw the way you ran that launch, it was fantastic. All these different benefits came from it. You're so organized, keep doing that." Or "The way you're keeping your peers in the loop, considering you've only been here three months is extraordinary. I've never seen someone so communicative. It's fantastic. Keep doing that. That's really working for you." If you do that often enough, you do get in the... First of all, it's positive, obviously. You become in the habit of getting better at positive feedback, which is extremely motivating to people at work. It helps them see their progress because that person I just mentioned, she's barely keeping her head above water and she's having trouble fitting in or whatever, but you come around and point out the things that are working. Again, it's very morale boosting. She knows where she stands, and then one day, if you have to give her these difficult messages, you've already sort of laid the reservoir of goodwill.

**Lenny Rachitsky** (00:37:37):
I love giving positive feedback. It's obviously so much easier, but to your point, it's like you have to really think about how to do it well. It's not just a, it's not that easy if you do it well, which is a really good point. And [inaudible 00:37:49] needs scripts for how to give really good positive feedback and have great conversations.

**Alisa Cohn** (00:37:53):
Yeah.

**Lenny Rachitsky** (00:37:53):
That's interesting. There's less demand for that. How do I have better great conversations or compliments?

**Alisa Cohn** (00:37:59):
Right, right. True.

**Lenny Rachitsky** (00:37:59):
**Alisa Cohn** (00:39:09):
[inaudible 00:39:09], I work with a lot of founders and so, don't forget that the entry-level position for a founder is leader, and they have it, they often not had a lot of other experiences being a leader or a manager, and so they're just doing the best they can. It makes sense, right? And they kind of get all this information from other people and their HR leader wants to have a happy engaged workforce and they don't want to upset people for all the reasons we talked about, why you don't want to upset people. Nobody wants to upset people.

**Alisa Cohn** (00:39:40):
And so there's this idea of, they're trying to now be now be the leader who everyone loves and makes people happy. So they would often bend over backwards to make people happy, to keep people, their morale up. But what really needs to happen very often is, we need to drive towards results. And the way this system is working is not going to drive us towards results or this employee continuing to not really do a great job at their job and not really pushing themselves. And you don't want to push them because you don't want to upset them, you don't want to give them difficult feedback, so you're just going to keep hoping it works out.

**Alisa Cohn** (00:40:23):
Ultimately, that leads to the demise of your company. I mean, ultimately right, as you're a startup? If you're not in a startup and you're a large company, it still is very subpar performance, obviously. And you're dancing around hoping and praying they're going to get there and they don't really know there's a problem. And so, I think it's very misguided for leaders to have this notion that their most important role is to keep people happy, is to create this high engagement workforce. High engagement workforce is great.

**Alisa Cohn** (00:40:56):
I think what that comes from is winning culture, which means we're set up for success. We've got the structure for success, we have the culture for success, everyone understands their role, they know the impact of their role. So doing the work to figure out and help them figure out the impact of their role and that when they work together and achieve these milestones, they win and then we celebrate the wins and then we do it all over again. And when you create that kind of a workforce, I think it's much more dynamic, even though sometimes in doing that, you have to redirect people and ruffle their feathers.

**Lenny Rachitsky** (00:41:26):
Essentially, the way I think about it is you think making people happy is not having hard conversations, not pushing them, when really, it's almost working backwards from, if we win and are killing it, people will be happy and what does it take to do that?

**Alisa Cohn** (00:41:40):
A hundred percent. And then the right people are going to want to join your team, people who like to win and like to get results.

**Lenny Rachitsky** (00:41:48):
Is there a story, an example of a founder you worked with or that comes to mind of this kind of where they thought this was their approach and then they shifted? Or is there kind of a pattern you see often?

**Alisa Cohn** (00:41:57):
One company comes to mind. One leader I worked with. Sometimes I think to myself, if I'm writing a book, the book would start with, "It all started with the avocado toast," because he wants to do right by his workforce. And so they have avocado toast at 10 AM, like tea time kind of a thing. And it became this great ritual where people would kind of hang out together and that was great. And then that turned into other longer periods of just hanging out together. Again, these are good things. And that turned into evening socials and everybody was enjoying spending time together, but they continued to be not fully clear on what they were actually supposed to do. And there began to be kind of a cliquey, gossipy culture of who's in and who's out. And that would take up a lot of the socialization time discussions. So rather than talk about expectations about the work and about results, and again, the results were not showing. So it wasn't a lot to celebrate.

**Alisa Cohn** (00:43:05):
They started at a culture committee. So they had a culture committee to talk about how we can make people happier around here. And you can imagine there's now layers and layers of things where we're trying to focus on engagement and we're trying to focus on the employees having a great experience. And the leader I'm working with is completely sincere. It actually want to have a great workplace. But I think the misguidedness was that he hadn't done a great job setting expectations. He had not done a great job of quote unquote "codifying their culture" because culture is not just avocado toast and working together and having socials, culture is things like, we go the extra mile or culture is we make sure, or it could be, we measure twice and cut once. Those are kinds of things that are really about the way we get work done around here. And certainly, a focus on results is like, are we following the process to then get the revenue and to then build a profitable company or are we just kind of hanging out together?

**Alisa Cohn** (00:44:07):
So he had to come to terms with his own discomfort of addressing this with employees and his own discomfort in being a corporate drone of, "Oh, expectations and in the workplace and how we do things." And it turned out that's the whole thing with coaching and with working with people is that you kind of see what their underlying assumptions and beliefs are and there's a reason everyone does what they do. So there's a reason he's doing what he's doing. We had to come to terms with that and then he had to really courageously make some changes about the way he was operating. And ultimately, they had to part ways with one or two really toxic people who were creating this gossipy culture and making people feel not included and not focused on results. And then when they all got on the same page, they were able to gain a lot more traction.

**Lenny Rachitsky** (00:44:56):
I feel like a lot of leaders and founders can relate to this, of wanting to create a great culture and keep it nice and friendly and everyone's a family and then things don't quite work out often in those cases. And there's a shift to, "Okay, we actually need to make a business that works."

**Alisa Cohn** (00:44:56):
Right.

**Lenny Rachitsky** (00:45:12):
It always reminds me, Sheryl Sandberg came to talk at Airbnb once and people are asking, "What do you do with all this...? We're just constantly in chaos. Things are always reorging or changing, just never... I'm on different teams every six months. Our goals are shifting. What do you do with all this... Our culture's changing as we grow." And she's like, "That is a sign of hyper growth and success. And the opposite is even worse when you are not growing and you don't want that. And so you should be happy this is the challenge you're running into."

**Alisa Cohn** (00:45:40):
I love that. It's so true.

**Lenny Rachitsky** (00:45:42):
So along these lines, you talk about how a lot of founders have to come to terms, and it's not just founders, it's just execs and leaders you work with, have to come to terms with, "Here's what I thought leadership was going to be and how to be a great leader, and here's what it really is." Is there anything more there that you find is commonly what they're wrong about or what they miss and what they have to realize?

**Alisa Cohn** (00:46:01):
Yeah. And I think as we grow as leaders, we all have to realize our own blind spots and the difference between what we thought and what is actually going on. So I worked with a founder who she wanted to be was a visionary leader, which is fantastic. I would love that. And she was an incredibly visionary person, very inspirational. But what she didn't see is that what her company needed was somebody to structure and hold people accountable and help them create goals and achieve milestones and course correct when they got off course. And she'd be very frustrated when all those things happened. People got off course, people didn't have goals, people weren't structured to work together. But what she didn't realize was that was, in one way or the other, her job to make that happen.

**Alisa Cohn** (00:46:49):
Now, maybe she needed to have, and I would talk to her a lot about this, a partner, like a COO or somebody else who could be the person who would be sort of managing the internal while she got to be more visionary, inspirational, but ultimately, it was her job to make sure that that was in place. And she didn't sort of see that and she did not adjust her style. And so there's a lot of wheel spinning that happens from that. Even though, by the way, she was an incredibly inspirational person and incredibly inspirational leader and she meant so well. There was nothing malicious about it. It's just that she didn't see the situation for what it was and then adjust.

**Lenny Rachitsky** (00:47:30):
It reminds me, we had this coach on the podcast, Joe Hudson, and he had this phrase that I think a lot of people use, but it just stuck with me. What you resist, persists. So if you hate confrontation, you're going to have much more confrontation. If you hate structure... Actually, this reminds me, Joe Gebbia at Airbnb. He was very anti-process at the beginning of Airbnb. He's like, "We're not going to have a process. I hate process. We're going to run... That's the big company stuff." And then it just chaos constantly. And then eventually it's like, "Okay, we need to have some process to how we build things."

**Alisa Cohn** (00:48:02):
Yeah.

**Lenny Rachitsky** (00:48:03):
And so it's interesting. A lot of people have to realize the thing they think was bad is actually, I see why people do it this way.

**Alisa Cohn** (00:48:10):
Yeah, totally. Actually, I'd like to say something about that because so many... Founders are kind of mavericks and they come into a situation or they start at this company and they want to do things their own way and that's fantastic. Otherwise, they wouldn't be a founder. That's actually fantastic. And so many of the founders I've worked with want to reinvent leadership. Right? They want to have it with no process, they want to have no hierarchy, they want to have autonomy, whatever it is.

**Alisa Cohn** (00:48:37):
And my feeling is like, "God bless. You should absolutely try to do that." But at the end of the day, what happens is, they kind of invent for themselves the understanding that they need to have process, hierarchy, roles and responsibilities, goals, OKRs, whatever it is. And I think it's helpful sometimes to go through that fire of thinking we can do it a different way. But ultimately, I think that the ways to structure a group of people and get them organized so that they can win, are kind of well trod. And I would say that it's helpful to get through that stage quickly so that you don't have to constantly reinvent the wheels of leadership.

**Lenny Rachitsky** (00:49:19):
Such an important context. Obviously, one of the... The most successful founders come up with, have first principles thinking into how to do stuff, and oftentimes they find something no one has ever thought about. So it's always this balance of try a bunch of stuff, a lot of it won't [inaudible 00:49:34]. Some of it was, what will help you win. And I think that's a really good point. I want to get into a couple more tactical things that you often work on with founders. One is, running meetings. Meetings come up a lot on this podcast. People hate them, people love them. There's some are great, some are bad, most are bad. You have some cool advice on just how to make meetings more effective and how to especially end a meeting to help you move forward. Talk about what your advice is there and just generally any advice for better meetings.

**Alisa Cohn** (00:50:01):
Yeah. I'm one of the few people that loves meetings. Or I should say I don't love meetings. I love the potential for meetings. We have all this smart people in the room. We have the potential to talk about these great things and make decisions. And unfortunately, they don't go that way. So what happens often, I mean there's so many downfalls with meetings, but one thing that happens is, we keep meeting. Either we make decisions or we don't make decisions, but then we come back to meet again and we don't have any continuity from the last. So then we re-meet, we re-decide, and that is a big problem. So my three questions to end the meeting are, what did we decide here? Who needs to do what by when? And who else needs to know? And if you can capture those, articulate those as deliverables, I promise you, you're going to have better meetings.

**Lenny Rachitsky** (00:50:51):
Okay, so it's, what did we decide here? Who's going to do what by when? So basically, action items with dates. And then, who needs to know about what we decided here? Is that how you put it?

**Alisa Cohn** (00:51:01):
Yes. Who else needs to know? There's so many executive teams that I've worked with and at first, they go into their room, they have their meeting, they make their decisions and then they leave and they don't tell anyone. "I made this promise for my team that you guys need to kind of go do." Or, "We decided on a policy of some sort and we forgot to tell everybody." And again, no, absolutely no maliciousness, just that they forget or they're too busy and there's not part of the protocol and the process inside of the company that encourages and really insists that people share important information, so cascading that down.

**Alisa Cohn** (00:51:37):
But even the first question, what did we decide here? If you really go around the room at the end of a meeting or six people in the meeting, let's say, and you say to everybody, "What did we decide here?" And they all write it down, you will get six different answers, even though we're in the same meeting. I love that it's so powerful, but also, so helpful to really raise that up, to surface that and then to figure out what to do about it.

**Lenny Rachitsky** (00:51:58):
I love that you highlighted that. I was going to say exactly the same thing, that everyone in their head has the thought of, "Here, oh yeah. Here's what we decided." And to your point, it's often not the same. So is the advice here, is this like a template or something you fill out at the end of a meeting or is it someone's job to make sure these three things happen or how do you operationalize these three questions?

**Alisa Cohn** (00:52:19):
I like it that it's someone's job, the person that I sort of think of as the meetings are. And typically, that's somebody who enjoys follow-up, who enjoys putting lists together and putting things into boxes and whatnot, and there's usually someone like that on the team. And so then it's kind of exciting for them to be the follower upper. But one way or the other, so you could use a template. I think that actually baking it in as a ritual to the meeting, because the other thing about meetings is that we never have enough time. We go right to the end and we don't leave the five or 10 minutes at the end to make sure that we ask these three questions and make sure that we have an understanding of what the follow through is on these meetings.

**Lenny Rachitsky** (00:52:57):
What I'm imagining is, say it's the product managers. Put this doc on the screen in the meeting as the meeting's ending and just have it filled out basically, and just confirm, "Does this look good to everyone?"

**Alisa Cohn** (00:53:09):
Love that. That's a great way to do it. By the way, with... Well, I just would say what's interesting about that is that if we ask people what did we decide here, I think there's value in just asking that question in particular because somebody might say, "We decided," I don't know, "Something." And other people would say, "No, we didn't. But that's actually a good idea. It sort of crystallizes what we did talk about in a more comprehensive way." I think there's value in raising the differences and I think there's value in stitching those together. So just putting it up on the board is good, especially if you're running short of time. I worry that somebody might not weigh in and say, "Actually, I have a very different point of view of what we decided here." So maybe it's also about building the culture to break in and say, "No, that's not what I see. Let's spend some time on that."

**Lenny Rachitsky** (00:54:03):
Let's actually spend more time on this because this is really, I think, really this specific detail I think could be really powerful if you do it right. So say you're the PM in the meeting, who do you ask? Do you say to the room, "What did we decide here?" Or do you look at the most senior person? Otherwise, it feels like it could just lead to a whole discussion the last couple of minutes, which I guess could be valuable, but who do you point this question to?

**Alisa Cohn** (00:54:24):
Yeah. So I picture this for let's say, a six-person executive team meeting, which means everyone go around quickly and say, "What did we decide here?" Now, if you're in a meeting with a large executive team, which I do work with sometimes or non-executive team, like a group of some sort, then you probably want to get a few people just to... I would just even say as a facilitator, two or three people, "Okay, two or three people, what do we decide here?" And if you can kind of get common, great. That's fantastic.

**Lenny Rachitsky** (00:54:52):
Got it. Okay. So if it's a small meeting, go around the room and everyone just shares, here's what we decided here. And they could just be like, "Yep, he's got it or she's got it." Awesome. Okay. This is great. So the advice here is, next time you have a meeting, especially an exec meeting, just at the end of the meeting, you, the listener of this podcast, just ask, "Okay, everyone. Let's just make sure we're on the same page. What did you decide here? Who needs to do what by when?" And then everyone chimes in and you're writing this in this doc, and then what else? Who needs to know about what we decided here?

**Alisa Cohn** (00:55:22):
Yeah. Lenny, I love that because also, do you have to be the leader of the meeting to do that? No. You could just be the person in the meeting and just chime in and just start it yourself. And if you do that and everyone kind of picks it up, it can become a ritual just by virtue of your own agency. So I love that you just encouraged everyone to do that.

**Lenny Rachitsky** (00:55:42):
And this is how you become a leader, is you just start doing these things and people are like, "Oh, Alisa is so helpful. She's just on top of it. Every time she's in a meeting, the meetings go better. We get things done." So I think just doing the thing that is useful to everyone is how you move up.

**Alisa Cohn** (00:55:56):
Exactly.

**Lenny Rachitsky** (00:55:57):
Amazing. Okay. Another topic that I know you spend a lot of time on is something you call the founder prenup. And what I love about this is, a lot of the problems that a company trickle down from the founders having their challenges with each other. And I started a company in the past and I don't think people realize how significant this decision is in your life. It's basically, you are marrying someone in a business context and you're stuck with this person for a long time and you basically came up with a prenup, which is a set of questions of just things you need to talk about to make sure you're aligned before you start this company. Is there any context around this thing before we talk through actually the questions that you recommend people talk through?

**Alisa Cohn** (00:56:41):
Well, I just want to reiterate what you just said. Exactly right. And it turns out that according to Noam, Noam Wasserstein, 65% of startups fail because of conflict with founders or the founding team. So it's really essential to get this right, and I agree that people step into this relationship with a lot less care than they should. And bad things can happen because you haven't done the work of getting to know each other before you decide to co-found.

**Lenny Rachitsky** (00:57:12):
Yeah. It's so easy just to like, "Yeah, I'll start a company. We have this cool idea. Let's just do it. It's going to be so awesome." And then you don't realize how much you're committing to and how often things don't work out because of that quick decision. And oftentimes, it's like friends and then it becomes even more challenging because I want to be friends, but we're in business together. So yeah. Let's talk about what you recommend folks talk through as much as we can on this podcast.

**Alisa Cohn** (00:57:36):
So I do have kind of an extensive questionnaire, so we just touch on a few things, but one thing I think first and foremost is, what are your values? And I think it's really essential to do some sort of values clarification exercise. You can find a ton of them online. You can find a list of values and just pull out your core values and just compare them with each other because when you are aligned, it's great. Or when you're adjacent, it's also great.

**Alisa Cohn** (00:58:01):
I might care a lot about excellence, Lenny, you might care a lot about learning. Fantastic. Those are great values that we can kind of, go together. I might care about excellence and you might care about work-life balance. Wow, let's talk about that because I think it's going to be really important as we go through our startup journey that we understand both of us, what does work-life balance mean and what does excellence mean? Because those two things can at times be at odds with each other, just as kind of an example.

**Alisa Cohn** (00:58:30):
So talking through those core values in advance and updating them regularly, even as you go down the path together is so essential. Just so you know where the other person's coming from. Because the other problem is, someone acts in a certain way, you don't know them that well maybe, or maybe you've known them as an eighth grader. A lot of founders do know each other from their youth and they've matured into different kinds of people. And so you think they're acting strangely, but actually, they're acting in accordance with their values. And so getting a handle on that upfront can solve, I would just say, solve a lot of problems before they start.

**Lenny Rachitsky** (00:59:08):
So signs that your values don't align. It's basically you both can't be true is almost the way I think about it as we talk. It's hard to be the excellent, focus on excellence and also not work long hours, which it's possible, but it's hard. Those are challenging and worth the conversation.

**Alisa Cohn** (00:59:26):
Yeah, worth the conversation because in fact, as you say that, I'm like, "Well, I guess you can do that. Right. You can do that." And so therefore, that's where the conversation has to figure out how you're going to marry these two values, which might be at odds or might be aligned, but let's talk through what work-life balance means to you and let's talk through what excellence means to me, and let's see if we can have a meeting of the minds about it or at least I know where you stand.

One of the founders I worked with, he would text or Slack his co-founder on weekends and the co-founder wouldn't respond. And that was extremely frustrating to the person, to the co-founder I was talking to. And it turned out, after they finally addressed it, it really was about wanting to have some downtime and some, quote unquote, "Balance." Nothing wrong with that, but because they didn't talk about it, both sides made [inaudible 01:00:20] big assumption about it and then it caused this conflict that didn't have to happen if they'd had the conversation in advance.

**Lenny Rachitsky** (01:00:25):
Comes back to where we started of having these conversations is necessary and almost helps the other person because this small issue could become a huge issue over time, if you just start assuming and it keeps happening and it keeps scratching and scratching at you. And letting that person's [inaudible 01:00:45] is screwed up because you're, "I can't do this with you anymore." Right? So it's just another reminder of how it's good for the other person for you to engage in a difficult conversation.

**Alisa Cohn** (01:00:54):
Yes, very true.

**Lenny Rachitsky** (01:00:55):
Okay, what else? So values. By the way, is there a values framework you most love that you can point people to or there just a bunch and don't worry too much about which one you go [inaudible 01:01:05]?

**Alisa Cohn** (01:01:04):
I mean, the one I use is super simple, which is on the thing called the internet. There's a lot of lists of values and I think when you see a list of values, you can pull out the ones that are most meaningful to you, and that's a very simple and helpful and free tool.

**Lenny Rachitsky** (01:01:19):
Got it. So you just Google list of values, there's a PDF, and just circle the ones that are most and pick whatever small number, don't... Half of [inaudible 01:01:27]-

**Alisa Cohn** (01:01:26):
Actually, well, just to give you the process, right? It's helpful to pick 20, for example. Great. And then you winnow them down to, let's say, 10. And then you do the difficult work of winnowing them down to three to five that you feel are core to you. And that's a good exercise for everyone to do actually, every year because things can change. It also forces you to make the difficult decisions about when it comes down to it, what are the things that really are important to me? The more you know your values, the more you can operate in the world with just more clarity for yourself.

**Lenny Rachitsky** (01:01:58):
Awesome. All right. So values. What else?

**Alisa Cohn** (01:02:01):
Yeah. So another one is, vision of the company. So when this company is successful, what does that look like? And what that might look like is, we're in control of our destiny and we are able to operate this business independently and we have a lot of freedom. What that might look like is a big venture outcome that we all read about. And if you are both assuming that you both think the same thing but aren't talking about it explicitly or talking about the trade-offs you need to make inherent in that, then what often happens if you have differences is they come home to roost while it's too late or when it's too late.

**Alisa Cohn** (01:02:40):
So an example is the two co-founders I worked with, one of them would said to me wistfully, this is like five or six years into the company, and the company was going well, but it was challenging and they had all their growing pains and like you mentioned about Sheryl said all the chaos. And he said to me, "Gosh, I don't see why we have to grow. I just wish we could actually have fewer employees. And I used to love it when I knew everybody's name and I would just much prefer an environment where we didn't have to grow." Well, unfortunately, they were already venture backed and also, the other co-founder had a very lofty ambition for a very big company. And since they hadn't talked about that, it was way too late to even have that conversation and it was a very painful reckoning for both of them to realize they were not on the same page.

**Lenny Rachitsky** (01:03:33):
Totally, see the value of this one. I could totally see how people would have different goals. I imagine it also changes over time, so there's probably an element of, if something has shifted for you, you should probably also have that conversation. I don't want to build an IPO venture scale business, I just want to build something chill. So basically, a line on what is... How would you phrase that? What does winning look like to you?

**Alisa Cohn** (01:03:52):
Yeah, what does success look like?

**Lenny Rachitsky** (01:03:54):
What does success look like to you?

**Alisa Cohn** (01:03:54):
Or what's the vision for the company when it reaches its full potential?

**Lenny Rachitsky** (01:04:01):
Okay. Great. What else?

**Alisa Cohn** (01:04:01):
Another one is, it's sort of a two part question. How do you handle conflict? So how do you handle conflict? But then, you might want to ask your spouse, someone close to you, "How do I handle conflict?" Because you might think, "Oh, I handle conflict with such an enlightened person. I'm so neutral about it. I'm so great at bringing things up." But the person who's close to you might say, "You seethe until you're ready to bring something up and it's really uncomfortable in the seething period." So it just gives you a little more self-awareness about how you actually handle conflict.

**Alisa Cohn** (01:04:42):
And that's really important because I might be the kind of person who wants to bring up conflict and talk about it immediately. The other person might be a person who totally wants to talk about the conflict but wants to let it settle first and wants to also go through their own thinking process about what's important to them and might actually feel like they've resolved it themselves without having to have a conversation with you. And if you're the person who's like, "Let's talk about it, let's talk about it, let's talk about it." And they're like, "I'm working through it myself." Now you have conflict over the conflict and it just turns into dynamic that's not necessary.

**Lenny Rachitsky** (01:05:22):
As you go through these questions, it's absurd to imagine people don't do this when they find a co-founder and work through stuff, and I know nobody does. The percentage of people that do this sort of work ahead of time, it's very low. And so I love that we're helping this percentage go up, but it also reminds me of just how crazy it is people don't have these conversations and how it explains why so many founder relationships don't work out. So these are awesome. What else? I know you have a whole list and we'll link to it, right? There's a PDF we can link to?

**Alisa Cohn** (01:05:22):
Yes.

**Lenny Rachitsky** (01:05:53):
With the questions or-

**Alisa Cohn** (01:05:54):
For sure.

**Lenny Rachitsky** (01:05:55):
[inaudible 01:05:55] post. Awesome. Let's do a few more.

**Alisa Cohn** (01:05:57):
Another one is, how do we decide when we disagree? And that is a very good thing to explore because there's actually a lot of different ways to decide when you disagree and they're all good. And if you have it sort of upfront and it's just for an ongoing discussion, but if you have it up front like when we disagree, because that's definitely going to happen, let's assume that the person who cares the most can win that argument. That would be a great way to do it. It might be, the person who's got the best perspective and the most expertise can win that argument. It might be, we'll go back and forth when we really disagree. First you win and then I win, like that, back and forth.

**Alisa Cohn** (01:06:41):
There's so many different ways to handle it and if you talk about it upfront, you'll be much more likely to be able to actually put that into practice when you do disagree because you will definitely disagree. There's no way around that. And that's not even a bad thing. You're smart people. You have this dynamic tension in the relationship. You bring different things to the table. You've got different perspectives. Disagreeing is normal. Working through it and having a practice and a process of working through it, will help it be a good conversation rather than this sort of sulky difficult conversation.

**Lenny Rachitsky** (01:07:12):
I love it. Maybe one more?

**Alisa Cohn** (01:07:14):
Yeah. So another one is, what kind of company culture do I think is important? People definitely don't talk about this before they found the company and they assume they're on the same page. So one founder might be, "I want to have this great company where everyone loves it and we're all loving together and working hard together. And it feels like a..." To use your word before, "It feels like a family." By the way, that's great. That's fantastic. "I want to have a get it done, results-focused culture where we're just executing the hell out of everything and that we're just focused on winning."

**Alisa Cohn** (01:07:52):
By the way, those two can actually exist together. But if you're pushing in one direction without the other and your co-founder is pushing the other direction without yours, it really can feel like two different companies. And that's... When I go into a situation at one of my client sites, often I will hear from the employees, "It feels like we have two different companies and two different cultures depending on whose team you're on." And that, of course, leads to lack of coherent working together and certainly even just lack of different standards and expectations.

**Lenny Rachitsky** (01:08:24):
Awesome. Okay. To kind of start to wrap our conversation, I want to take us to a recurring segment of this podcast that I call, Fail Corner. We've talked a lot about failure at this point and just all the ways people fail. I'm curious if, in your career or life, there's a story that might be helpful for folks to hear when things didn't go great and you've failed, and if you learn something from that experience. And the reason this is something I do is I feel like people listening to this podcast, everyone's like, "Sounds so amazing, everything's always going great. They're killing it." When in reality, that's not actually how things go. So these end up being really helpful for people like, "Oh, wow. Even Alisa had a really hard time sometime." Is there a story that you could share?

**Alisa Cohn** (01:09:04):
Absolutely. I mean, so many examples. I'm going to give two quick examples. One is, when I first started my coaching practice, I just kind of started and so I just did everything I could to get clients, to build a business, to build a practice, to build my brand, all the things. And I was working so hard and I think I'd had this conversation with somebody that didn't go very well. And I just thought, in my mind's eye, I thought, "Well, what will become of me?" That was my voice in my head for quite a long time, "What will become of me?" And I was living in Boston at the time. I got onto the floor, my hardwood floors in my Brookline condo, and I just balled in the fetal position. I just balled and balled and balled for an hour. It wasn't 10 minutes, it was an hour.

**Alisa Cohn** (01:09:50):
And I was so frightened and just upset. Am I going to be able to make this work? And it was a while and I got into the couch and took a little stress nap. And then I got up from my stress nap and I just started making more calls and doing more things. And that was definitely a rock bottom moment for me. And I think what I learned is, you have to literally pick yourself up from the ground and pull yourself forward. And when you keep taking action, action, action, win or lose, win or lose, you'll get where you need to go. And that turned out to be true. But in those moments, I was not thinking that was going to turn out to be true.

**Lenny Rachitsky** (01:10:32):
Wow. Amazing story. I imagine many people feel those moments and it's empowering to hear that it can all turn out really well, even when you're lying in the floor crying for an hour. An hour is a long time to cry on the floor.

**Alisa Cohn** (01:10:45):
It is a long time to cry. It really... I thought about it because most people just cry for 10 or 15 minutes. I was crying for an hour. I'm positive. Yeah.

**Lenny Rachitsky** (01:10:52):
Great story. You said you had another story.

**Alisa Cohn** (01:10:56):
Yeah. I'll tell you a second story, which is more focused on actually my work life. So one thing that I do is I do coaching of course and I do off sites. And this was early, early days of my coaching career and I was doing this off site and it wasn't going well. And I was debriefing with my client during the breaks and at one point she said something like, "I just think we should end this offsite. I just think we should just decide it's over and it's not working." And I felt horrible, obviously, humiliated, certainly, and just like that's a failure. That's like, "Oh, fail." And I know that what I took away from it was that I can improve my skills in every aspect of running an offsite.

**Alisa Cohn** (01:11:40):
So getting aligned with the client in advance, making sure that I had the right activities getting us to our goal, being very goal-oriented and focused, and making sure that I had kind of understood the rhythm of what it takes to bring people together. So I took some training on that. I worked my mentor on that, and I got so great at offsites after that experience. I'll tell you that was a real low because in the moment, in that moment, I'm not thinking, "I'm going to get great at offsites." In that moment I'm thinking, "Oh, my God. I'm going to get... What will become of me?" But I turned it into, in my mind's eye, or I should say, I turned it into the ability to build my skills. And I just want to tell everybody, even at your lowest moments, anything that you're learning from that, can then be turned into fuel to build your skills to get great at the thing that you're not great at.

**Lenny Rachitsky** (01:12:31):
What I also love about this is there's this feeling of imposter syndrome, is specifically this fear that I do something wrong and it'll all crumble and everyone will see I suck and I never... I don't know anything and everyone will see it. And I love both these stories are like, it doesn't go well and doesn't crumble. You build from there. And no one's like, "Oh, Alisa's terrible forever." No, it's like move on to the next thing. And then you use that as fuel to become really good at this thing that didn't go great.

**Alisa Cohn** (01:12:57):
Yeah, that's really well said.

**Lenny Rachitsky** (01:13:00):
Amazing. Alisa, we covered a lot of stuff. Is there anything that you were hoping to cover or you think might be useful for folks to hear before we move to our very exciting lightning round?

**Alisa Cohn** (01:13:10):
The only last thing I want to talk about, just sort of circling back to your role as a leader, I was one time working with the CEO who was handling the fact that this launch was not going well, as in the launch wasn't happening. [inaudible 01:13:25] foot off, foot off, foot off. And his point of view was, you need to have patience with it as it goes. And my point of view is, because I've talked to a lot of the people around, was that there was a massive process problem going on that he was not kind of touching into and really investigating because the product manager wasn't experienced, was kind of hiding it because he knew he didn't have the skills, was fighting with engineering, it just wasn't working.

**Alisa Cohn** (01:13:53):
And when the CEO was telling me, and we really had a long discussion about this where I kind of enlightened him about some of the issues that he needed to get involved and fix, he kept thinking, "I need to have patience." So what I want to say to everybody is, sometimes you need to have patience and sometimes you need to look at the process. And I think you, as the leader, need to have the wisdom to know the difference, but also your finger on the pulse to recognize, is this an issue with patience or an issue with process?

**Lenny Rachitsky** (01:14:22):
I guess, is there a sign that you're like, it's probably a process thing and you're just ignoring a glaring problem that everyone else sees?

**Alisa Cohn** (01:14:30):
I think the sign is when, if you search your mind, you don't really know how this thing is going to come together. There's no plan in your mind. You haven't touched in with people or talked to people about what's going on. You kind of hear this uncomfortable silence about it. Those are symptoms that you just need to dive deeper and just be a little more in touch with what's going on and talk to some folks and look at some data. And by the way, it might not be a massive process problem. It might just be one little thing that needs to get unstuck but you, as the leader, need to recognize that and figure out a way to make that unstuck. And if there's, of course, a big problem that needs to somehow be just surfaced.

**Lenny Rachitsky** (01:15:09):
So if there's this hope this'll work out versus I see a path to this working out, it's probably a problem. Awesome.

**Alisa Cohn** (01:15:16):
Yeah, well said.

**Lenny Rachitsky** (01:15:17):
Is there anything else that you wanted to share or touch on that you think might be helpful?

**Alisa Cohn** (01:15:21):
We talked a little about the co-founders prenup, which I think people would think, "Well, I'm not a co-founder, I don't need that." I just want to invite everyone to also think about a different tool that I have, which is called the Personal Operating Manual. And it helps prompt you to talk about working style together because you may not be co-founders, of course, but you're working on a team with a bunch of people and they all have their different working style.

**Alisa Cohn** (01:15:42):
So it's kinds of questions like, what communication style do you like the best? How do you like to work? Do you like large uninterrupted blocks? Do you like meetings here and there? When I'm trying to get a hold of you for something important, what's the best way to do that? What is one of your pet peeves or some of your pet peeves? How can I get a gold star with you? Also, this is my favorite. What's your delegation style?

**Alisa Cohn** (01:16:10):
Do you want me to check in with you regularly, like once a week as I'm working down the path of a project? Or do you want me to just let you know when it's done and just tell you at the end that it's been complete? So lots of different ways people assume other people work because it's like your style, but actually it's just your style. So those kinds of conversations can be great for working together and also be a great team activity.

**Lenny Rachitsky** (01:16:34):
So this kind of what goes into these READMEs people put together, here's how to-

**Alisa Cohn** (01:16:37):
Yes.

**Lenny Rachitsky** (01:16:37):
... work with me. I really love the gold star concept because I feel like people want to know how do I be super awesome? How do I be really successful working for you? And I like that visual of the gold star and the pet peeves. I feel like a lot of people will identify that. What are my pet peeves so that people don't do these things because they don't know, right? They don't know until you tell them.

**Alisa Cohn** (01:16:55):
Nobody knows what's your operating style until you tell them. And the more you can showcase, the more everybody will be able to do it right for you and you'll be able to do it right for them. And then you'll be able to have better workplace harmony and save your conflict with things that are really important. Not just because like, "Oh, you didn't text me when I wanted you to text me."

**Lenny Rachitsky** (01:17:13):
Being clear. What do you know? Is there anything else that you think might be helpful to share before we get to a very exciting lightning round?

**Alisa Cohn** (01:17:20):
No, just that.

**Lenny Rachitsky** (01:17:21):
Well, with that, Alisa, we reached our very exciting lightning round. Are you ready?

**Alisa Cohn** (01:17:24):
I can't wait. I'm ready.

**Lenny Rachitsky** (01:17:26):
Here we go. First question, are there two or three books that you find yourself most recommending to other people?

**Alisa Cohn** (01:17:33):
So we already talked about Kim Scott, the wonderful, amazing Kim Scott and her book, Radical Candor, is one I recommend a lot to people. It's fantastic. Working Backwards by gosh, Colin Bryar and Bill something, is about sort of the Amazon way of working backwards from the customer. Super geeky and tactical. I love it. I slurp it up like Harry Potter. It's so good. And I definitely recommend to my clients about Amazon's Management Science. And the third is Walt Disney by Neil Gabler because it really shows how Walt Disney, sort of it's everything about his youth and how he turned into a very bad entrepreneur and ultimately into a fantastic inventive entrepreneur. And it shows all the origins of how he invented these different pieces that now make up the Walt Disney Company.

**Lenny Rachitsky** (01:18:26):
The first two recommendations we've had on the podcast, Kim Scott and Bill Carr, is the other-

**Alisa Cohn** (01:18:29):
Bill Carr.

**Lenny Rachitsky** (01:18:30):
... co-author. He's been on the podcast and people love that episode. I haven't had Walt Disney on. I got to work on that.

**Alisa Cohn** (01:18:37):
Or the writer, Neil Gabler

**Lenny Rachitsky** (01:18:38):
Or the writer. Yeah, yeah. Good tip. Okay, next question. Is there a favorite recent movie or TV show you really enjoy?

**Alisa Cohn** (01:18:44):
Yeah, I enjoyed Inside Out 2. I thought it was fantastic, the idea [inaudible 01:18:49].

**Lenny Rachitsky** (01:18:48):
I could see why you love it. I feel like it's for all coaches in the world.

**Alisa Cohn** (01:18:51):
Totally. Just the idea that like, oh yeah, we're all this complex stew of emotions and it's okay.

**Lenny Rachitsky** (01:18:56):
Mm-hmm. I also love that movie. Next question. Do you have a favorite product you recently discovered that you really love?

**Alisa Cohn** (01:19:03):
Yes, the Ninja Creami. So good.

**Lenny Rachitsky** (01:19:07):
Say more.

**Alisa Cohn** (01:19:09):
The Ninja Creami turns anything into ice cream. So you can actually make ice cream. Good, God bless. But I take my protein shake, which is okay, and turn it into ice cream, which is delicious. And it takes 10 minutes and very little prep, and it's simple to use and it works as expected, which so many things do not. The Ninja Creami, go get it.

**Lenny Rachitsky** (01:19:28):
That's the first for the Ninja Creami. And I love, the holidays are coming around, so this is going to be good for people. Do you have a favorite life motto that you often come back to you find useful in work or in life?

**Alisa Cohn** (01:19:41):
This quote by Joseph Campbell animates my life, which is, "If you can see your path all the way through to the end, you are following someone else's path. Your path only becomes clear moment by moment as each foot hits the ground."

**Lenny Rachitsky** (01:19:57):
Wow, that's so good. It's so empowering because it helps you realize if you don't see where it's all going, that's normal and that's good. Wow. Great one, good one. I need to do something with all these mottos. They're so good. I need to create a poster or something.

**Alisa Cohn** (01:20:14):
That's a great idea. Or your newsletter.

**Lenny Rachitsky** (01:20:16):
Here we go.

**Alisa Cohn** (01:20:16):
Send them out.

**Lenny Rachitsky** (01:20:18):
Yeah, that's the easy path. Okay. Last question. So I'm curious, and not to create more competition for you, but I feel like a lot of people think about becoming a coach of some kind, like a product coach, exec coach. If someone is thinking about going down that path, is there one piece of advice you could share to help them pursue this path, even explore if it's right for them?

**Alisa Cohn** (01:20:38):
If you think you want to become a coach and you immediately want to build up your coaching skills, listen to people more deeply and ask deeper questions, not just respond to what they just said, but why do you think that? Or where is that coming from? And you will see if you enjoy that process of really going deeper with people. I think that would be helpful for everyone to do. But certainly if you want to become a coach, I think that's essential to be able to get really beneath the surface.

**Lenny Rachitsky** (01:21:13):
I love how your energy just changed into coaching mode when you said that. I love that. That was such an interesting thing to see and that was great advice. That's easier said than done. And it's interesting, you could tell people are so good at that specific skill versus not. And so I love that that's the thing to work on, is ask better questions, think deeper about the person and what they're coming from. Alisa, this was incredible. Two final questions. Where can folks find you if they want to reach out, maybe work with you, what kind of people do you work with in case people are interested in that, and finally, how can listeners be useful to you?

**Alisa Cohn** (01:21:47):
Oh, thank you. Well, I work with executives at startups and also at large public companies, so feel free to reach out if you want to have a conversation about coaching. And you can find me at alisacohn.com. And actually, I'm going to take some resources and put them at a special link, which is alisacohn.com/lenny. If you want to download the Co-Founder Prenup. I also have a Personal Operating Manual and a few other resources I will put there. So alisacohn.com/lenny and you can also join my newsletter from there.

**Alisa Cohn** (01:22:20):
And I think in terms of helping me, I guess there's two things I want to say. My life's work genuinely is to make a difference. When I became a coach, it was because the music in my head was to make a difference. And so I hope I've made a difference for all of you today and I would invite you to try one thing that makes you uncomfortable, this week. As soon as you hear this, this week, try something that makes you uncomfortable and feel free to let me know on LinkedIn or even send me an email and let me know what you did that made you uncomfortable.

**Alisa Cohn** (01:22:51):
So that would be very meaningful to me. And the second thing that would be very meaningful to me is if you would go find my podcast called, From Start-Up to Grown-Up and give it a listen. Maybe give it a rating and review because as you know, Lenny, the way people find your podcast is when other people are interested in your podcast.

**Lenny Rachitsky** (01:23:09):
From Start-Up to Grown-Up. I love that title.

**Alisa Cohn** (01:23:11):
Thank you.

**Lenny Rachitsky** (01:23:12):
Alisa, thank you so much for being here. This was awesome.

**Alisa Cohn** (01:23:15):
Thank you so much for having me, Lenny. It was great.

**Lenny Rachitsky** (01:23:17):
Bye everyone. Thank you so much for listening. If you found this valuable, you can subscribe to the show on Apple Podcasts, Spotify, or your favorite podcast app. Also, please consider giving us a rating or leaving a review as that really helps other listeners find the podcast. You can find all past episodes or learn more about the show at lennyspodcast.com. See you in the next episode.

---

## [2/15] Behind the founder: Drew Houston (Dropbox)

**Lenny Rachitsky** (00:00:00):
People just don't realize the wild journey that you have been on over the past 18 years building this company. It feels like there's almost been these three eras of Dropbox. The first era of you're killing it.

**Drew Houston** (00:00:10):
For the first several years, it was doubling, 10-xing every year. Taping user counts that we printed out to the wall, and then running out of space on the wall. Having to put 100,000 users, 200,000, 500,000, 1,000,000, 10,000,000 on the ceiling.

**Lenny Rachitsky** (00:00:23):
And there's the second era, which I'll just say, everyone's trying to kill you.

**Drew Houston** (00:00:26):
We started getting all the incumbents. Apple, Microsoft, Google. All of them launched competing products, but weirdly, it was sort of like you see the videos where there's the mushroom cloud in the distance. You see it. But you don't hear, or notice it. It was also clear that winter was coming.

**Lenny Rachitsky** (00:00:41):
It feels like the year 2015 was a pivotal year where things started to shift.

**Drew Houston** (00:00:45):
I'd start to hear a louder set of critics inside, and outside the company. Less than a year later, Google Photos launches. And not only does it provide a lot of the same value, but they also gave you free unlimited storage for life. And so, they just totally nuked our business model.

**Lenny Rachitsky** (00:01:02):
You end up fighting wars on three or four fronts against the big kahunas that have infinite cash, and can do whatever they want.

**Drew Houston** (00:01:08):
So it killed Carousel, killed Mailbox, went all-in on productivity. And I wish I could say, "Then, everything got better." It was the opposite, actually. The narrative completely flipped on the company. Suddenly, your employees don't want to wear your T-shirt anymore. Everybody's looking to you, and is wondering, "How the hell did you get us in this situation?"

**Lenny Rachitsky** (00:01:28):
Today, my guest is Drew Houston. This may be the most interesting, and most useful episode of my podcast so far. Especially if you're a founder, or if you someday want to be a founder. Drew shares the very real talk story of what it's been like to build Dropbox over the past 18 years. Including the ups, and especially the downs. He shares stories that he's never shared before, the struggles he's been through that very few people know about, what it's like to compete with big tech, how he's thinking about the future of the company. And also, what he's learned about himself throughout the journey.

**Drew Houston** (00:04:51):
Oh, thank you, Lenny. It's great to be here.

**Lenny Rachitsky** (00:04:53):
I have been so looking forward to this conversation for so many reasons. One, is I feel like people just don't realize the wild journey that you have been on over the past 18 years building this company. You told me a few of these stories when we had dinner recently, and I was just like, "You need to come on the podcast, and tell this stuff. I think it'll be really useful to a lot of people."

**Lenny Rachitsky** (00:05:12):
And then also, I feel like you're just a very real talk founder that isn't afraid to share what's really going on. I think a lot of founders don't actually tell you the things they're going through. It's always, "Killing it. We're killing it all the time." So for all those reasons, I'm really excited to have this chat, and this opportunity to stand with you and hear these stories. So thank you again for doing this.

**Drew Houston** (00:05:29):
Me too, and I've been a big fan of the podcast.

**Lenny Rachitsky** (00:05:31):
Oh.

**Drew Houston** (00:05:32):
I learned a lot.

**Lenny Rachitsky** (00:05:33):
Oh, wow. I appreciate that. The way I'm thinking we structure this conversation is, as an outsider, it feels like there's almost been these three eras of Dropbox. And tell me if I am missing something, but essentially, it feels like there's the first era of you're killing it. Just up and to the right killing it. Dropbox is on fire.

**Lenny Rachitsky** (00:05:52):
And then, there's the second era that I think fewer people know about, which I'll just say, everyone was trying to kill you. All the incumbents are coming after you. I like that it leads you to sigh. This is going to be good. And then, there's the current era. I'll say the third era of just rethinking what Dropbox could be. Does that sound roughly right? That's a good way to think about that?

**Drew Houston** (00:06:11):
I think that's right. Yep.

**Lenny Rachitsky** (00:06:12):
Okay, awesome. So let's start with this first era, and spend some time here. Which is, I think, the era most people know you for. There's the big Hacker News launch, the demo video of the thumb drive you're always losing, the referral program everyone's always studying. You're almost the epitome of viral growth. And so, let's just talk about just things here that maybe people don't know about. Maybe some moments that are really important to you, memories that stand out to you. So maybe just start wherever you want to start about this time of the history of Dropbox.

**Drew Houston** (00:06:41):
I mean, I started Dropbox more out of just personal frustration, and it really felt like something that only I was super interested in as far as file syncing. And focusing on one customer, which is myself. And then there's the story of me forgetting my thumb drive on a trip to New York, and things like that, and coding. I won't get into all that. I had a lot of friends who were in Y Combinator, and many friends who had moved from Boston where I was living out to California. And doing that whole pilgrimage, or kind of maybe one-way pilgrimage.

**Drew Houston** (00:07:14):
And feeling a little bit left out, but then also having this idea for Dropbox. And then I think some of the most memorable things were the moments where Dropbox really felt like it was taking on a life of its own, or maybe it belonged to the internet, and not so much to me anymore. So for example, we had a lot of success with these demo videos. First, was just getting into Y Combinator. It sort of worked backwards from thinking about, "What does Paul Graham do all day?" And my hypothesis was that he just hits refresh on Hacker News like everyone else. Like me, like everyone else.

**Lenny Rachitsky** (00:07:46):
This was you trying to figure out how to get Paul Graham's attention?

**Drew Houston** (00:07:49):
Yeah.

**Lenny Rachitsky** (00:07:49):
I love that. "What does Paul Graham do all day?" Okay.

**Drew Houston** (00:07:52):
Yeah. Well, my first company was doing online SAT prep, and I was 21 when I started that. But in the world, there's a lot that Y Combinator has in common with college admissions. You have 1,000,000 people applying for very few spots. And the more you can get some kind of hook, or find some kind of side door. That was the thinking, and I was like, "All right. Well, if I can create some kind of viral video, put it on Hacker News, get Paul's attention, that'd be one way to do it."

**Drew Houston** (00:08:21):
And that was inspired by a book called Guerrilla Marketing that I had read, which is basically how to do marketing if you have no money. Which was a good fit for where we were. I say, "We." It was just me at the time. And sure enough, I made this video that was basically a pretty straightforward screencast of Dropbox. Showing it working on my computer, and then it hit the top of Hacker News for two days. I don't think it can really happen anymore, but sure enough, we got a note from Paul saying, "Hey, this is interesting, but you need a co-founder."

**Drew Houston** (00:08:56):
Which was a problem because it's clear that the YC application deadline for the next cycle was maybe a week or two away. So Paul was basically sending me a helpful note that, "I know you're not dating anyone, but you need to be married in the next two weeks if you want to get into YC." So I ended up finding my co-founder, Arash, and there's just story after story like that in the early days. So I'd say the first chapter was characterized by just feels like one moment I'm sort of paddling in the ocean alone on a little board. The next, I'm just like a hundred feet off the ground on this tidal wave trying to stay on.

**Lenny Rachitsky** (00:09:32):
How long does this period last of just Hacker News to something starts to change? How long is this up and to the right period?

**Drew Houston** (00:09:40):
It was the first several years. So I started the company back in 2007. I might have been 24 at the time, moved to California. And so, I'd say it was first probably seven years from 2007 to 2014 were really that crazy fever dream .com experience where it's a blur. I mean, basically, get into Y Combinator. After we finish Y Combinator, that culminates in Demo Day, and getting your first investors. At Demo Day was this guy named Pejman Nozad who is an angel investor who also owned a rug store in Palo Alto. He runs Pear there.

**Lenny Rachitsky** (00:10:24):
Oh, and he runs Pear. [inaudible 00:10:25].

**Drew Houston** (00:10:24):
Yeah, and he runs Pear now. He introduced us to Sequoia. He came with us to the pitch, which I learned later it was unusual. And that was on a Friday, and then Saturday Mike Moritz is in our apartment. But anyway, so we raised money from Sequoia, and that was a seed round in 2007. So right after we finished Y Combinator. And then, basically, you start with a pretty narrow circle of what you're working on.

**Drew Houston** (00:10:53):
I mean, right in the beginning it's just really just coding, and talking to customers. But that circle kept expanding pretty rapidly. So in 2007, we were just building the first prototype of the product. It was in closed beta for about a year. And in end of 2008, we launched at what's now TechCrunch Disrupt. Our demo totally failed. The wifi wasn't working on stage, so a live demo is quite underwhelming. So that took a few years off my life, but fortunately, we had accumulated this big beta waiting list from basically doing another version of that Hacker News video.

Engineered to be even more viral, and have all these memes, and things in it. But similarly, a few minute demo video of, "What's Dropbox? And here's moving stuff between a Windows PC, and a Mac. And here's all these little Easter eggs about the HD DVD encryption key, or [inaudible 00:11:50] who was one of the first YouTubers. Tom Cruise jumping on a couch in Scientology." This was a long time ago. 15 years ago, or something. But we put this video on Dig, and Reddit. And our beta waiting list went from 5,000 to 85,000 people overnight.

**Drew Houston** (00:12:09):
So we got this initial seed audience by chasing that early adopter set. And then, we also figured out these viral motions around our referral program, and shared folders. And so Dropbox started expanding virally for the first several years. And then a lot of the engineering that we applied to the product of Dropbox can't have a bad day when it comes to your wedding photos, and tax returns, and all the things people put in Dropbox. So we took a while to get the beta right before we felt comfortable opening up to the broader public.

**Drew Houston** (00:12:40):
But then, we applied that same engineering mentality to these viral loops. And this was all the era when social media was exploding, and Facebook, and Facebook platform. All these startups were setting new land speed records for fastest to 1,000,000 users, or 10,000,000 users. Things like Zynga. And all of this was built on this emerging playbook of virality, which in turn came from epidemiology. The study of the spread of viruses turned out to be a good parallel for the consumer internet. Draw your own conclusions.

**Drew Houston** (00:13:13):
And then we were like, "Oh, there's no reason this wouldn't work for Dropbox." And some of our early investors, Hadi and Ali Partovi, talked to us about how Facebook thought about growth. And then there were a lot of great people in the early days who had really fine-tuned, and mastered a lot of that. We tried many things. Conventional, and unconventional. But the things around virality, and the referral program really worked.

**Drew Houston** (00:13:39):
And so for the first several years, it was just doubling, 10-xing every year. Taping user counts that we printed out to the wall, and then running out of space on the wall. Having to put a 100,000, 200,000, 500,000, 1,000,000, 10,000,000 on the ceiling. So it was wild. And it was super fun, but it was also super stressful. But going from maybe $6,000,000 valuation in 2007, to $27,000,000 in 2008, to then $4,000,000,000 valuation in 2011, being on magazine covers. Just that whole experience was wild.

**Lenny Rachitsky** (00:14:19):
The visual of the user numbers extending onto the ceiling is such a good one. Of just how quickly things grew. Okay. So let's transition to the second era. So things have been going great, keeps growing. Obviously, challenges along the way, but it feels like it's just grow, grow, grow. It feels like the year 2015 was a pivotal year where things started to shift. Does that sound right?

**Drew Houston** (00:14:43):
Yeah.

**Lenny Rachitsky** (00:14:44):
Okay. Yeah. Let's talk about that.

**Drew Houston** (00:14:46):
Yeah. Maybe the end of the first era, the start of our teenage years, would've been around 2013, or 2014. And maybe before that. 2011, 2012, we started getting all the incumbents, or all the big platform companies. Apple, Microsoft, Google. All of them launched competing products in one form, or another. But weirdly, it was like you see the videos where there's the mushroom cloud in the distance. You see it. But you don't hear, or notice it. So it mostly just seemed like nothing happened.

**Drew Houston** (00:15:27):
When Steve Jobs was on stage in 2011 announcing iCloud, calling out Dropbox by name as something that will be viewed as archaic. And similarly, we always felt like we were in the shadow of the hammer of Google launching Google Drive, which had been rumored long before we even started the company. For the first several years, we were just sort of quivering, waiting for the shoe to drop. And the product's launched, but you would never be able to look at our numbers and see when that happened. And the press often writes about competition like, "Oh, it's a shotgun blast."

**Drew Houston** (00:16:06):
When, later, I would learn it's more of a boa constrictor. But the end of that first chapter really culminated in us recognizing that we experienced a lot of benefits of being this kind of product for everyone. And in fact, it would be kind of hard for me to describe in the early days who Dropbox is for, or what it does. It was similar to what's a phone for, or what's a computer for? And in the beginning, that was a blessing. It meant that those viral loops would really work. Pretty much everybody you would hit would have a need for something like Dropbox. Either a personal need, or a work need, or both.

**Drew Houston** (00:16:43):
But we recognized even back then before competition that, "Hey, people are using us primarily for things like backup, backing up their devices, or storage. People are using us for photo sharing. People are using us as a collaborative space at work, and often, in huge companies." And then, we felt that there was a lot of tension between these use cases. So for example, the ideal file server replacement for an IT admin is going to look quite different from the ideal consumer photo sharing app. And we also recognize things like, "Well, we're going to be competing in many cases with the device, or the operating system."

**Drew Houston** (00:17:22):
When you look at something like iCloud, it's like, "Yeah. When you turn on your iPhone for the first time, it's probably not going to give you an ad for Dropbox." So we're like, "Hey, we need to address some of these issues," and expanded to some new areas to diversify. The first two things we did were, after getting enough complaints from IT administrators asking us what the hell these photo sharing features were for, we pulled all the photo sharing functionality out into a separate and new app we called Carousel. Where the basic value prop was phones, to that point, were limited in their storage by the amount of physical storage on the device.

We're like, "This is silly. You should be able to have your whole life in your pocket." So be able to store everything in the cloud, but have the experience be as if everything were locally there through a lot of caching, and sleight of hand, and thumbnails, and things like that. So there's much engineering there. And then, secondly, we saw people were using us at work, and we're like, "Well, there's a lot of different adjacent workflows." And there was a startup called Mailbox that built the first great mobile email client. Had a lot of funny parallels. They had a [inaudible 00:18:26]-

**Lenny Rachitsky** (00:18:26):
Oh, that's right. They're famous for that waitlist.

**Drew Houston** (00:18:28):
Yeah.

**Lenny Rachitsky** (00:18:28):
That's a really good point.

**Drew Houston** (00:18:29):
Which was crazy.

**Lenny Rachitsky** (00:18:30):
Yeah.

**Drew Houston** (00:18:30):
And so we're like, "I don't know. Maybe this is going to be our Instagram," and so we bought those guys. And then 2014, I'm on stage painting this picture of Dropbox's future. I'm like, "We're going to help be the way that you remember your life. We're going to be your productivity. The new productivity suite on your phone," and all these things. But then it was this dissonance where there were so many things that were going right, and certainly the numbers, user numbers, revenue numbers. We were sort of accidentally cash flow positive maybe a year after launching.

**Drew Houston** (00:19:06):
It was also clear that winter was coming, or that things weren't exactly as they seemed. Then the start of the second chapter, 2015, I'd start to hear a louder set of critics inside, and outside the company. And I'd been thinking for a long time like, "All right, man, we're really fighting wars on all these very disparate fronts." We're with storage. We're competing with the device to back up the device with photo sharing, or competing with Facebook, Snap, Instagram, Google, Apple.

**Drew Houston** (00:19:47):
On productivity, we're competing with Microsoft, and Google. And then, there's a whole new cohort of companies like Slack. And then, there's this experience of one day I am standing on stage talking about how Carousel, and Mailbox, and everything are the future of the company. Less than a year later, Google Photos launches. And not only does it provide a lot of the same value, and in many ways very inspired by what we had done, but they also gave you free unlimited storage for life. Not just photos, but video.

**Drew Houston** (00:20:24):
And so they just totally nuked our business model in ways that were bad enough in terms of just their obvious impact, but even worse because it was so easily anticipated. So this became a very public, and personal embarrassment for me. How could we not have predicted that, or been out in front of that? And then, that started this period. I'd say that was the beginning of chapter two where it was we went from the company that could do no wrong to the company that could do no right, which was a big flip. It was probably early summer, maybe late spring, when Google Photos launched.

**Drew Houston** (00:21:10):
At first, I'm just thinking like, "Okay. This was a big miss on my part. How do we get out of this? We need to tackle some of these competitive issues much more ahead on." And then, what's the problem? Well, the problem is that every incumbent is going to copy your product. They're going to bundle it with our platforms, and then they're going to kill the economics. And that was clear what was going to happen with Google Photos. It was very similar product experience bundled with Android, bundled with all of Google's different touch points, and then free.

**Drew Houston** (00:21:51):
So that was problematic enough for Carousel, but I'm like, "Wait, this is going to happen with everything that we're doing. Same thing with Mailbox." I had even pitch the founders to join Dropbox by saying, "Look, you're going to wake up tomorrow, and Gmail, and Apple Mail, and everything is just going to have these swipes and snoozes. The UI, it's not a durable source of advantage. We'll buy that problem from you."

**Drew Houston** (00:22:19):
And that's exactly what happened. So I'm like, "All right. Even in theory, how do we deal with this?" And I'd gotten the criticism over the years that Dropbox is a commodity, and investors in the early days would be like, "Hey, this is kind a graveyard of a space, and DOA." And I thought, "Well, how do other companies deal with this kind of competition, and commodities?" So there's a great book by AG Lafley and Roger Martin called Playing to Win.

**Lenny Rachitsky** (00:22:53):
Roger's been on the podcast.

**Drew Houston** (00:22:54):
Yeah. I'm a huge fan of Roger's. And I was like, "All right. Well, if we think we're selling a commodity, try literally selling paper towels." Which is what AG, and Procter & Gamble, and a lot of CPG companies do. And AG was the CEO of Procter & Gamble at the time. And basically, he and Roger did this download of how they think about competition, and markets, and advantage. And a lot of it's thinking really critically about where do you play, and how do you win?

**Drew Houston** (00:23:28):
So being very selective about the markets you're in, and only being in markets where you can have a leadership position. And then, being really crisp about what is your leadership position? And so that was a very timely thing to read in terms of, "All right. Dropbox looks like one product, but it's really participating in many different markets. And our big risk might be that we're the number two best thing in each of those markets, which would be a bad situation."

**Drew Houston** (00:23:55):
Another really influential book was Only the Paranoid Survive by Andy Grove, and I'd been a big fan of Andy's. High Output Management is another great book of his, which was my introduction to management. At least, his theory of management. So just loved that book. And then Only the Paranoid Survive talks about Intel's experience where they actually had something like this happen that I wasn't aware of, and we all know Intel as the microprocessor company, Intel Inside. At least in the 80s, 90s.

**Drew Houston** (00:24:27):
But before, they were in microprocessors. They sold memory, RAM. In the 70s, they were running into a situation where they were really high growth, successful business selling memory. But then, they had these Japanese competitors that were just building memory faster, better, cheaper. Just on every dimension. And potentially, also, things like anti-competitive things where the government might be subsidizing these manufacturers. This feeling not only they might just be better at the game, but there's also not a level playing field.

**Drew Houston** (00:25:05):
But meanwhile, you wouldn't see it in their numbers. They're still growing. And just that there were these weird dissonant experiences where salespeople just suddenly have problems selling into accounts that used to be a slam dunk, or things that used to work just stop working. And this happens. Blackberry's, and Nokia's best sales years happened in the years after launching the iPhone. So these things don't actually change immediately. And so Andy labeled these moments strategic inflection points for a company, and I'm like, "Yeah. Then, we're definitely in a strategic inflection point."

**Drew Houston** (00:25:40):
But they were dealing with, "Oka., our whole business is memory. How do we deal with this competition?" And there's this little vignette where he and Gordon Moore, one of the other co-founders of Intel, said, "Hey, let's pretend we're consultants to ourselves. What would we do?" And what they immediately decided was, "Oh, well, we clearly get out of the memory business, and put all of our chips on this sketchy little microprocessor thing that was much smaller. But very high growth, and potentially big market." And then they're like, "Well, why don't we do that?"

**Drew Houston** (00:26:10):
Only problem with that is it's like Google saying, "Yeah. Let's get out of search, and go all in on Gmail, or something, or YouTube." So it just seemed insane. Andy cautions in the book that, "Look, most of the time CEOs want options. They want to hedge their bets, but what you really want to do in these strategic inflection points is go all in one thing." Or as Mark Twain put it, "Put all your eggs in one basket, and watch that basket." And I was like, "All right. This makes sense to me, but man, this is going to be painful." So we go away for 4th of July. I'm with my family in New Hampshire. I reread the book, I come back. All right. I'm like, "Well, we really got to do it." Then, I just killed Carousel, killed Mailbox. Went all in on productivity...

**Drew Houston** (00:27:00):
Killed mailbox, went all in on productivity. To some extent, that was a relatively easy decision because most of our subscribers, 80% of people paying for Dropbox were using it at work. But that meant foreclosing on photo sharing and consumer and storage and all these things that Dropbox had become synonymous with and to this day are some of the things that we're most recognized for. And I wish I could say then everything got better. It was the opposite, actually, the narrative completely flipped on the company. The press started, we killed these new products, and then internally and externally, the narrative became super negative. Articles would come out every week or two, like, oh, Dropbox could be the first dead deck of corn. And then sometimes we've all seen this in tech companies get in this washing machine of self-perpetuating negative press. And if it goes deep enough, I mean, Uber had this for a while later, Meta's had this over their history, lots of companies where you just can't get the monkey off your back.

**Drew Houston** (00:28:06):
And then because what the reporters end up doing is they basically park their metaphorical van behind your office. They interview all the people that you just fired and then print everything that they say anonymously as if it were facts. And there's a lot of truth to what the press were saying, so I can't really blame them or saying they were being unfair, but it immediately put recruiting into this deep freeze. You're just in the situation. You've started this company, it's been super successful, and then suddenly your employees don't want to wear your T-shirt anymore. And frankly, you don't even want to wear your T-shirt anymore. It's just your pride in your own company takes a big hit.

**Drew Houston** (00:28:50):
And meanwhile, I've talked mostly about the external market competitive forces, but inside the company was a mess too. The business's revenue had scaled so much faster than our ability to hire and build the right infrastructure and operations internally. And so everybody's just panicking and being like, "All right, well good, Drew, we're not doing Carousel, we're not doing Mailbox. What are we doing?" And the truth was like, if I knew the answer to that, we would be doing it, but it's just this is going to take some time to figure out. So it was pretty tough when everybody's looking to you as the founder and CEO looking for quick fixes and answers, and also just wondering, how the hell did you get us in this situation?

**Lenny Rachitsky** (00:29:36):
Wow. So we're going to talk about how you're actually turning things around and the work you're doing now, learning from this experience. But I have a lot of questions about this part of the journey.

**Drew Houston** (00:29:45):
Yeah, yeah.

**Lenny Rachitsky** (00:29:46):
Just to almost summarize the journey, it's like launch, killing it, viral, find all these opportunities, find all these big markets. You start three different product lines. There's the enterprise use case, there's the consumer file storage use case, there's photo sharing, there's productivity, and then basically every big incumbent's like, okay, great, we're going to come eat your lunch. And you end up fighting wars on three or four fronts against the big kahunas that have infinite cash and can do whatever they want. It's interesting the way you described the Apple launch of just like it was this mushroom cloud you didn't quite see for a while, but the Google Photos launch was like, okay, that was more clear of like, okay, this is bad news. I know it's always easy in hindsight of like, oh, I should have done something different when Apple launched this thing. But is there anything maybe you could have done, or was it even possible to have adjusted course at that point? Or was it like, okay, this is tough and we need time to figure it out?

**Drew Houston** (00:30:45):
Well, we panicked about the Apple thing too, because that also, that created a much smaller version of the tempest in a teapot that the Google Photos did. It was just much less public. Or it was just less obvious that we would be so out position. And actually we were more surprised to the upside when it's Google Drive launches, iCloud launches, OneDrive launches. And even as they build version two that we just didn't really notice that much of a problem. And I also studied over the years what happened in Netscape and companies like [inaudible 00:31:27] MySpace or these other cautionary tales. And interestingly, Internet Explorer was a thing that eventually undermined Netscape enough to permanently throw it into a negative trajectory. But Internet Explorer 1.0, 2.0, even maybe 3.0 just weren't that good and didn't really do anything but Internet Explorer 4.0, 5.0 and all the bundling, that really made a big difference, or big negative difference.

**Drew Houston** (00:31:55):
And then another thing that was interesting was, so there's a big time lag where between when these products launch and when they actually have an impact, because often it's not so much the existence of the product or it's availability, it's the constant bundling or the constant iteration, or it's the boa constrictor, in any given second, it's not much tighter, but over the following day, you're in a bad place.

**Drew Houston** (00:32:24):
But another thing that was interesting was one cool thing that we got to do and moving to California is you get to actually meet people who had been at these companies. And one of those people was a guy named Bill Campbell who was at Netscape during that whole period, and I had asked him, I'm like, "Man, that's really unfair. That sucks what happened with Microsoft and bundling with Windows and all these things and the antitrust and this and that." And he's looked at me, he laughed at and snorted and he's like, "Microsoft did not kill us. We killed ourselves." And so that was the other half of what was the problem within Dropbox was a lot of our wounds were self-inflicted in that we were struggling to keep scaling and launching all these products. And the Dropbox products had stagnated.

**Drew Houston** (00:33:25):
And I started to learn the hard way that, okay, yeah, I now understand what Bill is talking about, because we were hiring all these smart people, but the things we would do as a company would seem really stupid. And then some of these things were me personally, like yeah, I knew about from reading all these different things, but for whatever reason, I guess we just kept going anyhow. We were just too confident in our ability to respond, or we got lulled into this complacency, we were like, " Oh, well." Because it hasn't been a problem, it won't be a problem in the future.

**Drew Houston** (00:34:03):
But I'd say there's something about the Google Photos launch that just really set things off kilter. And then also for me personally, I think, I'd go from feeling good but stressed out all the time to mostly feeling bad all the time. And I just remember leaving, really trying to get away from the office for a little bit during this period, picked my teeth up off the ground. And the good news, I bought a place in Hawaii, so I was doing that in Hawaii, but didn't feel very good in any way.

**Drew Houston** (00:34:42):
And then I learned. I had to really figure out a lot of different things for myself to be the leader that could get us through this. Because mostly what I felt was like, man, I felt like we'd been doing such a good job, but now, man, if I really screw this up, or maybe I don't know what I'm doing. And then I had to really reset on a number of fronts. One is I just remember thinking my 18-year-old self would be like, what the hell are you complaining about? You did it. This is so great, and yet I felt so bad about things. And so I'm like, yeah, both of those things are true. How do I navigate that? And I think one part that was really helpful is getting a sense of equanimity basically, and doing a lot around mindfulness and meditation to distance... One thing that happens when you're a founder and your company succeeds is your identity is fused with the company. And so it's easy to get into a situation where you only feel good if the company's... Or how you feel is how the company is doing, and you need to separate that a little bit. So what that looked like for me is recognizing, yeah, there's really not an easy button that keeps things up and to the right forever. And most of the entrepreneurs that are my heroes had various periods of wandering in the desert. Those things instead of just being problems were probably the crucible that forged the people that they became. So the presence of badness is not necessarily you are bad. And it's like, yeah, now you're just getting your stripes as an entrepreneur.

**Drew Houston** (00:36:41):
And people were really helpful during that period. So I mentioned Bill Campbell. He was nice enough that we just stayed in touch and he would take me out to dinner every now and then. And I would be freaking out, but I was always surprised he'd never seemed to be freaking out. And mostly he'd just be saying like, all right, he'd dust you off and smack you and say, get your ass back out there. And he was literally a coach of the Columbia football team in addition to being a great technology executive and advisor to many of the great founders that I looked up to. He was really helpful because he helped me believe in myself even when I wouldn't believe in myself. And then people like Andy Grove who had written all these books, I'm like, well, at least I'm having this problem. But yeah, then that raised a question for me. I'm like, okay, I've been through this chapter one where everything's up and to the right, but I ran out of merit badges or Xbox achievements to collect. I was like, we did it. We got the product to this many users, we got it to where we got this valuation, or let me back up. So my whole life, or it's common, I think for a lot of founders, especially if you get good grades and stuff, you follow pretty linear path from childhood, from first get the right test scores, then get into the right college, get into the right college, take the right classes, get the right internship, get the right job. And early startup life is like that a lot. It's like, okay, have an idea. Find a co-founder, get into Y Combinator, get funding from Sequoia.

**Drew Houston** (00:38:22):
Watch an MVP just in 10 million, a hundred million, billion valuation. And I felt like we'd cleared all that. And so actually one of the other problems was, I don't even know what Dropbox needs to be and I don't even know what I need that I want to be in the world. And so part of this was also reconnecting to some kind of sense of purpose, because if I was really honest with myself, I was like, well, I'm just checking boxes and advancing and leveling up like a video game or something, but now I'm here. I'm like, all right, is it just big numbers bigger? It's also not that fun to just launch something that you put your soul into and then just have it get crushed with these competitors. It's even worse when you're bringing all these hundred other people who worked on that thing through that meat grinder too.

**Drew Houston** (00:39:10):
So are we just inventing things 10 minutes before Google and Apple do? If not, what are we doing? And then if we're really, really honest with ourselves, I think it's hard to argue that no one would've invented cloud photo gallery that wasn't constrained to your local storage. So I had to really rethink what does the company need to do. That actually took a longer time to answer than I wanted, but then also just what do I want to do? And I was like, well, you know what? I'm going to do this.

**Drew Houston** (00:39:45):
What I ended up settling on was like, look, coming back to my 18-year-old self, I was like, all right, stop complaining. This is a pretty good situation. You wanted something where the learning curve would be steep. Here it is. You get to do this really meaningful work to build things for millions of people, work with awesome teammates. There's always good things about it, and you've chosen this life, so stop complaining about the burdens because it's really dangerous if you start to resent the job or feel like a victim because that's just corrosive to everything else that you can get in this spin cycle.

**Lenny Rachitsky** (00:40:20):
What helped you come to that place? Because that sounds like a very hard place to get to and a very important place. Was that Bill Campbell advice? How long did it take you to go from this is not going well and life sucks to, okay, let's look at it this way?

**Drew Houston** (00:40:35):
I think there were a couple unlocks that were pretty fast, but I think again, this was not like, oh, then I felt better and we lived happily ever after at all. There were some moments of recognition where I was able to, I always try to get away from the mayhem or the whirlwind and recenter myself and do think weeks and things like that. And so those were really helpful because I felt when I diagnosed, well, when I'm on this treadmill and just firefighting or how did I miss this important thing? It's like, well, basically the root causes of being on the treadmills, I was too busy firing to aim, and then I thought I would do my aiming on vacations and things like that. But no, I need to get off the treadmill every now and then and make the space to really address some of these bigger questions.

**Drew Houston** (00:41:31):
Because it's not like I didn't forget about them. They just were always below the fold on my to-do list to deal with. And then until a few years go by and you're like, wait. Yeah, we really were just strategically out of position. The other way I came to some of these conclusions is a combination of things. So I think picking up a meditation or mindfulness practice, having therapists personally, having coaches, having friends and mentors, other founder friends. And so I wouldn't say it was any one thing, but I think you have to build this whole ecosystem of support around yourself and recognize no one's going to do that for you. And it's super important.

**Drew Houston** (00:42:12):
And then to Bill's other point, he was like, yeah, Microsoft wasn't killing us, we were killing ourselves, I think having to ask some hard questions to myself of, yeah, some of these wounds were self-inflicted. As much as there might be, I could look at this and say, oh, Microsoft is mean, or Google was mean, or Apple was mean, but I'm like, I drove the ship towards these rocks. So that's one downside of being founder, CEO. You can't blame the last guy. I think one of the big questions every founder has to figure out is not just, how do I have equanimity? What's my purpose? But then also, what am I doing that is destroying the company?

**Drew Houston** (00:42:57):
Because everybody's got strengths and weaknesses, but as CEO, both your strengths and your weaknesses are massively amplified and a lot of blind spots in your personality can become huge cultural dysfunctions in the company. And so step one to dealing with that is building awareness. And one thing that floated around the company for a while was the Enneagram, which on the surface is a personality typing thing like Myers-Briggs. I often describe it as it's like Myers-Briggs, but actually useful. The way it works is it's a personality typing system. You get a number from one through nine that's your dominant... Or there's nine types that are numbered. It's similar to Myers-Briggs in that Myers-Briggs has 16 types of letters like ENFJ or different things. But I find that Myers-Briggs is more descriptive. So it tells you, okay, you're an introvert, not an extrovert or vice versa. You're judging instead of perceiving. But I could never really work out what that meant. And so it's descriptive versus predictive or causal.

**Drew Houston** (00:44:06):
Enneagram I find is the theory that is really about your fundamental motivations. What are you running towards and what are you running away from? And the theory goes through some combination of your wiring from your genetics and your early childhood. Your autopilot is fixed after your childhood in that you just have instinctive responses to things. Now it's autopilot. It doesn't mean it's your destiny. You could override it, but it's super important to become aware of what are you running from and towards. And the Enneagram gives you a really good map to do that.

And so I'm reading this book and it starts by you type yourself, and I'm rolling my eyes. I'm like, okay, I'm a seven. What's a seven? And then you read, each type has its own chapter and you read the chapter and I was like, oh my God, this completely describes who I am down to my core. And I was almost looking over my shoulder. Who the hell did this? I thought I was a special snowflake, but apparently not. Okay, I'm listening. And [inaudible 00:45:16] with a lot of my early coaching experiences, which are, you do a 360, your coach sits you down, they're like, here are your strengths. In my case, it would be things like, oh, you're really creative and you like new ideas. That's okay. And second, you really love people. You build great relationships with people. I'm like, okay. Uh-huh. Third, you're really comfortable in chaos and resilient. I was like, this coaching thing is great. See, yeah, that was my first coach. And then you flip to your development areas because no one calls them weaknesses.

**Drew Houston** (00:45:49):
It's like, all right, one is you're bored by routine and really undisciplined. Second is you're really conflict avoidant. And third is you're intuitive, but you're also pretty chaotic and don't create enough structure for people so they're mostly confused. And when you pair the two together, you're like, oh, wait. Yeah, I'm creative and like new ideas, but I'm bored by routine and I'm a bit of a space cadet. All right, you love good relationships with people, but you don't want to make them unhappy so you don't tell them the truth basically about things that are difficult to hear. And it's going to be a similar map of things depending on your personality type or circumstances. But the Enneagram was that map for me to be like, okay, I've got these things that are genuinely great, but I need to address these downsides of the company's conflict avoidant, so we're not telling the truth and then making a bunch of predictable mistakes, or I'm creating this really chaotic environment where people don't know what they're supposed to be doing.

**Drew Houston** (00:46:57):
That's a problem. My personality is quite directly, I'm going to torpedo the company unless I do something about it. The same is true for any founder. And I'm a type seven, it's the enthusiast. It's like, again, central casting is like a lot of new things and creative and love just interesting things. But then a space cadet, FOMO, some of the shadow side of it, and every type has a thing like that. So really understanding, okay then, I was also just frustrated in the day-to-day of my work, I'm like, yeah, I'm not really getting to do these creative things where I'm too busy firing to aim. I'm missing these bigger picture things that are really taking the ship way too close to the rocks. Okay. Then I need to then understand what those things are and either work on myself or hire the right people who don't have those issues or one way or another, just make it so the company is not so exposed to my personal dysfunction.

**Drew Houston** (00:47:56):
And so yeah, it was a whole constellation of things like that where it's like, all right, we have to get... I mean, after this whole tailspin, we got to work. I'm like, all right, first business issues. The most urgent issues were just blowing all this cash. We need to get the P&L to look a lot better and get out of this land grab mode or fighting with Google on giving more free storage away or just being in these fundamentally unprofitable things. And so we did that. We cut the unprofitable parts of the business. We turned cash flow positive in 2016, maybe several months after this reckoning, and that set the tracks for getting into a billion run rate in 2017, going public in 2018.

**Drew Houston** (00:48:40):
We had to come up with a new vision and mission for the company, which I'll maybe save for chapter three. And then I was also just trying to do things that helped sustain me through the difficult things. I'm like, well, I am an engineer. As a little kid, I was coding since childhood. I knew I wanted be a founder. I didn't know I wanted be a CEO, so I backed into being a CEO. And then, so I'm like, man, there's a lot of tedious things about being an executive. So I was like, I'm going to find ways to automate tedious parts of my job and learn machine learning in 2016 and '17, and that ended up being really important later on. But then also it's just coming to terms with my job and what I wanted to do from a personal perspective, which included things like, oh, I'm lucky in that being a CEO is, I always wanted a steep learning curve.

**Drew Houston** (00:49:33):
In a lot of professions like in sports or math or chess or certain academic fields, you peak in your twenties or thirties. CEO's not like that. You can go your whole life and still not be a master. So I'm like, for better or worse, it's pretty cool that I can do this kind of work. And then I don't really get that much out of just chasing bigger numbers. The thing that is my favorite thing about Dropbox is looking over someone's shoulder in Starbucks and seeing if the little Dropbox icon is there. Building something that becomes a verb, taking some of the pain out of technology, things like that. So really reorienting. Yeah, it's not about just the external scoreboard just much. It's more about the craft of being a great CEO and building things I'm really proud of or really making a difference. And so those kinds of things, none of them all coalesce. It's like this lens that slowly and stubbornly comes into focus, but making the time to have some of that reflection was super important.

**Lenny Rachitsky** (00:50:36):
So much of what you're describing makes me think about founder mode in a lot of different ways. One is your point about how you didn't have a lot of time to think about where things might be going and how the markets are shifting, and that has to happen, basically, no one else is going to do that if it's not you. Also, just this idea of how much of the solution was you understanding yourself better, reflecting on where you have strengths and weaknesses. Also, I'll mention, so before I had Brian Chesky on the podcast, I asked you, "What should I ask Brian?" And you asked me to ask him basically about founder mode, which was the first time I think you talked about it, and now it's like this whole thing. So you've been in the middle of all that for a long time. I guess just any thoughts on the importance of that and how you think about that as a founder?

**Drew Houston** (00:51:19):
So founder mode means a lot of things to a lot of people. So it's a bit of a Rorschach test, but I think parts of it that really resonate with me are, there's this evolution you go on as a founder where you don't know what you're doing and everything is new and unfamiliar. And then you're also very involved in all the details because there's no one else to be involved in the details. It's not like someone else is coding for you if you're just in your room at home. But then over time, you have to, there's this, Ben Horowitz calls it the Product-CEO Paradox, where the first way that companies die is from founders not letting go. And so you need to learn to scale yourself and hand off your responsibilities and operate at a higher level of abstraction. But then the second way that companies kill themselves is the founders get too far away.

**Drew Houston** (00:52:23):
And I felt like I had done that. That's a big part of the problem that led to a lot of this chaos at the end of chapter two, where I was like, oh, man, I'm on this treadmill. I'm doing stuff, but I'm clearly not setting the right direction. Or people are confused or it's not working, and I was too distant from the product. And then there's all kinds of issues you can get into as you hire an executive team and get that to function well. And so I think there's a debate to what extent founder mode is a mindset or a destination. To me, I think the destination part is pretty important. I think it's like after that learning journey, suddenly you have this conviction from actually knowing, from having lived experience and navigating a lot of these things where suddenly it's a lot clearer.

**Drew Houston** (00:53:10):
You don't have the same kind of confusion or learning curve problem. And then if you start out being too far leaned in, then you lean out too far. And I think a lot of that founder mode flip is when you're like, hey, this is not the company I want to be running. I need to be more involved. I need to stop making excuses. And basically, I also don't want to apologize or negotiate all the time for the kind of company that I want to be part of and run. So I had that too over time, maybe more towards a chapter three, some of those elements of... And I don't know if there's a way to get to this destination without some level of pain. Certainly the Elons and Steve Jobs of the world had big spans of wandering in the desert where they were not as cool as they are-

**Drew Houston** (00:54:00):
... spans of wandering in the desert where they were not as cool as they are today. The things they were doing weren't working, and they managed things very differently in that sort of post wandering phase. Suddenly had this level of conviction, and intuition that they might not have had before, or they sanded down some of their more counterproductive instincts. Again, there's a lot of surface area.

**Drew Houston** (00:54:27):
I don't know if anybody has a common definition of founder mode, but certainly that evolution, the way I think about is it is very difficult. One of the hardest things when you're running a company is that you're hiring all these execs who the only thing you know is they know a lot more about the subject matter than you do. What does it even mean to bring someone in, bring an exec who had managed a double-digit billion dollar P&L at Google, and what am I supposed to tell him how sales works or marketing? It's easy to be too leaned out there where you're not really setting a clear direction. Each founder is going to have their own things that are easy, things that are harder. Step one is being aware of them, and then being intentional about, how do I offset those things? Then you eventually get to a point where the learning curve flattens out. You actually do have some experience, and you can have a lot more conviction about what the company should do, and where you want it to go.

**Lenny Rachitsky** (00:55:25):
Awesome. The way you described it to me, which has really stuck with me, is as a founder, as you said, you're in there leaning in, doing, on top of everything, micromanaging. Then you lead out, and hire execs that you think are smarter than you, and delegate more. Then it's like, "Oh, things aren't going great," and come right back into it.

**Drew Houston** (00:55:40):
Yeah. Specifically, it's often well-intended or it's often even not conscious, but as you hire these execs, and then you also build up an HR function, you get coaches, and things like that, and then the things are not going well, one thing that can happen is you start collecting all this feedback. People are like, "Wait, if I can give Drew some negative feedback, then maybe I can displace some of the ..." It's sort of easy to get into a situation where you're sort of being my to-do list of problems, just weaknesses, and flaws I needed to work on was super long, way longer than anyone else's. On the one hand I'm like, "Yes, I am wholly accountable, and responsible for everything in the company as its CEO." That's not wrong. But you can end up just carrying too much of the water, so a lot of it ended up having to be this pushing back a bit. Be like, "All right, yes, there are things I need to work on, but that can't be an excuse for my execs to somehow evade accountability."

**Drew Houston** (00:56:52):
There's these sorts of systemic things that can creep in for sure. Then you get this feedback, and you're like, "Now, I'm doing all this behavior," or, "Now, I'm behaving in ways that I don't think are really right." Or I'm like, "Everything is just negotiated compromise with everyone on the roadmap or on how we behave or strategy." At some point you end up blowing that up as a founder, and that was certainly my experience.

**Lenny Rachitsky** (00:57:19):
**Drew Houston** (00:58:42):
My playbook as an entrepreneur is to get really frustrated by something, and then try to solve that personally. I'll cover the team, and what direction we ended up choosing together. My first frustration was forgetting a thumb drive, and then wanting to never have that problem again.

**Drew Houston** (00:59:03):
But my second big frustration was like, "Man, this scaling this company was way harder than it should have been." Thank God I had people like Bill, Andy, friends, mentors to guide me along. A lot of people don't have access to that. But I feel like I could have written myself postcards that would've made this a little bit easier. Or, there are just some things that don't make sense about how you scale a company. There seem to be things that are really broken about the way we work in general.

**Drew Houston** (00:59:40):
As I was piecing through a lot of the challenges, I'm like, "All right, what should Dropbox do?" We should probably not do things that would've been invented by everybody else 10 minutes later. Where are the problems that are not solving themselves? One problem I felt like was not solving itself was this problem of being just on this treadmill where I'm like, "I'm working really hard. I'm in meetings all day, emails all night. Is this it?"

**Drew Houston** (01:00:08):
First of all problem, no complaints. But I'm like, "This is so weird. Yeah, I'm making huge strategic mistakes, because I'm too busy firing the aim, and I'm upset at myself about a lot of these things." But then I look left, look right, I'm like, "Wait, everybody is too busy firing the aim." Everybody is on the same treadmill.

**Drew Houston** (01:00:28):
I'm like, "This is bizarre. Who's winning here?" It's like if I was working hard, and the company is benefiting, then no problem. But I'm very busy, but I'm not really productive. I'm not putting in a lot of creative input. Then the company's not getting a lot of creative output. It just seems like lose, lose, lose all around.

**Drew Houston** (01:00:50):
I'm like, "Why are we on this treadmill?" We know from brain science that people are most happy, most productive, fulfilled, engaged, kind of everything you want, where we'd rattle off the same list, like when people are focused, when there's some kind of flow state, when they've had enough sleep, when they have a sense of purpose. We know all that. Yet we go to work, and it's like this cage fight of who's busiest, who's gotten the least sleep, who's most inbox zero. Then you look at the environment on our screens, now, especially after COVID, we live with that.

**Drew Houston** (01:01:33):
That's where we work. We don't work in an office, no matter what flavor of hybrid you are, you're working on a screen. COVID relocated us from offices to screens. You look what's on that screen, it's like, "Yeah, if you wanted to design a working environment that made it impossible to ever focus, ever get into a flow state, bombard you with just constant interruptions, distractions, and busy work, then yeah, you can squint and see me even before COVID, but definitely after.

**Drew Houston** (01:01:59):
But that people are like, "Yeah, I'm really not feeling great." I'm not feeling super engaged because half of my work is bullshit, the other half I'm just distracted to death. I'm like, "Why is this this? All right. Maybe Dropbox doesn't have to exist. Maybe we did the thing. Maybe I should go." Maybe I'm ready for the tech bro ascendancy. Maybe I should be flying cars, space, cancer, climate ...

**Lenny Rachitsky** (01:02:26):
Ayahuasca.

**Drew Houston** (01:02:28):
Yeah, right. But then I would talk to people who were actually working on those things, and they were on the same treadmill. I interviewed this director of engineering at SpaceX. I was like, "Oh my God, you're actually going to Mars. This is so cool. How do you guys work together? How are you going to get to Mars?" He's like, "I don't really understand the question."

**Drew Houston** (01:02:50):
I'm like, "What tools do you use? How do you work together? How are we going to get to Mars?" The answer was basically, "We're going to get to Mars through a lot of emails, and a lot of files." I was like, "Oh, my God." Most of the time, we correctly think about technology as just like force multiplier, like the amplifier of our abilities. But from another perspective, we're only as good as organizations as our tools, or tools can become the limiting factor. It's really a problem when the mechanism through which we were getting work done has gone from being a force multiplier, and amplifier to a suppressor. We are the frogs in the boiling water where this kind happen one day at a time.

**Drew Houston** (01:03:36):
All the people working on those things, and there are many others, they're all dealing with the same issues of, "I can't focus my tools. I'm fighting with my tools." Somehow the tools went from helping us get the work done to becoming the work, as we layer. It went from 5 of them, to 10 of them, to 500 of them.

**Drew Houston** (01:03:51):
Yeah. I'm like, "All right. If Einstein were alive today, what would his day be like?" He would wake up. He'd have to delete a bunch of LinkedIn notifications. He'd get down to work, start writing equations, and then someone would Slack him, and interrupt him. Then he'd get back to work. Would we still understand relativity if Einstein were living in that kind of working environment? Probably not.

**Drew Houston** (01:04:17):
Even just visiting my dad at work when I was a kid, a lot of things were the same. He had a phone, a PC and an office, but he could turn his phone off. He got 5 emails a day, not 500. He would literally come home from work with a briefcase, put it down, and stop thinking about work. Then when he would be in the office, he could close the door, and actually get stuff done.

**Drew Houston** (01:04:40):
I'm not saying we should go back to the early 90's, but I'm saying it wasn't always like this. The whole frontier of productivity at the time felt like it was just taking us in the wrong direction. Things like Slack, which are wonderful tools, also chop up your day into little fragments, and make you cognitively diabetic. There's a lot of empty carbs in our collaboration, and our collaboration tools.

**Drew Houston** (01:05:05):
I'm like, "Okay, here is a problem that's not solving itself. Actually, none of our competitors even are framing the problem correctly. What we should be recognizing is that our ultimate non-renewable resource is our time, our attention, mostly our brain power, our creative energy. I think as a civilization, we're going to, hopefully in 5 or 10 years be like, "Man, we took this crazy detour where we basically hooked up that brain power to this thing that was just burning off half of it as friction with our tools." That was dumb. Yet no company was even talking about this, let alone fixing it.

**Drew Houston** (01:05:49):
Then we came up with a new Dropbox mission. It became Dropbox is designing a more enlightened way of working, because the way we were working is unenlightened, unexamined. Meanwhile, we were patient zero as the company. No one was doing a lot of thinking. Everybody's kind of busy. Everybody's bumping into each other. There were a lot of cultural problems. Like that we were faster. Your company grows, and rises, the easier it is for new people in the company to think they hit a home run instead of starting on third base. A lot of complacency, entitlement, and preoccupation with things other than our customers or other than our products set in.

**Drew Houston** (01:06:37):
It wasn't all bad. I think there were a lot of good things we had to pay attention to. We were scaling up the business financially. We were getting ready to go public. We had to mature a lot of things, but we lost sight of the place that money is coming from is from our customers, and the thing they really want is a great experience.

**Drew Houston** (01:06:58):
We've been tackling this in a lot of different ways. First is just things like after COVID hit, suddenly the world is working in a completely different way. We saw this as a big opportunity, not just a crisis. Or once you got past the crisis part of COVID, there was a big opportunity. COVID kind of wrecked everything, but we didn't have to put the floorboards back down in the same places. I'm a big fan of people like Peter Drucker, and lots of people who think about the nature of work, and this idea of we're finally decoupling work from our physical location. It's a big deal. It's probably one of the biggest changes to knowledge work in our lifetimes. Or even since Drucker coined that term knowledge work in 1959.

**Drew Houston** (01:07:44):
It's like, "Cool. For the first time, we can actually design how we work in ways that our parents couldn't, or the way that most of us just receive how we work from our parents." But like, "Hey, we can actually do things completely differently." We came up with this whole virtual first model where we're 90% remote, but then most importantly, we're trying to rebuild the new product stack for distributed work. I'll talk a little bit about where the new things we're doing, and then I'll come back to how we're rebooting some of the core.

**Drew Houston** (01:08:16):
But initially, after COVID, we studied a lot of how other remote first companies had worked. We synthesized with a lot of primary research of other companies like GitLab, Automattic, and others who had been doing this for a long time. We collated all that, synthesized a common playbook, open sourced it. If anybody's curious about this, you could find Dropbox. We've open sourced our virtual first toolkit. Then since COVID, designing a working model very in line with our new mission with a lot of benefits. It really works. The employee retention, satisfaction, engagement or offer separates all the things we care about have been significantly up since COVID. There's many factors to that, but there's a lot that's working about the model. But really, we're approaching it from a product standpoint.

**Drew Houston** (01:09:07):
In deciding to turn Dropbox into this lab for distributed work, we're like, "Okay, when you open up your laptop in 2025, 2030, what do you see?" Hopefully, there's some new stuff or some things that are different on there. What are those things, and how do we bring them back?

**Drew Houston** (01:09:20):
One of the first things you lose when you go distributed is context. We get a lot. You get a lot for free from osmosis. Just, "Oh, hey, how are you doing? How's this?" Or, "Oh, do you have that thing?" But then when you're remote, then a lot of that gets replaced with endless video meetings, and lots of Slack messages, which is super inefficient.

**Drew Houston** (01:09:41):
Also, one thing that was clear across the board from all the successful remote first companies was you have to be more documented, but then you have a lot more documents. This started turning up very obvious problems, which is like, "Wait, why do I have 1 search box at home, and 10 search boxes at work? Why is it easier to search all of human knowledge than my company's knowledge? Why is this problem getting worse not better every year?" If you think about it, we've got 10 search boxes that you search 10% of our stuff. I'm like, "We should just fix this."

**Drew Houston** (01:10:14):
Again, even just as a frustrated user, maybe 2018, 2019, I'm having this problem myself. I'm like, "It's really hard to find stuff." Plus there's always cool vector search stuff, and deep learning stuff happening. Even for ChatGPT, I'm going to make a personal search engine that takes us, so I shouldn't have to, right? If I search for strategy, and the title of the document is plan, it should get it right, and the matching should be fuzzy. Then I built something like this. I'm like, "Oh my God, it completely works."

**Lenny Rachitsky** (01:10:42):
You built this yourself, you're saying?

**Drew Houston** (01:10:45):
Yeah, it went nowhere for a couple of years, but then after COVID, and after we're dealing with all these information problems, I'm like, "Yeah, we should be the company that helps you get the right information at the right time to the right person." I built that little hacked up search engine thing that I made, might be an ingredient to this.

**Drew Houston** (01:11:08):
We bought a little company called Command E that was doing universal search, and we created this new product called Dropbox Dash. Basically, Dash connects to all your different apps. It gives you universal search. Then obviously after ChatGPT, not only can you do conventional search, but you can ask questions in natural language, and answer a lot of the questions that ChatGPT can't because it's not connected to your stuff. If you ask ChatGPT like, "When does my lease expire? Where's that slide from last year's product launch," can't tell you because it's not personalized. Because it's not connected to your information. We did all of that for this whole connector platform where we index the known universe of SaaS apps, have this kind of intelligence engine that brings all of it into a common representation, and lets you do conventional search, natural language search.

**Drew Houston** (01:11:58):
Then more than that, we want to organize your working life for you. Stepping back even further, often it happens that a company likes to starts solving a problem, and that problem is never permanently solved. In a lot of ways, Dash is the biggest, best embodiment of solving the same kinds of problems I started Dropbox to begin with. Because on the one hand, I was talking about thumb drive, but I wasn't, really. If you think about the higher level job to be done, I was like, "No, the real issue is that it's super hard to find my stuff, organize my stuff, share my stuff, and secure my stuff."

**Drew Houston** (01:12:36):
In the beginning that was like my stuff was my files, and the scattering was across devices. Now, the stuff is tabs in my browser, and cloud tools. But it's a lot of the same problems. We've talked a lot about search, but then organize is still a big problem. There's no desktop folder or desktop as in the physical realm, when you're in your browser. It's just this ocean of content that washes by you. Then you get tired. You nuke the whole environment. But there's nothing to come back to. There's no collection concept. Files have folders, songs have playlists, links have ... If you're getting ready for board meeting or remodeling a house, there's no common container. If you have a Google doc, and a 10 gig, 4K video, and an air table, this is a new problem.

**Drew Houston** (01:13:25):
Dash also has this thing called Stacks, and has basically smart collections that helps make sharing a lot easier. But again, a lot of this was an intersection of like, "Yeah, how has work changed since COVID? What crazy little science projects have I done? Then how should one find, organize, and share their information in the cloud era? If I were to build Dropbox today, what would it look like?" Dash is first part of the answer. We're super excited. But we just launched it two months ago, and signed up our first customers. Then I can come back to just rebooting the core business too. I think one of the big challenges or lessons from this whole period is that one, is as part of the problem we had with competition, and this chapter two was external, but a lot of it was internal, where we couldn't organize ourselves effectively. The business has scaled so much faster than our ability to manage it. First is to recognize this is not an unusual problem. Many companies, not all, but certainly many, and actually many in SaaS productivity have this sophomore slump where you have the super successful first product, but then it's challenging to build the next platinum album. Dropbox has fallen into that category. You could say Zoom, Slack, others.

**Drew Houston** (01:14:58):
First, it's easy to feel bad about it yourself, but no, you're not. This is actually a very common problem. Then there's a lot of solutions. People devise lots of solutions to this. Part of it's structural. Your company has to go from a pretty fun, and stable state of a functional organization where you have engineering in product design, there's one product, and there's one customer. Everybody's working, we are all in the same direction for the same team to then when you have multiple products, then you have all these conflicts. How much should we invest in the core business versus the new thing? Who should have authority in the company? You try to build multiple products in a functional org, but then you lose accountability.

**Drew Houston** (01:15:50):
People are like, "I'm spending a lot. I'm trying to make Dropbox work. I'm trying to make mailbox work. I'm trying to make paperwork, and all these things." Then you respond by breaking up into groups or business units. We have a product business unit, product GM structure. Then you have to refactor things to have a common technical platform, and all the GNA functions. There's this metamorphosis, and maturation you need to go through that is mega obvious to experienced people. That you're just totally blindsided by this if you're solving it by trial and error. There's a great book, Zone to Win, that talks about some of this, what needs to happen internally by Geoffrey Moore.

**Drew Houston** (01:16:35):
Then there's a lot of cultural problems that set in. We just touched on some of them at a high level. But whether you're in a company or an empire, a civilization, what gets you to the top turns out to be pretty similar things. You have this outsider, challenger mentality where you have to eat what you kill, and the odds are against you. Through hard work, learning, and grinding, you start moving up the curve faster. But then once you're successful, then there's a temptation to take the foot off the gas or enjoy the finer things in life or just focus on other things, be it other than what got you there. I think what we experienced, many other companies experienced, Bill Campbell experienced in Netscape, the success plants, the seeds of failure in terms of complacency, entitlements, or taking your eye off of what got you to be successful in the first place, yeah, I went from earning a lot of the early stage startup founder merit badges to earning the stagnation, and irrelevance merit badge. Now, the turnaround merit badge. Part of that is the turnaround playbook.

**Drew Houston** (01:17:47):
Okay, first, it's just getting out of this delusional state where you think you're great, just splashing cold water on everyone, and be like, "Look, we haven't shipped any. Last three years, the only thing our customers have really meaningfully seen from our product is a price increase. What the hell are we doing?"

**Drew Houston** (01:18:08):
A lot of it is getting back to like, "Hey, we have to focus on craft. We have to embrace a growth mindset and learning. We have to stop blaming external factors or displacing blame. We have to be a high agency culture." There's a cultural transformation that you have to personally embody, and be clear about.

**Drew Houston** (01:18:29):
Then in our case, we had to reboot the whole exec team. We've had all the time getting the right kind of leadership team in place for the end of chapter two and three. Part of what happens in conjunction with a lot of the cultural stuff I just talked about, is also you can have this as the talent flywheels flying forward, and flies in reverse. You have a new set of issues where you get all these amazing people in the company, but they want to work for Facebook, not MySpace. You start having retention problems when the narrative goes negative. We start bleeding talents, and then it's also really hard to hire people for the same reasons. Especially, as I think about where the big lessons from chapter two that we've addressed in chapter three, the seniority gap is something that can accumulate because much of the talent can fly to the next shiny thing. What you do in response typically is, you promote a lot of people internally, which is good for them. It seems good for them in the short term, because they're just getting a lot more authority, and responsibility quickly. People like that on day one. But it's a problem because suddenly people are solving problems through trial and error. Things may be new to them that are not new to the industry, and suddenly no knowns to the industry are unknown and knowns to those leaders. Then usually you can get around that by hiring a layer of experienced execs or having enough of them in the company. But in the talent wars of the late teens, we were in the situation as were many companies where, any exec you talked to with any experience had five offers from fan companies for three times the comp, and a third of the workload, or 10 offers from pre-IPO startups, and C-level roles. These are offers in hand. The only thing that really worked to even just keep the lights on was basically giving people double promotions so that they could justify someone who's a director at company X, making him a VP at Dropbox. Then you had to do that because they literally had offers in hand of many millions of dollars a year. Otherwise, you'd just hire no one. That cooled off a lot after the correction in SaaS, after we got more of our vision straight, and certainly after we've been working on things like Dash.

**Drew Houston** (01:20:59):
But that seniority gap is really rough. Where ...

**Drew Houston** (01:21:00):
Seniority gap is really rough where you need to have enough experienced people in the company who can then train your high potential people. And again, that sounds obvious, but it's very difficult to do that in an environment where the talent flywheel is going in your direction and then starts flying away from you really quickly. And then through the battlefield promotions and these double promotions on hiring, you can end up a situation where there's a huge voltage drop in terms of people's matching between what your company needs and what they know how to do. And that also doesn't mean hire only experienced people either because then you're splicing in a lot of outside DNA into your organism. But what you want is to keep it roughly in balance. I mean 60-40, 40-60, 50-50, whatever. But your high potential people also lose out if they don't have experienced people learning from.

**Drew Houston** (01:22:00):
So that's a little bit, I think that it was a big part of the stagnation. Again, it's not because the people, but this was something that I was doing to my team. So it's not them or certainly on my watch, but I think it's a dynamic you have to be careful of like do we have the right balance of people who are super high potential and talented, but doing their thing for the first time? Are they paired with people who have been there, done that, this is not the biggest job of their life, and that they can help those people up the learning curve faster such that the aggregate learning curve or the aggregate learning rate of your company is what it needs to be.

**Lenny Rachitsky** (01:22:41):
Man, I feel like there's a book here of lessons for founders, things that you never hear about until-

**Drew Houston** (01:22:51):
Mind the seniority gap.

**Lenny Rachitsky** (01:22:55):
That's one of so many. People think about becoming founders, they start companies, "Oh, I'm going to start a company, there we go." And then, they think about that first part of the journey that you talked about. And most people don't even get that part. And the best case scenario is you experience a ton of growth. And then, the way you described it to me previously is you hit the final bosses where showing all the success and they're like, "Oh, I see there's a big market here. Let's go after it." And you're battling Apple and Google and Microsoft and Meta, and a lot of people don't think about that and how that's the future if you do well.

**Drew Houston** (01:23:28):
Yeah, every time you move up a league, your reward is a stronger and better opponent and potentially more unlevel playing field. And that's just the way it is. You can't control that. All you can control is how you respond. And so, you want your company's response to embrace that challenge and use it as a mechanism to get stronger. Easy to say, hard to do.

**Lenny Rachitsky** (01:23:51):
Yeah. So maybe just to start reflecting on some of the lessons from this journey, because there's a lot here. One question people may be wondering is, should I even be a founder? Should I start a company? A lot of people, there's a lot of pain that you've been sharing of the challenge of starting companies. What do you tell people that are asking you like, "Hey, should I start a company? Am I right to be a founder?" Is there any advice there that you could share?

**Drew Houston** (01:24:18):
I don't know what made me want to be a founder. I just showed up that way. And so, there's probably some chemical imbalance of some form that causes people to do it, but I had a lot of insecurity about should I be a CEO? Because coming from the technical side, because I was like, I can do the engineering and technology, but all I know about being a CEO is I do not know how to do all of those things that CEOs do, and I do not look like that, I do not talk like that, it's intimidating.

**Drew Houston** (01:24:46):
So I think first is it doesn't have to be an all or nothing decision. I mean, obviously it's all encompassing decision to become, to start a company, but you'll have many points along the way where you can keep calibrating. And I think the biggest thing is, or advice I got from one of the founders of a company where I worked was I was like, oh, should I be CEO? Should I not be CEO? Should I be CTO? He's like, "Look, just try it, be CEO, see how it goes."

**Drew Houston** (01:25:17):
Initially, I thought I was like, all right, well I'm going to get the company to 100 million valuation and the trend will just be going faster, faster and faster, faster. And instead of just getting thrown off of it, I'll just hit stop and retire. And I thought that sounded like a great plan. But then, moving to the Bay Area, you suddenly meet, you throw a rock and hit 10 people who've done that. And what do you learn? This was just shocking to me because I talked to one of the advisors I mentioned earlier. He's like, "Yeah, the day I sold my company was the saddest day of my life." I'm like, "What?" As a 24-year-old, and he's like, "Yeah. Well, I just felt like the company had drifted away from what it was supposed to be and I just didn't like it." I think they're certain in the founder mode trough where you're like, wait, things are screwed up and it's on my watch, but I'm not really sure what happened and what do I do about it? So I think burnout is the biggest thing that will kill you. And so, I think that's why these coping methods and getting your own head is important, but if you do that, then it's also this amazing experience. And then often so many founders go back and start another company and another company. And you don't have to, but there's a reason that people do that because there's just a lot of rewarding aspects of it, building great things with great people that impact huge populations.

**Drew Houston** (01:26:42):
What I found is that founders keep doing this because they love it in some way, or at least have a love-hate relationship with it or can't imagine something else that they would otherwise be doing and I'm probably the same way. But I think to get through that, you have to, Ben Horowitz also said the hardest thing for a CEO is to manage your own psychology. And we've talked a lot about, yeah, first you have to be aware of what your psychology is, and then how do you make sure you're not resenting or hating your company or yourself and view...

**Drew Houston** (01:27:18):
I think the hardest thing for a founder, challenge is not optional. You're going to be challenged, but the suffering is optional. You don't actually have to suffer. I mean, look, there's crunch periods or times where it's taxing, but it doesn't have to be this experience of suffering all the way or burning out or being really angry and sad all the time. Although certainly, in my experience that I did get in periods where I was just sad and angry all the time. So you need to figure out how do you work your way out of that.

**Lenny Rachitsky** (01:27:47):
So along those lines, I'm curious, so you've shared a few things that have helped you level up as a leader and push through that. You mentioned meditation, Bill Campbell, a coach. You mentioned a few books. What has helped you continue to level up and stay ahead of where the company needs you? Sometimes you fell behind, sometimes you're ahead. What are some tips for founders that are trying to stay ahead of where things are going?

**Drew Houston** (01:28:12):
Well, I think first thing is I think a theme that's carried throughout a lot of these different chapters is you have to figure out what game you're playing and what the rules are when the game keeps changing. So I like in addition to the Enneagram or coding, I also like a long time video game player and there's a great game called, everybody's familiar with StarCraft. It's an awesome game. I think it has many lessons and riddles for an aspiring entrepreneur, many parallels with running a company. I'll save that for another one.

**Drew Houston** (01:28:54):
But this idea, Mindshare is a big resource. It's not actually how much money or economy or land or whatever or-

**Lenny Rachitsky** (01:29:03):
Actions?

**Drew Houston** (01:29:03):
Military. Yeah, it's like you can only do so much or think about so much, your attention, but that's actually not the big lesson. In StarCraft and other games, there's this concept of there's a micro and macro, and I would add to that, there's also the meta game. Okay, so what's that? Micro in these games, something like StarCraft is like, all right, can you click really fast and move your people and build things really quickly? And it's like the mechanics. Every second, how many hundreds of things are you doing per minute or how many things are you doing per second? And that's like a conditioning and practice and one kind of thing.

**Drew Houston** (01:29:43):
Then, there's the macro game. Well, could come back to the micro, in a product context or in a startup context, that might be things like, all right, how do I, often the stuff around product design and technology and distribution. So early founders especially are going to be totally fixated on here's how I make this great design and here's how the engineering works and here's how we get users and here's how our sales motion works, or here's how our viral loop works or things like that. And it's very in the details.

**Drew Houston** (01:30:12):
And then, there's the macro game. In StarCraft, that might be things like, well, clicking a lot is important, but overall you're really managing your economy. So do you have more expansions and resources than the other player? Are you building up your military? Are you getting your balance of the investment in economy and military? Are you scouting? Do you know what the other person's doing? It's more strategic and conceptual. So things in the startup realm that would feel more like that are more like, all right, not just the mechanics of how does this feature work, but more like, all right, what's the business model? What's the market? Who are my competitors? How do I differentiate myself? How's this all going to evolve over time as the category goes from super high growth to more mature, things like that? And you respond, okay, we got to go from one product to multiple products, so then we have to reorganize the company that way and we blah, blah, blah.

**Drew Houston** (01:31:11):
And then I'd say there's also this, the meta game, which in gaming is a pretty specific thing. It's like some combination of the game itself gets updated so that as the creators, they make this unit stronger and this area unit is stronger, and this ground unit weaker and try to keep this big system in balance. And then, also as this community of players figures out new strategies and this big repeated game of rock paper, scissors, people figure out, okay, this one strategy is now categorically better. And it's always this adversarial thing that's always shifting. And so, playing StarCraft in 2020 is pretty different from playing StarCraft in 2015 or 2010 because everything, both the players are shifting and the game itself is shifting. And the game, it's usually not shifting a lot. So it's really more of an ecological effect.

**Drew Houston** (01:32:02):
So I think it's super important. And then, in a startup context, what's the meta game? Well, I'd say some of it is just business cycles. So I think you'll have, as you see, there's this boom in the late teens in SaaS and in tech. There's a bit of an AI boom happening right now. It's not that different from what, or parts of it are not that different from what happened with the dotcom era in '99, 2000 or the web 2.0 era in 2007 or others. And then, you have these bust times like 2008 and 2000, 2001, et cetera. So there's a market cycle thing that often creates similar dynamics where it's like five years ago every tech company was hiring the maximal number of people that it could. And then, that created one talent dynamic. Now, it's a lot of companies are keeping high comp flat or negative, so that's just a very different thing. But even those are cyclical.

**Drew Houston** (01:33:03):
Then, you want to identify how is this game of business actually fundamentally permanently changing? So for example, we launched Dash, we usually do a press thing, conventional press, tech press, we got almost no coverage for it. And yet all the stuff we did on social or going direct was way, way, way more impactful. And it's not just Dropbox or in tech, but it's also presidential election. All the candidates are on podcasts. Yes, they're on CNN too, but not really. So that's just a new thing. How do you manage the brand of your company or launch products, things like that. The nature of marketing is changing. And that changed when, in 2007 too, instead of hiring PR firms and buying AdWords and stuff, we were creating these viral videos and we were using math from epidemiology to create these viral loops. In a lot of ways, Dropbox took that, transplanted that consumer internet playbook into business software.

**Drew Houston** (01:34:10):
But the thing is you need to understand what game you're playing and you need to get good at the micro. It's not that the meta is more important than the micro, it's like you need to do all three at the same time. And so, that's really the hard part. And then, when things are shifting. And the way you know what game you're playing is to be systematic about training yourself and probably the single most useful, or at least important to me piece of advice I could give is you have to figure out how to keep your personal growth curve ahead of the company's growth curve. The single most important impactful thing for me has been reading because talk about this micro, macro, meta games or abstractions, this is the kind of stuff you can learn from history.

**Drew Houston** (01:34:55):
Learning what happened in Netscape ended up being pretty important to what happened at Dropbox in 2014. Learning what happened to Procter & Gamble and selling makeup ended up actually being pretty relevant. So I think having a broad information diet is really important. And I've collated a list of books that were really impactful to me. You can find a way to share that. All right, I'd be happy to share that.

**Lenny Rachitsky** (01:35:18):
Absolutely.

**Drew Houston** (01:35:20):
But in a lot of ways, as a CEO, you have to be right about a lot of things, especially and including things you haven't done before. And I think from that perspective at least, I'd say there's things that are challenging for your head and things that are challenging for your heart. The head challenge is one that's more like, all right, how do I cultivate wisdom really quickly? And so, in addition to a lot of technical books about here's how you do marketing or product or management or things like that. So it could be around higher level, almost philosophical when it's things like Buffett or Munger or Bezos would be more on that end of the spectrum. But it's not just reading. I think having a community of people that you can learn from is really important. And I've also found it's interesting, you learn having stable of people that are at the same stage as you, couple people, a couple stages ahead, 2 years ahead, 5 years ahead, 20 years ahead.

**Lenny Rachitsky** (01:36:23):
And this is founders, other CEOs?

**Drew Houston** (01:36:25):
Other founders, founders and CEOs, exactly, because you'll learn different things from them. So first of all, when you're in the early innings, you'll actually get more useful advice often from your peers because they're going through the same thing. So it's like, how do I raise a seed round? You're going to get a lot more out of someone who did that a year ago or is doing that now. Then if I asked the doc, "How do you raise a seed round?" He'd be like, "I don't remember. Just build a good company."

**Drew Houston** (01:36:55):
And you'll see that as you talk to people at different phases. I think again, early stage it's more about the micro, about the product and distribution, then there's this continuum. So then it's about your business model and monetization and the financials, then it's about defensibility and maintaining that advantage. And as you talk to, as I talked to my peers would be more about the product. That was the scaffold I would see over founders.

**Drew Houston** (01:37:26):
And then, people that were further out, actually they still knew all the micro stuff and had feelings about it, but they're almost philosophical in nature like philosophers and just very broad intellectually and drawing lots of distinctions from lots of different things and knew a lot about a lot. And I think most of the tenured founders that I'm, most of the founder CEOs who've been at it for 10, 20 years plus, I'm at 17, almost all of them read voraciously. So I think combination of reading and community are the most important things.

**Drew Houston** (01:38:02):
But then, the last thing is being systematic about it. So in line with what does it mean to keep your personal growth curve ahead of the company's growth curve? I think one way to do that exercise is think about is always be working back from, in one year from now, what will I wish I had been learning today? Two years, five years? And often, those to-do lists are pretty different.

**Drew Houston** (01:38:23):
So in 2008, I would've been focused on just getting users for Dropbox, but then looking ahead to a year later, I would've been building the first business functions in the company, and then five years later, it would be thinking about how do we fight with Google and Microsoft and all these other things. And then, for my own skills, it would be like 2008 where like how do I raise a scale to venture round and what do all these terms mean and things like that. But then, how do I be a great leader, great manager, comfortable speaking publicly, things like that. And the things that are further out are often the most intimidating, but you also have the most time to learn and you often can psych yourself out in terms of like, oh, this is just new and uncomfortable. And I did say this is one of the hard challenges. You're always going to have that feeling of discomfort. And instinctively, what you're going to want to do is run away from it because you're human. It's uncomfortable to feel like to confront these things that you're not good at or might be embarrassing or threatening. And so, I think an important part of being a founder is learning to run towards that feeling not away from it. That's a big part of your learning rate is the extent to which you're pushing yourself beyond discomfort.

**Drew Houston** (01:39:43):
And so having that list of a year from now, two years from now, five years from now, what can I start learning today? And then, recognizing these things, they're all trainable. So in five weeks you're not going to be a great guitar player or surgeon or manager, leader, but in five years you can put a pretty big dent in those problems. And people who enter college and leave college often have a lot more relevant knowledge after those four years. So just having that growth mindset.

**Drew Houston** (01:40:16):
And I'd say the last piece of it that is also tricky in addition to going from a heart or mindset perspective going towards discomfort and in addition to a lot of the other stuff we talked about being like how do you have a sense of equanimity, you're not burning out and finding a sustainable pace. I think the other thing is often smart people have more trouble learning than otherwise. And there's this great article that I, it's probably my most handed out articles probably from the '70s or '80s or something called Teaching Smart People How to Learn. Because what ends up happening is the more book smart you are or the more you were identified as gifted or intelligent as a kid, the more not knowing something or being wrong, it's not just not knowing something or being wrong, it's like an assault on your identity.

**Drew Houston** (01:41:11):
And so, I found that one of the best predictors of an exact continuity of this scale, it's not just what they know or what they can do, but the extent to which they can actually be aware of their failures and not blame or dismiss things because smart people have a really fast rationalization hamster. They can convince themselves that, well, here's how they were technically right even though clearly they were wrong or clearly the thing didn't work out and they let themselves off the hook. And all of this happens unconsciously as a protective mechanism.

**Drew Houston** (01:41:47):
So I think finding ways to take responsibility or always have a mindset of what if I were a hundred percent responsible for this? What if it was no one else's fault? What if it's entirely in my control? And those things are never true, but that's always... Or with perfect hindsight, what would I have done differently? And just owning things ends up being more painful in the short run, but then some painful hours can save painful years. And there's a great book on that mindset stuff, the 15 Principles of Conscious Leadership, Diana Chapman and some co-authors, and she's been a coach and friend of mine. She's amazing and it has really helped me on that front. But really, how do you train your head and train your heart? There's different things you do for each, but all in service of keeping your personal growth curve ahead of the company's growth curve.

**Lenny Rachitsky** (01:42:39):
I love how tactically your advice is. There's this piece of advice of think what do I need to know a year from now, two years from now, five years from now, this reading list, this idea of having staggered staged founders in a community to help you along this journey, this article about how smart people learn and how to get past this block. So I love just how concrete a lot of this stuff is. So then, just zooming out and maybe final reflections on this journey, as you described this, it's basically the epitome of the hero's journey that you've been on here. Things are good, trouble appears, you enter this other land of everyone coming at you trying to squash you in every direction, battling them, and then emerging now into this new land with all the things you've learned, coming back to the original idea of what Dropbox could have been. So I think that's partly why it's so interesting is it's the epitome of-

**Drew Houston** (01:43:29):
But we still got to make it fully out of the, we're on our way up, we're clawing our way back up.

**Lenny Rachitsky** (01:43:37):
The dragon's still out there. The incumbents still exist.

**Drew Houston** (01:43:41):
And by the way, it will always be like that. You're never done with that. It's not like you reach the A-top or you don't have to if you don't want it.

**Lenny Rachitsky** (01:43:53):
And it makes me think about, as you're talking, people think about product market fit and something, a lesson from this journey is just like product market fit is not a binary. You will have it now. You've got it and you have it and will last forever. It's constantly being broken by other people or if there's something you've discovered that is a big market. And maybe final reflection on the story you've shared things that might be helpful to people, just the journey you've been on that might be helpful to founders who are going through this right now or will probably go through this.

**Drew Houston** (01:44:20):
I think it's important to remember we're all really lucky to be able to do this kind of work. And the things that I got out of... When I was 18, the things I thought I would get out of starting company would've been the things you would think. It'd be like, oh, I'll be really well-known or really rich or build these awesome things. I think there's benefits to that, but what I've grown to appreciate a lot more is over time is this thing or the experience and the journey of starting a company can really be of forge for a better character or there's transfer. A lot of what you learn here is transferable in many different domains.

**Drew Houston** (01:45:11):
I just had a little kid, Charlie, he's one-year-old, so I know we're both in new dad mode, but I'm like, it's made me a better husband, it'll make me a better father, made me a better person. And then, I just say, listen, there's no easy button where things are just up to the right and so don't look for one and don't feel bad if things are difficult. But ultimately, one of the biggest surprises about being able to spend time with all these interesting people and founder CEOs who've created these iconic companies is at the end of the day, they do it because they love it.

**Drew Houston** (01:45:56):
And so, I think learning to decouple... That doesn't mean they're always having a good time at all. Most of them use metaphors like Jensen or Eli, like chewing broken glass, staring into the abyss, and yet for a lot of these people, there's at the end of the day, nothing they'd rather do. So I think finding the gun in that.

**Lenny Rachitsky** (01:46:19):
I'm very excited to follow the next chapter of your journey. I am really thankful that you shared these stories. I think this is going to be helpful to a lot of people. I think this is a story that people study for a long time because it's common and not talked about much. Drew, thank you so much for being here. For folks that want to maybe check out what you're building, anything you want to share, where should they go?

**Drew Houston** (01:46:40):
I'm just Drew Houston on all the socials and then we got our new product, Dropbox Dash, dropbox.com/dash. Right now, it's for companies. We'll have a downloadable version. You could just get on your phone or your computer soon. But Lenny, this is awesome, super fun, really big fan of what you do.

**Lenny Rachitsky** (01:46:59):
Awesome. Same. Drew, thank you so much for being here.

**Drew Houston** (01:47:03):
Thanks.

**Lenny Rachitsky** (01:47:04):
Bye everyone. Thank you so much for listening. If you found this valuable, you can subscribe to the show on Apple Podcasts, Spotify, or your favorite podcast app. Also, please consider giving us a rating or leaving a review as that really helps other listeners find the podcast. You can find all past episodes or learn more about the show at lennyspodcast.com. See you in the next episode.

---

## [3/15] How to break out of autopilot and create the life you want | Graham Weaver (Stanford GSB professor, founder of Alpine Investors)

**Lenny Rachitsky** (00:00:00):
You are ostensibly a professor at Stanford Graduate School of Business, and you shared that when people come ask you for advice, the most common question you get is, "What should I do with my life?"

**Graham Weaver** (00:00:10):
Imagine that you're walking home from work, you see this bright, shiny object, and you realize it's a magic lamp. And you rub the lamp and this genie comes out and the genie says, "Hey, I can give you one wish. Whatever you throw yourself into with your whole life and your career, it's going to turn out great." If that were true and you had that genie, what would you wish for? At some point in this one life we get, you want to get yourself on that path of that journey.

**Lenny Rachitsky** (00:00:35):
This whole exercise connects to something that you're a big advocate of, this idea of getting out of autopilot mode in your life.

**Graham Weaver** (00:00:40):
You're unconscious, and you may not even realize why you're doing what you're doing or even realize what you're doing. So for example, I get up, work out, drive into work, fight traffic, commute, maybe I return some emails, fight traffic on the way home, rush through dinner, go to bed. It's not a day that is intentional. It's not a day where I've said, "Where do I want to be going with my life? What's important to me in this world?"

**Lenny Rachitsky** (00:01:02):
You another quote, which is, "Everything that you want is on the other side of worse first."

**Graham Weaver** (00:01:07):
Pick anything. You want a better body? Okay, you're going to need to go to the gym. When you go to the gym the first few times, it's going to not be that fun. The first move is negative. If I'm optimizing for tomorrow and I just want to have a great day tomorrow, I'm going to stay exactly where I am. So many people I see have this happen, where they hit a plateau and they never move past it, because they're not willing to have that hard day, month, week, year.

**Lenny Rachitsky** (00:01:35):
When should you quit something, because some things are just not worth it.

**Graham Weaver** (00:01:37):
I think the time to quit is when you can no longer ...

**Lenny Rachitsky** (00:01:44):
Today, my guest is Graham Weaver. Graham teaches a top-rated course at Stanford's Graduate School of Business, a course which is technically called Managing Growing Enterprises. But as you'll hear in our conversation, he ends up mostly helping students figure out what to do with their lives and how to get out of the autopilot mode that most people are in. He recently won Stanford Graduate Business School's 2024 MBA Distinguished Teaching Award. And teaching is actually his side gig. His full-time job is founder and CEO of Alpine Investors, a private equity firm, which based on my research, is one of if not the top-performing private equity fund in the world. So the advice you're going to hear today is coming from someone who is actually doing the thing, not just teaching the thing.

**Lenny Rachitsky** (00:02:28):
In our conversation, we cover practical exercises that can help you figure out what you should do with your life, including something he calls the genie framework and the nine lives exercise. We talk about why life is suffering, and you may as well choose something worth suffering for, also, why most things in life that are worthwhile take more time than you expect, some practical advice for creating accountability in your life to help you achieve your goals, and so much more.

**Graham Weaver** (00:05:38):
Thanks so much for having me, Lenny. I'm super excited to be here.

**Lenny Rachitsky** (00:05:40):
Okay, so you are ostensibly a professor at Stanford Graduate School of Business, of a class called Managing Growing Enterprises. But I was reading an interview with you, and you shared that when people come ask you for advice, they're not asking, "How do I start a company? How do I manage my growing enterprise? How do I make my company grow faster?" Most of the questions, the most common question you get, which is surprising to me, is, "What should I do with my life?" First of all, is that accurate?

**Graham Weaver** (00:06:10):
Yeah, that is accurate. About half of all the times I meet with students, that's the question they ask. It's really funny too, because sometimes they'll come with a PowerPoint presentation and a two-by-two matrix and an expected value and all that. But really they're asking the question of, "Hey, what should I do with my life," which is, by the way, a great question.

**Lenny Rachitsky** (00:06:29):
Why do you think that's the case? That's not what I would imagine someone at an MBA school in a business class asking for advice on.

**Graham Weaver** (00:06:36):
My typical meeting with a student will say, the student will start to tell me about these two or three different career or job alternatives that they have, A, B and C. Let's just use two, A and B. And then they go through and they start talking about A, and they tell me all the pros and the cons of A, and they go through it. And then they go through B, and they start telling me all those things. And I start asking some questions.

**Graham Weaver** (00:06:56):
And after about five or 10 minutes into that, I can tell that their heart and their soul and their energy is really for B. That's really what they want, but they're talking themselves out of B, and they're going to talk themselves into A. So what I try to do is, A, first, I try to let them realize that their real energy is for B, just let them feel that and understand that. And then secondly, I try to figure out, what are the limiting beliefs they have? What are the fears? What are the obstacles? What are the voices in their head? What are all the societal pressures that are keeping them from doing B? And then we try to deconstruct those and get them to go do B. So that's the process.

**Lenny Rachitsky** (00:07:41):
Is there an example of a student going ... just this actual conversation you had with someone?

**Graham Weaver** (00:07:45):
Absolutely. So there's a great example of a student of mine was from Brazil. And he came in and his prior job prior to business school was he worked in consulting. And that was more or less what he wanted to do. And then at some point I asked the students, "Okay, imagine you knew you were going to be successful and you were going to have a dream. And whatever you dreamed was going to come true, what would you dream for?" And he wanted to start a nonprofit in his home country of Brazil to help students have more access to education. That was what came up for him.

**Graham Weaver** (00:08:27):
And over the course of our class, we just chipped away at all the fears and limiting beliefs of why he shouldn't or couldn't do that. And by the end, that's what he did. And so that's a real life example. And there's all kinds of other examples, but that second one of going and starting a charity, it's not on the beaten path. It's not what probably your parents are thinking you should do. There's probably 100 reasons not to do it. You probably don't know exactly how to start, and so it's intimidating. But if you start with this idea, "Hey, five years out, 10 years out, I know I'm going to be successful," and work backwards from that, you're going to come up with a better answer.

**Lenny Rachitsky** (00:09:07):
This is a framework that you call the genie methodology or the genie framework, this question that you just asked that we should spend more time on, which is ... I guess, you tell the story of the genie and how to think about this [inaudible 00:09:20].

**Graham Weaver** (00:09:19):
Well, when I was 13 years old, I used to listen to these motivational tapes, mowing lawns. And I'm pretty sure it was Brian Tracy had this exercise. And I've adapted it, so I'll use my exercise now. So I say basically imagine that you're walking home from work and you see this bright, shiny object. You walk over and you realize it's a magic lamp. And you rub the lamp and this genie comes out and the genie says, "Hey, I haven't been in this bottle for 10,000 years yet, so I'm not fully formed. So I can't give you three wishes. But what I can do is I can give you one wish. And the wish I can give you is whatever you throw yourself into with your whole life and your career, it's going to turn out great. It's going to work out great. It's probably going to take longer than you think. It's going to be harder than you think, but you're going to be really happy you did and it's going to work out beyond your wildest imagination." If that were true and you had that genie blessing you with that wish, what would you wish for?

**Graham Weaver** (00:10:19):
And then the students come up with an answer that's really close to their heart. It's a thing they would do, absent the fear of failure. And then the second part of the exercise is basically that's what you should go do. You should be spending your life in pursuit of your genie goal. Maybe you can't start that tomorrow; you have financial obligations, maybe there's some experience you need. But at some point in this one life we get, you want to get yourself on that path of that journey. And that's the exercise that I go through with the students.

**Lenny Rachitsky** (00:10:56):
It's such a simple idea, that I can totally see how it could be so powerful. And I love the way it's framed as not like it'll guarantee you'll be successful. It's instead, I'll guarantee you'll be happy.

**Graham Weaver** (00:11:09):
You'll be happy that you took the path. And the reason I say that is that usually the genie goal is the not well-trodden path. So you don't even know exactly what the goal is. This charity to start education for underprivileged kids in Brazil, that takes its own form over the course of a decade. And it will almost certainly turn out differently than you think. So it's more that you'll be really happy you went and started that journey, and it will go great. It'll probably go differently and take longer than you think though.

**Lenny Rachitsky** (00:11:46):
What are some other examples of genie goals folks have followed that are kind of non-traditional and they've ended up being happy about it?

**Graham Weaver** (00:11:52):
Mine was buying companies in my dorm room at business school. I have a student who's starting an amusement park in Texas. That's a pretty crazy example. Many students who are leaving their job and doing startups. A lot of students who going in the nonprofit world. So really just lots of examples. It's as variant as the number of students I have. And that's the magic of it, because you have inside of you some really unique dream that you maybe haven't even shared. And the goal is that that's your uniqueness, and that's what you should be bringing into this world.

**Lenny Rachitsky** (00:12:37):
This whole exercise connects to something that you're a big advocate of, this idea of getting out of autopilot mode in your life. And the way I think about it is people ... and I'm going to ask you to describe it, but just it's almost like you're driving and you've never turned off the cruise control in your car, and you don't realize it. Talk about this idea that most of us are in autopilot and why it's so important to realize that and get out of it.

**Graham Weaver** (00:13:00):
You start off and you're unconscious, and you're kind of going through the motions. And you may not even realize why you're doing what you're doing or even realize what you're doing. So for example, a typical person gets up, they have whatever their morning routine ... I'll use myself. I get up out, work out, take a shower, drive into work, fight traffic, commute. I'm late. I get in, I'm late to a meeting, I'm kind of rushed meeting; meeting, meeting. Okay, quick break for lunch. Maybe I return some emails; a few more meetings, a couple of Zoom calls. Fight traffic on the way home, rush through dinner, get back on email. Go to bed. Okay, that's my day. And that's a busy day. I felt like I did a lot. I'm exhausted, but it's unconscious.

**Graham Weaver** (00:13:45):
It's not a day that is intentional. It's not a day where I've said, "Where do I want to be going with my life? What's important to me in this world? What are my values? What 10 years from now will I wish I was starting to embark on?" Adding that level of intentionality and then working backwards from that is really the magic of getting to that 10 years from now and looking back without regret and getting to a later point in your life and feeling like you're doing the thing you're put on the earth to do as opposed to just going through the motions.

**Lenny Rachitsky** (00:14:22):
So the question then here is, how do you get out of that autopilot mode? Because first of all, no one really realizes this is the case. And I'll tell a quick story. We're going to different preschools and daycares for our son. He's a year and a half oldish. And we went to this Montessori school, and the teacher's like, "I'm going to be very clear: what we're doing here is informing your child's subconscious. That's what they're learning here. And that's a huge responsibility. We put a lot of love into that idea, but it's very important you understand that's what we do at the school." And I never thought of it that way.

**Graham Weaver** (00:14:56):
That's amazing that they said that. So you just opened up another part of this, which is depending on what research you read, somewhere between 95% and 98% of our thoughts are subconscious. And those get programmed in somehow, some way. A big part of that, by the way, is media, our friends, our parents, our boss, our coworkers, what we think we're supposed to do, social media: "This is cool, buy this Ferrari," all these different things. And then you're just operating out of that. And so the idea of being intentional is create space, get out of that, get out of the fog of war, make some space. We'll probably get to this in a little bit, but I do it with an executive coach, and really ask deep questions, make space, ask questions, create the intention that you want in each of the areas of your life. And then start having your calendar reflect that intention.

**Lenny Rachitsky** (00:15:58):
And so this genie exercise is one approach, is just ask yourself this question. Can you say it again just for folks so they don't miss it, what's the question you should ask yourself?

**Graham Weaver** (00:16:08):
I mean, the biggest question I think with respect to your career is, within reason, what would you do if you knew you wouldn't fail? That's the biggest question. I'll give you a few more if you want.

**Lenny Rachitsky** (00:16:24):
Yeah, please.

**Graham Weaver** (00:16:26):
Different questions trigger different things with people that could be helpful, trigger in a good way. So here's a few more. If you didn't have to make money, what would you do? And that'll answer what you enjoy. Naval Ravikant has a great question, which is, what's play for you that is work for other people? So for you, Lenny, that might be a podcast. That might be play for you; that's really fun. You're always going to do better at that. You're going to spend more time, you're going to enjoy it more. That's a good one.

**Graham Weaver** (00:17:00):
Another one is, what's the thing you really want to do? But you're just too embarrassed to say it? And my answer to this question was that I wanted to be a motivational speaker like Tony Robbins. And I was super embarrassed to say that, but that actually works into a lot of my life at this point. Another question, who are some people that you admire and want to be more like, and what do they do? Where do you find those people? What are some things you want to learn and how do you want to grow over the next five, 10 years? So I know a lot of your listeners are working in tech. So five, 10 years from now, you're amazing, you're best in the world at X. What's X? And how do you start to work toward becoming great at X? So there's just a few more other questions you could throw in to help you figure out some things you're excited about.

**Lenny Rachitsky** (00:17:55):
You mentioned this idea of limiting beliefs. And I think a lot of people listening to this are probably having these beliefs right now of, "Okay, but I have a family to support. Am I going to go start a charity in Brazil? That's absurd. It sounds easy, but I can't actually change my life this radically." Can you just share something to help people get past that in some way?

**Graham Weaver** (00:18:14):
Sure. The first thing I would say about limiting beliefs is they're the most powerful and the most dangerous when you don't even know what they are. So when they're in the recesses of your subconscious mind, that is 95% of your thoughts, that's when the limiting beliefs are the most dangerous.

**Graham Weaver** (00:18:32):
So a simple example might just be like I have a limiting belief that I'm not funny. That's in my subconscious and I don't even realize I have it. So I avoid things where I'm trying to be funny or tell jokes. I mean, that's a silly example. But let's use your career now and come up with some maybe deeper ones. Let's use that charity example, the Brazilian charity, "I'm going to start a Brazilian charity." What are the limiting beliefs? Well, oh my gosh, there's a million. "I don't know how to start. I don't know how I would pay for myself. I have business school debt. I don't really know what I'm even talking about. I don't even have a plan for it. How would I get funding?" Those are all these things that are flooding your mind.

**Graham Weaver** (00:19:15):
So the first exercise is just write all that down. Just get it down on paper, and then two things will happen. One is when you get it down on paper, it will almost immediately strip that limiting belief of a lot of its power and a lot of its scariness. Because now it's just something like, for example, "How would I fund this?" So the second thing that is that a lot of that scariness becomes just a to-do item. In the recesses of your subconscious, that is a very scary, limiting belief that will actually keep you from doing the thing you love. Once it's on paper, now it's just a to-do item that you can actually deal with with your conscious mind, just like you do anything else. So, "How would I fund this," just becomes a plan, like, "I need to design a plan where I'd get funding for this charity." And then that just is a problem like any other problem. It's not this nebulous, scary fear. It's just literally a to-do item.

**Graham Weaver** (00:20:14):
So the first part of limiting beliefs, write them down, understand what they are, look at them in the cold day of light on paper, and then translate them into things that are just obstacles to be overcome. And by the way, if you're listening to this podcast, you've overcome millions of obstacles in your life, and these are no different once they're down on paper.

**Lenny Rachitsky** (00:20:34):
Are you actually doing these exercises with your students? They're taking a class about growing their enterprise and then it's like, "Okay, let's analyze what you want to do with your life"? Is that how this class goes?

**Graham Weaver** (00:20:42):
That's a really good question. So I'll give you a little bit of background. So I was a case guest at Stanford Business School while I was buying companies in my dorm room. And it was talking about all the things that went wrong. And so it's kind of a really fun case. And I did that for 12 years. And I started to realize that was my most energetic day of the year. And so long story short, I started teaching a class full time. So I did that for four years. And I was teaching the Xs and Os of being a CEO, basically: hiring, firing, having hard conversations, managing a board, fundraising, selling, all the things you would imagine that a young CEO would need to know.

After about four years, I started to realize that was great, except that no one went and did it. So the class is on entrepreneurship, and they learned how to be an entrepreneur. There was only one problem: they didn't actually become entrepreneurs. So then I said, "Well, wait a sec. I have to readjust my class a little bit, and I have to spend some time on the stuff we've just been talking about: finding out what your dream is, your limiting beliefs, starting to map out goals toward your actual entrepreneurship dream or whatever your genie goal is."

**Graham Weaver** (00:21:53):
So the way I say it is the university allows me to teach the class because I teach them entrepreneurial tricks, tactics, tools that will help them become a great CEO. But the real reason I teach is because I'm trying to help people really go find the thing that they're excited about and get into the life path of doing that thing. So I do both, but I do the second one kind of ... maybe that's not the headline of the course, although I think that's probably why people like the course so much.

**Lenny Rachitsky** (00:22:30):
It's like a Trojan horse element.

**Graham Weaver** (00:22:32):
Exactly, yeah. Tony Robbins used to say that people hire him for success, and he has to give them that so that he feels like he earned his money, but what he really delivers them is fulfillment. And it's a little bit like that. People take my class to learn how to be a CEO, but what they really get is hopefully on the path of doing the thing they want to do.

**Lenny Rachitsky** (00:22:54):
A lot of what you've been talking about, it's almost an assumption that you'll be more successful and happier if you follow your energy, your passion, versus the, "Here's how I'll make a bunch of money. Here's how I'll move up the ladder." Can you speak to that?

**Graham Weaver** (00:23:09):
Yeah. Gosh, Lenny, it sounds so cliche when you say it like that, that I almost cringe a little bit. Let me try to give it a little bit different framework where it won't sound as ... because cliches are cliches because they're true. But let me try to give you a little bit different framework to think about it.

**Graham Weaver** (00:23:29):
I think about it as you have sort of your heart or your soul or your internal scorecard, and then you have your head and your mind and the world's external scorecard. And I'll speak from this from experience. So when I graduated from business school, I took the job I was supposed to take. And it was the safe job at the big private equity firm that paid well and looked great on my resume. And I took that job. That was the external scorecard. My head said that. I built an expected value calculation, all that stuff.

**Graham Weaver** (00:24:02):
The problem was it wasn't my internal scorecard. It had nothing to do with what I actually cared about and wanted to do with my life. And so the way that shows up is just this tension, friction, stress, anxiety, burnout, all those things. And you can will that, you can will yourself through for long periods of time. In fact, if you want, you can will yourself through that your whole life.

**Graham Weaver** (00:24:31):
But then once I got into the path of the thing that I was excited about, that's when I really felt my energy change dramatically. And I developed almost a superpower in that thing, because I had more energy. I was willing to work longer, I was willing to do it. I've been running my company now for 23 years. I was willing to do it for a longer period of time. I thought about it in the shower. I thought about it when I went on runs. I talked to people. They wanted to join because I was excited. And it was a whole different level of power and life force that I was able to bring to that.

**Graham Weaver** (00:25:14):
So I'm saying this from experience, that you're going to show up so much differently in the thing that you're excited about, that that alone is going to make your life a lot better. But the great irony is you're going to do so much better in that thing than you are the thing you are, quote, "supposed to do." And I certainly, certainly understand that people have real life constraints on their finances. And I 100% get that. And so a big part of what I do with students is work through those things and say, "Okay, great. Let's talk about this job you're going to take for X years and you're going to pay off your loans or whatever. But during that time, let's get you on the path of the thing you really want to do." So sometime soon in your life, you want to get on that path.

**Lenny Rachitsky** (00:26:02):
There's a bunch of directions I want to go here. One is I'll just share, what you're describing is exactly what happened to me, without knowing this advice. I just started writing things online because the poll was there and people seemed to enjoy it. So I just kept following that path. And the whole time, my wife's like, "You can't make money writing on the internet. That's not a thing. Why would you do this? You have all these skills, that you can make a lot more doing other things." But I just kept doing that, and that's what led me to this life now where I make much, much more than I made as a product manager at Airbnb. And also, it's a lot less stressful. And so, I'm a living example [inaudible 00:26:39].

**Graham Weaver** (00:26:38):
Yeah. And if you go back to what I was saying before, you probably were answering the question, "What would I do if I didn't have to make money?" You just did that because you enjoyed it. What's play for you that's work for other people? What do you do in your free time? You were answering all those questions. And then I think a lot of people just say, "Oh, well, that's just a side hustle or a hobby." It is until it isn't, right? And you're a great example of that.

**Lenny Rachitsky** (00:27:04):
Just to set it some expectations with folks, you tell me, how often does this actually work out for someone where this ended up being the right path for them, that it ended up working well for them? Just because people can hear this and be like, "Yeah, okay, I'm sure it works for some people; probably not for most people." What's the success rate, however you define that, for your students?

**Graham Weaver** (00:27:25):
Okay, so the short answer is I don't know, because I don't have full information about how my students do or what their path would have been had they not done it. So it's a kind of difficult question to answer. But what I would say is that the formula you're solving for is you, excited about something for a decade or more. So what has to happen? You excited about something. We just talked about what that is. The decade or more is going to come true more likely if you're excited about it. But also, you have to go in at the beginning with that mindset and the structural ability to stay at it for a long period of time. So the missing ingredient in most of the people that fail is time.

**Graham Weaver** (00:28:18):
And I'll use myself as an example. I started Alpine. We lost money on our first fund. We started doing well. We got hit by the recession, we started digging out, whatever. But long story short, I was 14 years into running my firm until I could say with confidence we were going to even stay in business, let alone be really successful. And probably 18 years until we were what I would say really successful by external standards. And now 23 years, we've had a great run. But if you take away the time period, then I would've gone down as one of the, quote, "failures," as opposed to, I would say, one of the success stories using this methodology.

**Graham Weaver** (00:29:10):
So time is the variable. And I think the biggest part of that is it's really, believe it or not, not typically the finance or the structural piece. It's the entrepreneur or the individual's willingness to actually stay with it. And then upstream of that, it's their belief about how long it's supposed to take. And I really hate this about the social media and just media in general, where they try to paint this picture that things are going to happen overnight. And we've invested in 600 businesses. Let's say 550 of those are founder-started businesses where we're the first money. I've never, in one of those examples, seen anyone who did something quickly. They've all been very, very long stories to get to that point.

**Lenny Rachitsky** (00:30:12):
This reminds me of a quote I read. I don't know if you wrote this or you shared it, this idea that the life is suffering, so choose something worth suffering for.

**Graham Weaver** (00:30:20):
I wrote that, yeah.

**Lenny Rachitsky** (00:30:22):
You wrote that. Talk about that, because it feels like it's exactly what you're describing. It's going to take a long time to figure something out.

**Graham Weaver** (00:30:28):
Exactly. Yeah, exactly. I'm glad you brought that ... Yeah, I mean, think about it. Again, using myself as an example, the first job I had, I wasn't suffering any less. I was getting on planes, I was working late hours. My time was not my own. If I would've had kids, I'd be missing their little league games and all that stuff. So I'm doing that anyway. I'm just doing it for something I don't care about. And then I start my own company and I was, quote, "suffering," getting on planes, doing all that. It was just something that I cared about. So yeah, the quote, "Life is suffering. So figure out something worth suffering for," you're going to suffer either way. And that's another thing I think people don't realize, is there isn't really a path that is easy that I've ever found.

**Lenny Rachitsky** (00:31:16):
You have another quote along these lines, which is, "Everything you want is on the other side of worse first."

**Graham Weaver** (00:31:21):
Yeah, I mean, I know those are two non-super optimistic quotes perhaps, but I think they're true. The second one, everything you want is on the other side of worst first, and this is something where I almost can't think of many exceptions to this. Pick anything. You want a better body? Okay, you're going to need to go to the gym. When you go to the gym the first few times, it's going to not be that fun. You're going to set your alarm, you're going to get sore. It's not going to be great. You're going to have to probably make some changes to your diet. That's not going to be fun, at least initially. And so that's one example.

**Graham Weaver** (00:32:05):
You want to change careers, you're going to have to learn a new career. You're going to have to leave your career. You're going to have to maybe interview for new jobs, or whatever it is. In each case, the first move is negative. The first move is negative to getting in shape. The first move is negative to get out of a bad relationship, to get into a career you want to be in. The reason I think that's important to say is because if I'm optimizing for tomorrow and I just want to have a great day tomorrow, I'm going to stay exactly where I am, because my life will be better tomorrow if I don't make any changes. I don't have to break up with my girlfriend, have a hard conversation, have the tears, be alone, go on dating apps. I don't have to do that if I just stay in it one more day.

**Graham Weaver** (00:32:55):
So if you realize this and instead ask the question, the version of myself five years from now, what would they wish I was going to do right now? So I can guarantee your five-year version of yourself will say, "Get out of that toxic relationship, no matter how painful it is for the next two months." And if you can make decisions from that, and then on top of that, realize it's going to get worse first, then that's why I say everything you want is on the other side of worse first. But if you don't do that, you just end up plateauing. And so many people I see have this happen where they hit a plateau and they never move past it because they're not willing to have that hard day, month, week, year whatever it is.

**Lenny Rachitsky** (00:33:45):
I am imagining many people hearing this right now are like, "Yeah, I see what I need to do now." That was really powerful advice. It makes me think of parenting advice, some parenting advice I recently saw. Dr. Becky has this advice of your job as a parent isn't to make your kids happy, but it's to make them resilient.

**Graham Weaver** (00:34:02):
I love that, yeah. By the way, watch how people parent. They parent exactly the opposite of that.

**Lenny Rachitsky** (00:34:07):
Exactly [inaudible 00:34:09].

**Graham Weaver** (00:34:09):
You and I both live in Marin. I don't know if you have kids or how old they are, but when you get kids in school in Marin, you see parents, they try to clear all obstacles away from their kids. It's the worst thing you could do.

**Lenny Rachitsky** (00:34:25):
**Graham Weaver** (00:35:48):
Well, I think one of the things I would say is that I never stand up in front of students or be on a podcast like this and say anything that I'm not doing myself or that I'm advising students or people who work at Alpine to do. And so I appreciate the kind words. And I think that the things that I'm talking about are rooted in real results. And this is not just happy talk podcast. The formula, I think, for greatness is to be intentional, get in the path of the thing that you're most excited about, and then give yourself several decades to do it. And that's based on investing in 600 companies and building my own business. So I appreciate the kind call out.

**Lenny Rachitsky** (00:36:37):
Yeah. For people that are listening and be like, "Okay, I'm motivated. I want to do this," other than taking your class, anything you can recommend to just do these sorts of exercises, ask yourself these questions that you've seen work?

**Graham Weaver** (00:36:49):
I mean, the answer is accountability. How do you keep yourself accountable to living the life you want to live? And the analogy I would use is let's say that your number one goal in life is you've got to get in better shape. You just have to. Let's say you have even medically, you're going to have real health problems if you don't do that. I would say you hire yourself a personal trainer, pay what you need to pay. Okay, maybe you don't have a ton of money, but that's where I'd spend it. That person, A, they're going to hold you accountable to showing up at a certain time, and B, they're going to show you the exercises. They're going to call you if you're not there. You're just increasing the chances of success. Plus you spend some money, you want to get your money out of it.

**Graham Weaver** (00:37:40):
The equivalent of that in your life is an executive coach. And I figured this out in 2009 in the dark recesses of the recession. I hired my first executive coach. And I was like, "Wow, it's a personal trainer for me for two things." Number one, make space to ask yourself the big questions in life about your career, your relationships, your health, your spirituality, your children. Whatever the big things are in your life, ask the big questions, find out what your intention is. What are you looking for in those areas? And just have, in my case, several hours a week to get clear on those things. Okay, so that's part one.

**Graham Weaver** (00:38:25):
And then part two is that person can hold you accountable. I have one coach that I can't even have the call with the coach unless I fill out a piece of paper or an online form that says, "Here's what my one year goals are, outcomes I want to have this year are. Here's what I did last week, based on those. Here's what I'm going to do next week toward those. And here's the outcomes I want to have for the call we're about to have." And even if I never had the call, just having to fill that out every single week is incredibly powerful and allows me to hold myself accountable.

So 100%, that's what I'd recommend. Let's back up and say, okay, let's say you can't afford a coach or you're worried about that. And this is the same thing as, let's say you couldn't afford a personal trainer. I would give you the same advice, which is find a very like-minded friend of yours and sit down and do it for each other. So if you're using the workout example, okay, you're going to go on a run Tuesdays, Wednesdays and Fridays with your friend, and they're going to meet you at 7:00 AM at this trail, or 6:00 AM at this trail. That's your accountability. And you're going to be more likely to do it. You don't want to let them down. They're going to beat you up if you don't get there.

**Graham Weaver** (00:39:44):
Same is true with this. This is how I started. I did this with my roommate in business school. And we would go on a walk for 30 minutes and talk about my dreams and hopes. Then we'd turn around and talk about his. And it was great, because we made room for each other to have those conversations, and we also developed a great friendship. So short of having an actual executive coach, find a really-minded person that could get into this with you, and that would be another thing. But accountability is huge. I'm going to just say one more thing. I'm sorry. I'm going on a little bit long on this.

**Lenny Rachitsky** (00:40:20):
[inaudible 00:40:20] please.

**Graham Weaver** (00:40:19):
There's another thing that happens that's kind of magical, which is you activate a different part of your brain when you talk. You actually activate more of your brain when you talk than when you think or write. So thinking activates the least amount of your brain. Writing is a little bit better, but talking activates a whole different region of your brain. So that's the other big benefit of not doing this just on your own, is being able to talk about it with someone.

**Lenny Rachitsky** (00:40:54):
I love that there's a whole spectrum of ways to create accountability for yourself. I love the second coach you've shared, where just filling out that form was basically the biggest benefit.

**Graham Weaver** (00:41:07):
I learned this from these audio tapes. When I was in college, I had this green notebook. And I was trying to row crew, and I never row crew before. And I wrote down at the top of the page every single morning, I wrote down, "I am the number one rower in the country." I wasn't. I was a freshman novice, 135 pound Midwesterner, never had been in a boat. But I wrote that down. And then I wrote down three things I was going to do that day to move toward that goal. And I did that every single day that I was in college. And it's just incredible. We talked earlier about your subconscious mind. You're just locking your subconscious mind into your goals and where you want to go and who you want to be and how you want to show up. It's really powerful.

**Lenny Rachitsky** (00:41:46):
And that's advice anybody can implement. "I want to be the best [inaudible 00:41:51] founder"-

**Graham Weaver** (00:41:50):
You could do that. You could do that [inaudible 00:41:50] exactly.

**Lenny Rachitsky** (00:41:53):
Yeah. It's such a simple thing. "I want to be the best product manager. Here's three things I'm doing today to help me along the lines."

**Graham Weaver** (00:41:59):
Absolutely. My students have to do this twice a week. It's one of their assignments, and they actually have to turn it in. And I have so many students that say five years later, "I still do this a couple times a week, and it's been unbelievable." I'd say this to all your listeners: you will get more done writing down your goal and three things you're going to do to move toward that goal, you'll get more things done in three months than you will in three years without that.

**Lenny Rachitsky** (00:42:28):
That's an awesome thing to just try tomorrow at work. Write this down, see how it goes. The coach point, I just want to highlight that. What convinced me to get a coach back when I was working is I just realized, one, every athlete has a coach that tells them, "Here's how to become better," slash, everyone that you work with that is a leader and exec basically has a coach. And the people that have an exec coach will do better in their career and life than those who don't. So why would you not have someone there just helping you become better at this craft? It just makes so much sense if you can afford it. There are different price points for different coaches. Most people can probably afford it in some [inaudible 00:43:09].

**Graham Weaver** (00:43:08):
That's right. That's right.

**Lenny Rachitsky** (00:43:11):
Sweet. Okay. Let's talk about another framework that you have that you call nine lives. And this is essentially another way, it's like another way to hack your brain to come up with things you really should be doing, probably. Talk about this exercise.

**Graham Weaver** (00:43:23):
So this idea of what's my passion and what's my career goal, it can be really intimidating. And it is intimidating. And so this exercise is to make it less intimidating. And you basically come up with nine lives. So you say your first life, life one is the life you have now. So when I did this exercise ... let's pretend I did this exercise when I'm right out of school, taking that job. So life one is I'm working at this big firm, I live in the Bay Area, here's what I'm doing. That's life one.

**Graham Weaver** (00:44:05):
There's two rules. The first rule is all the lives have to start from today. So you can't go back in time. They all start from today. And the second rule is you have to be excited about all these lives. So I might say, "Hey, my second life is I want to start a private equity firm and be a founder and be a CEO." And that's life too. And life three is I want to be an author and I want to write whatever fiction or nonfiction. And then life four would be I want to be a professor and teach. And life five is maybe I want to make videos and be on social media. And life six, I want to be an actor. And you go through this whole list of these lives.

**Graham Weaver** (00:44:43):
And the idea is that ... A couple of things from this exercise. One is let's say that you're in a position where you need to be working in the job you're in. There might be one of these lives that gives you the most energy, and it's the thing if you knew you wouldn't fail, you would do this one. And it's good to recognize what that is and then pull that life a little bit into your current life.

**Graham Weaver** (00:45:10):
So let's say, Lenny, that you were a product manager, but you really loved doing podcasts, and that was one of your nine lives. Do a podcast every other week, and just pull it into your life, which it sounds like what you did. You started doing it as a side hustle. And then that'll have two amazing effects. One is you will just have more energy everywhere else in your life. Forget if you ever do this full time or not; just the act of pulling something into your life you're really excited about will give you a tremendous amount of energy.

**Graham Weaver** (00:45:45):
And then the second thing is you'll ideally find the path that is the thing that gives you the most energy. The other thing that I've learned in doing this exercise is you actually can have pretty much all nine lives. You can't have them at once, but if you're fortunate enough to live long enough, you can have all of these things. So I have had the corporate job, been the founder, been the professor, been a writer, had videos, taught people. I've been able to bring most of those lives into this current life.

**Lenny Rachitsky** (00:46:25):
There's a couple things there that come up as you talk about. One is, on that point, what I do now is my fourth career. First I was an engineer, then I became a founder, then I was a PM, and now I do whatever the heck this is. And I think people don't realize that's how life often goes. You think you're going to do one thing, and then you have many different careers that pivot into [inaudible 00:46:46].

**Graham Weaver** (00:46:45):
Exactly. And I think that's the thing. It's just trying to make it a little less intimidating. You don't have to have this one life purpose, passion, thing that you do for the rest of your life. If you just follow the thing that gives you energy at each time, it'll probably be a good indication of where you want to be going.

**Lenny Rachitsky** (00:47:03):
The other thread there is, I know one of your other really important pieces of advice is to avoid this not now idea, where everyone's like, "Here's the thing I should be doing, my genie goal, but not now." Thoughts on just how to think about that, of just like, "Okay, I know I have these lives I could live, but not now," on this or that?

**Graham Weaver** (00:47:21):
Yeah, I mean, in 20 years of teaching, I've never had a student come to me and say, "Hey, Graham, my real dream is to do X, but I'm just going to give up on it. I'm not going to do it." No one's ever said that. Instead, they say, "Not now." And not now, if they're not careful, will turn into not ever, because not now is just really another way of saying, "I'm not going to do it." And then there's a million reasons why you can't do it now. And those reasons, some of them are legit and some of them are just fear in another form.

**Graham Weaver** (00:47:58):
In terms of how to overcome that, I think it's kind of hopeful to realize that it's never really the right time. When you're making a change or you're going to go do something different, it's never going to feel secure and safe. You're always going to have some fear. You're always going to feel like you're not ready. You're going to feel like it's too soon. You're going to feel like you don't know exactly what that path looks like. And so just understanding that's normal. That's called entrepreneurship. That's called life. And if you wait for the clouds to part and this ray of sun to come down and say, "Now is the time," you're going to wait your whole life.

**Graham Weaver** (00:48:43):
And so I think that realization can maybe be helpful. And then try to figure out, what are the things that need to be true for you to launch? And usually for my students, it's something financial that is the big bottleneck. And what I tell them is, "You know what? People have raised money to start a business before. That's happened, where you've had people who've started businesses without their own money that have been able to pay themselves. And that's not a reason to not do it. It's an obstacle. It's something you have to solve, but it's not not an insurmountable obstacle."

**Lenny Rachitsky** (00:49:30):
This touches on a quote I definitely wanted to get to, something that you wrote not long ago. Here's the quote: "The most important thing I've learned in the first 50 years of my life is that the true game of life is an internal one, not an external one. And that journey starts with three powerful words: I am enough." Talk about that. Why is that so important?

**Graham Weaver** (00:49:51):
Well, I mean, that's a really a deep topic, but I'll talk about the internal and external journey. So life presents itself as a series of external obstacles and events. And it feels very much like an external journey. And it can feel that way your whole life. But I think what you'll realize, and I started realizing this when I really started meditating and spending time distancing myself from the subconscious thoughts, is a very, very large part of life is internal. I say a very large part because obviously you need some food, clothing, shelter, some basic needs that are external. But for most people that have the ability to even listen to this podcast, I would say the vast majority of your life is internal.

**Graham Weaver** (00:50:52):
So what do I mean by that? I mean that you're writing a story about what it is you think you need to be happy, or you're writing a story about things you think you need to be to be enough or to be respected or to feel worthy or to get admiration of other people. You're writing that story, and it's just a story. And if you really follow this logic, you'll realize that. You'll realize it is 100% just a story, or it's a story that you should even care. And then that opens up a lot more agency that you have over deciding what is important to you, what is your internal scorecard, what are the things that matter to you, not what the external world thinks or the story you've been writing for a long time. When you start to open this up, it's really kind of scary at first, because you'll start to realize most of the things you're operating from are really just stories that have been written at some point in your life. And so it's actually terrifying at first, and then it starts to become really liberating.

**Lenny Rachitsky** (00:52:10):
Was there an example of that in your life? Because externally, it feels like you're killing it: a killer PE fund, teaching at Stanford. The scorecard is looking good. So it's interesting you say that.

**Graham Weaver** (00:52:21):
The first time I really realized this, it was in 2015. I mentioned to you it took me 14 years to be successful. So we had just sold the last company from our second fund, which is where we really got paid. I had a financial event that was ... it wasn't like I never have to work again in my life. It was just like I could exhale a little bit. I knew I was going to be able to pay my mortgage and put my kids through college. It was that kind of an event. And for a couple days I was euphoric, because I felt like I'd worked so hard for this. I'd been, again, at this for 15 years, but really longer than that, if you go back to getting into college. And the whole thing had been a long journey.

**Graham Weaver** (00:53:05):
And then it hit me that nothing changed. Nothing internally changed at all. I still had the same problems. I still felt the same way about myself. I still had a lot of negative thoughts about myself. This goal that I thought that I had for this whole long period of time, it didn't actually change anything. It changed externally, for sure. Like I said, I could exhale and pay my mortgage. And those are all really good things. But that was the first time when I realized, "Oh, wait a second. It's really up to me to find things that are going to give me joy. And the achievement of some kind of external event is not one of those things."

**Graham Weaver** (00:53:54):
And I know that sounds really, really weird, but there's so many people that I've heard that have had very, very similar stories. And so it was really disorienting for me. And actually it was the first time in my life where I experienced depression, because I just had this thought of, "I think I was working my whole life for that, and it wasn't what I thought it was going to be." And so now I was thinking, "Well, what is it then? What is the thing that's going to give me joy?" And that takes some introspection to ask those questions.

**Lenny Rachitsky** (00:54:34):
I was just having a conversation with a friend who's an angel investor, and he just had a bunch of exits. And he's like, "Cool, I got some money in my bank account now, but I don't feel anything. I thought I'd be like, 'Holy moly, this is exactly what I was hoping for. And nothing changed.'" Exactly how you're describing.

**Graham Weaver** (00:54:53):
I think you get a little bit of peace of mind when you have some financial security, which is valuable. But in terms of now my life changes, now I'm enough, now I am happy, now I feel good about myself, none of that changes, for really anybody that I know.

**Lenny Rachitsky** (00:55:15):
And the hardest part, as you said, is you think it will. You think, "Oh, I'll be so happy once I achieve this thing." And I think an example of this is there's a lot of miserable billionaires, from what I've read and see. And that should tell you a lot.

**Graham Weaver** (00:55:26):
Yeah, exactly.

**Lenny Rachitsky** (00:55:27):
Maybe a second-to-last question; I'm curious if there's anything recently you're focused on, have been thinking a lot about that maybe you changed your mind about, or has changed the way you think about the world?

**Graham Weaver** (00:55:38):
The last two years, I had my two oldest boys go off to college, 2022 and then 2024. And that really hit hard. You would think that I would've been preparing for that for 18 years for each of them, but for some reason it just really hit me really, really hard. And I think it was a real wake up call of mortality, I guess, and to realize that nothing goes on forever. And these wonderful people that I had lived with, each for 18 years, were no longer going to really be a part of my daily life.

**Graham Weaver** (00:56:17):
And that really set me off on a journey of a lot of spiritual work, doing a lot of meditating and working with some gurus. And it's been really profound. And it's put me a lot closer in touch with the things that really matter to me. I've given myself more permission to spend time doing those things than the normal external world type things. So that was a pretty profound change for me. And the spiritual journey is arguably really the important journey. And this could be a longer conversation. And maybe you have the luxury of doing that journey as you get older or something, but it's been a really profound, profound journey.

**Lenny Rachitsky** (00:57:17):
I'm excited to see what insights come out of this part of your life. Final question before we get to our very exciting lightning round. We have this segment on the podcast called Failure Corner, where people come on this podcast, they share all these wins: "Oh, I have this amazing PE fund that's killing it. I teach at Stanford. I've launched all these things. All these students, they're so great. Life's amazing. Nothing ever goes wrong," when in reality, it does. And those stories often aren't told. So I'm curious if there's a story you could share of a time you failed in your career and, if you learned something from that experience, what you learned.

**Graham Weaver** (00:57:51):
Yeah, I'll tell a couple stories. So when I was in high school, I wrestled. And I cut a lot of weight to make the varsity team. And I was not in the best mental place because of cutting weight. But anyway, I lost a big match my junior year. And I quit and I never wrestled again. And that haunted me. So first of all, that was a failure, a big failure for me. And it really haunted me. And so I, after that, made a promise to myself that it wasn't going to have that happen again. When I went to college, I tried to row crew. I failed year after year trying to make the team, trying to make the boat. Eventually had some real success my senior year, but prior to that, just failure after failure.

**Graham Weaver** (00:58:44):
And then at Alpine, I mean, we lost money on our first fund. We had real trouble during the recession. I think five of my first eight investments I ever made in my life, I lost money. And in venture world, that's one thing, but in private equity, that's a whole different ratio, which is not a good ratio at all. When I first started teaching, I wasn't good at teaching, had a lot of insecurities. I was really young when I started, and I didn't feel like I had really anything to share with the students. And I think that showed up, and it took me a long time to kind of figure that out.

**Graham Weaver** (00:59:26):
So I guess almost my entire track record is one that starts with things not going well, and then just over a long period of time of chipping away, looks like a success on paper. But anytime early in the process would look like an abject failure. So I'm quite familiar with failure in the form of setbacks. I think the ultimate failure though was the wrestling one, where I just quit. That was really the only one I would characterize as a failure. The other ones, because I kept going and staying with it, turned out to work out well, with a lot of scars and bruises, but the failure would've been quitting.

**Lenny Rachitsky** (01:00:15):
I love that the circles back to your core advice of stick with it. Most things that are important take a long time and there's a lot of suffering [inaudible 01:00:24].

**Graham Weaver** (01:00:24):
Yeah, they do. Yeah.

**Lenny Rachitsky** (01:00:26):
Something I wanted to ask, I can't help but ask at this point, because I think a lot of people are wondering this, just when do you quit? When should you quit something? Because some things are just not worth it. Is there any advice there you could share?

**Graham Weaver** (01:00:37):
Yes, for sure. I think the time to quit is when you can no longer see the vision and you can no longer really believe the vision. And then when that happens for a long period of time ... or maybe you're no longer even excited about the vision; somewhere in there, I think. The excited one, you have to be a little careful of, because in the dark days, exciting is not the word you're going to use. But at least in our company, for the first 10 years it was not going well. But each time, we'd make fewer mistakes, we'd start to see something working. We'd do one really good deal in this fund, we'd start to learn from that. We'd get one really good hire, we'd learn from that. We would little by little start to see these green shoots.

**Graham Weaver** (01:01:24):
And I have this unbelievable statement. I didn't write it. Dan and Chip Heath wrote it in their book, Switch, which is "Scale your bright spots. Find what's working and do more of that." And as you start to progress, for me, for example, at Alpine, almost all the time, we always had at least a small glimmer of a bright spot. And then we'd scale that and then we'd continue forward and we'd find some more and we'd scale those. And over time, all those bright spots became our business. That became what we did. That became our strategy. That became how we hired people. That became where we recruited from. All those bright spots just started to magnify until the entire business was pretty much a bright spot. But it took time, because we had to figure out where those were. And we had to do a lot of things wrong to figure out where the bright spots were.

**Lenny Rachitsky** (01:02:19):
Graham, is there anything else that you wanted to share or you think is important to leave listeners with before we get to our very exciting lightning round?

**Graham Weaver** (01:02:25):
I mean, I think what I would say is just in general, you got one life, you get one shot. And so take the time to really figure out and answer the question, what does a wonderful, amazing, incredible life look like? And just get as clear as you possibly can on that. No matter how crazy or aspirational it seems, write it down. Write down that thing is write down that thing that would make this life amazing. And write it down for your life, your career, your relationships, your friends, your body, your spirituality, your financial situation. And just the first magic is just knowing what you want. And I'd say 90% of people never even know what they want. So take the time to do that. And the more clear you are on that, the more invested you are in that, the more likely you are to make it come true.

**Lenny Rachitsky** (01:03:18):
What I love about that is you don't have to do this thing, just step one is understand what it could be if you could do that.

**Graham Weaver** (01:03:18):
Exactly.

**Lenny Rachitsky** (01:03:27):
And it's almost like understand where your Google directions could take you if you turned off autopilot. Oh, man. Okay. Well, with that, Graham, we reached our very exciting lightning round. Are you ready?

**Graham Weaver** (01:03:37):
Let's do it.

**Lenny Rachitsky** (01:03:39):
First question, what are two or three books that you find yourself recommending most to other people?

**Graham Weaver** (01:03:44):
So in the realm of a lot of the topics we've been talking about, which is your internal and external game, I love the book Untethered Soul. And I love the book, Don't Believe Everything You Think. They have very similar themes, but they come at it differently. But I think both of those will really change your perspective if you read them.

**Graham Weaver** (01:04:08):
And then a very, very practical book that's probably the book I've read more than any other book, is How to Win Friends and Influence People, by Dale Carnegie, which was written in like 1930. There's no other book it. There's a reason that people are still recommending it 100 years after it's written. So it's definitely worth checking out.

**Lenny Rachitsky** (01:04:31):
Yeah, that book, I still think about it often, even though I read it 30 years ago at this point. I love that recommendation. And it's like a very old book to read, but you have to get past the fact that it was written a long time ago. Okay, next question. Do you have a favorite recent movie or TV show you really enjoyed?

**Graham Weaver** (01:04:48):
I, for the first time, watched the movie Where the Crawdads Sing. And I just loved it. It's kind of a romantic comedy, or not comedy, sorry, romantic love story meets murder mystery, meets coming of age. And it really, really touched me. Then I promptly read the book as well, so I love that.

**Lenny Rachitsky** (01:05:14):
Do you have a favorite product you've recently discovered that you really love?

**Graham Weaver** (01:05:17):
I'm a really big fan of sleep. I think it makes a massive difference in your life. And that's a whole other topic we could go down another time. So I have a few things that have helped me on that. So if you saw me sleeping, I have earplugs, I have a noise machine. I have a sleep mask, and then I have a Chilipad that goes on my bed to keep the bed cool. And I sleep great. All those things really help. The earplugs and mask and noise machine allow you to not hear the ambient noise. And then there's a lot of research actually on the temperature at which you want to sleep. And your body goes up and down throughout the night. So this Chilipad that goes under your mattress, there's a whole bunch of versions of that. And that helps a lot.

**Lenny Rachitsky** (01:06:11):
I also sleep with an eye mask. My wife and I rotate the earplugs, because someone has to pay attention to the baby, in case he wakes up.

**Graham Weaver** (01:06:18):
There you go. Yeah.

**Lenny Rachitsky** (01:06:20):
And this Chilipad, is this the Eight Sleep, or is this in just a cold bag?

**Graham Weaver** (01:06:24):
Well, no. So I actually bought the Eight Sleep, and it was too much.

**Lenny Rachitsky** (01:06:27):
Yeah, it's a lot.

**Graham Weaver** (01:06:28):
It would turn on and off. And I would wake up. And then it would track my sleep, and then I'd start to freak out because it'd tell me I wasn't sleeping well. So I actually returned it and I got a really simple one called OOLER. And it just turns on and off. There's no timing. There's no any weird functionality. And it was a lot cheaper, and it works better for me. So everyone use their own thing, but that one worked better for me.

**Lenny Rachitsky** (01:06:55):
Okay, two more questions. Do you have a favorite life motto that you often think about that you find useful in work or in life?

**Graham Weaver** (01:07:02):
I love this quote that sums up a lot of what we talked about in the podcast. It's by Howard Thurman. And he says, "Don't ask what the world needs. Ask instead what makes you come alive, because what the world needs most is for you to come alive." And I think that just talks about it's really about you coming alive that's the most important thing. And that is going to have so much positive exhaust in the world. And things from that will come that you can't even imagine right now.

**Lenny Rachitsky** (01:07:38):
I was thinking of that quote as you were describing your philosophy. Final question. I feel like a lot of people might be listening to this being like, "I came here for one of the most legendary private equity investors of all time, and you don't talk about private equity at all." So let me just use this opportunity to ask you just a question here. What do you look for in a company that you want to buy that maybe other people don't? Is there some insight you could share?

**Graham Weaver** (01:08:04):
Yeah. Well, I'm happy to talk about private equity, by the way. It's just that we talked-

**Lenny Rachitsky** (01:08:10):
We'll do another episode on that. This could be a whole podcast episode. I understand.

**Graham Weaver** (01:08:13):
Really happy to talk about that about. So I'll answer the question more like, what's a little bit of a different philosophy that we have? So when we were coming out of the recession and I hired this coach, we looked at all of our companies. And we were looking for where did we make our most money, and what was the most consistent trend? And we looked at valuation, growth rates, capital structure, geography, industry. We cut the data every way you could imagine. And we had these three companies that kept showing up on all these lists that were three of our top performers, but they didn't seem to really have anything in common.

**Graham Weaver** (01:08:53):
And then we're like, "Well, they have one thing in common, which is they started off really badly, so badly in fact that we put our own person from Alpine in to go run the company, and then they ended up becoming our best companies." So we said, "Wait a second. Maybe that's the highly correlated thing, is us putting our own team in place, and even upstream of that, maybe just having an incredible management team." So that was foundational. And now, we put our own leadership team in 100% of the time. And not only that, but we have spent an inordinate amount of time trying to build a program to help people who are in their late 20s, early 30s learn how to become CEOs. And that's been foundational.

**Graham Weaver** (01:09:38):
So the thing that we probably believe to be true that not that many people agree with us on is that the management team is really where we think all the alpha comes from. You can't get the industry wrong, because if you hire the best management team in the world to run a typewriter business, you're going to lose money. So you can't be wrong on the industry, but you also don't have to be perfect on the industry. You have to have a good enough industry and then a world-class management team. And we found that to be a really good formula for consistent returns. And it's way more fun because you're literally in a board meeting with someone that is on your side of the table, because you hired them and put them in. And so you're building the company together. And they're bringing a lot of similar values. And so it's been a real differentiator for us.

**Lenny Rachitsky** (01:10:32):
I have so many questions, but I'm going to cut it off there. We could do another episode going deep on all this. Graham, this was amazing. I think we're going to be helping a lot of people with what they want to do with their lives, and if nothing else, give them a little opportunity to break out of autopilot, at least for a little bit. Two final questions. Where can folks find you online if they want to maybe follow up, ask maybe some other questions that they are thinking as they hear this? And then how can listeners be useful to you?

**Graham Weaver** (01:10:56):
My website's grahamweaver.com, and I have a blog on there as well as a lot of videos and different things. And then on Instagram and TikTok, I'm grahamcweaver. And then on LinkedIn, I'm Graham Weaver. And on YouTube, I think I'm Graham C. Weaver. So Graham C. Weaver will pretty much get you on all those channels.

**Graham Weaver** (01:11:16):
How can listeners be helpful? I would say I'd love to hear from you. So my best way to reach me is grahamweaverblog.com. And tell me what's on your mind. I may not respond 100%, but I will read all the emails. And then if you're interested, subscribe to my blog. So go to my website, grahamweaver.com, and I have a blog where I talk about a lot of the topics that we're talking about today.

**Lenny Rachitsky** (01:11:44):
Amazing. Graham, thank you so much for being here.

**Graham Weaver** (01:11:45):
Thank you, Lenny.

**Lenny Rachitsky** (01:11:46):
Bye, everyone. Thank you so much for listening. If you found this valuable, you can subscribe to the show on Apple Podcasts, Spotify, or your favorite podcast app. Also, please consider giving us a rating or leaving a review, as that really helps other listeners find the podcast. You can find all past episodes or learn more about the show at LennysPodcast.com. See you in the next episode.

---

## [4/15] An operator’s guide to product strategy | Chandra Janakiraman (CPO at VRChat, ex-Meta, Headspace, Zynga)

**Chandra Janakiraman** (00:00:00):
I started noticing that there was a certain mystique and aura about product strategy. There was this perception that some people were intrinsically really good at strategy and others were not. It was almost as if there was a strategy gene you needed to be born with to be good at it.

**Lenny Rachitsky** (00:00:16):
Say someone's sitting down, okay, I'm going to start developing a strategy for our product. Where do you begin? What does this process look like?

**Chandra Janakiraman** (00:00:23):
In terms of what product strategy is? There is a smallest flavor of it which focuses on solving problems, they're called present forward, and it typically operates in a two-year horizon. We use a five-stage process to get there and it takes about eight to 12 weeks. The reason I think this process works is there is a ton of alignment built in. It goes back to human psychology of just something that comes from you, feels a lot more familiar and easy to accept.

**Lenny Rachitsky** (00:00:49):
Let's talk about big S strategy. When should you approach strategy this way?

**Chandra Janakiraman** (00:00:53):
There's this interesting quote by Elon Musk, which is-

**Elon Musk** (00:00:56):
"Life's got to be about more than just solving problems."

**Chandra Janakiraman** (00:00:58):
I think this is true of every sort of company. There needs to be an aspirational and cool component to strategy. What does the product look like in five to 10 years? Why is the world better in 10 years? And what is the most exciting version of that view?

**Lenny Rachitsky** (00:01:15):
Today my guest is Chandra Janakiraman. Chandra is chief product officer and executive vice president at VRChat. He was a product leader at Meta, chief product officer at Headspace, a GM at Zynga, and a senior PM at Amazon. And the way this podcast episode happened was an avid podcast listener, Karthik Suresh, told me about Chandra at a community meetup. And when I connected with Chandra, it was clear that I needed to get him on the podcast.

**Chandra Janakiraman** (00:04:55):
Pleasure to be here, Lenny.

**Lenny Rachitsky** (00:04:57):
I want to share actually the context and how this conversation happened. I was at a meetup of my readership, of my community and someone came up to me and they're like, "Lenny, you need to get this guy Chandra on your podcast. He's the most amazing playbook for developing a strategy." He's gone through it with you at a company he worked at with you once, and he's just like, "People need to learn this because it's so good." And to me, if someone can get better at strategy, it feels like it just makes so much of the way the company operates and the way that people work better. So we chatted, we met, I was like, "I completely agree. We definitely need to get you on this podcast to share your approach to the strategy." So we made this happen, and so we're here. So again, thank you for doing this and sharing.

**Chandra Janakiraman** (00:05:37):
Thank you. Thank you, Lenny.

**Lenny Rachitsky** (00:05:39):
First of all, I wanted to ask just you're very passionate about strategy and developing a way to consistently create great strategies. What got you so interested in this stuff in the first place? What kind of sparked your interest in this area?

**Chandra Janakiraman** (00:05:53):
Yeah, yeah, it's a very interesting story. It goes back all the way 10 years ago, but I remember it vividly like it happened yesterday. And I was a relatively new VP of product at Headspace, and we had this amazing company vision and mission that the founders had laid out, and I had come in and established goals for the team in terms of our key metrics, and we had a very buttoned up roadmap in my mind that fed into that company mission and vision. And I was feeling pretty good about how things were shaping up.

**Chandra Janakiraman** (00:06:37):
And on a particular Monday, the founder, CEO pulled me aside, and in his usual disarming style, he sort of made a short but profound statement. He said that, "Hey, CJ, I'm hearing that a lot of people don't really understand why we are working on what we are working on," and that was it. That was really the extent of what he shared. And it was a little bit of a bubble bursting moment for me because we obviously had spent a lot of time building the plan and I was feeling relatively good about the plan.

**Chandra Janakiraman** (00:07:21):
And so I spoke to a few people. I sort of wanted to understand it a little bit deeper, like, "Hey, what's happening?" And he was right. He was right. A lot of people didn't really understand why we were working on the things we were working on, and it led to some soul-searching. And basically I was lucky because there was actually a board member of Headspace who had a product background, kind of knew what good looked like, and I came to the conclusion, "Hey, we needed a strategy for Headspace."

**Chandra Janakiraman** (00:07:56):
So with extensive work with her, with the board person, we built the first written product strategy for Headspace. And that, and the subsequent actions on the product led to a complete re-imagination of the product. And basically we were able to create a new product, we call it the Next Generation Headspace, which on one hand it could support a comprehensive library of content, not just meditation, but non-meditation content as well. It had the sort of home experience where everything was incredibly personalized for the individuals, and there were several motivational elements built into the whole product experience. And it was very transformational for the company and the product because it changed the product from being a meditation app to a broader health and wellness service and really put the company on a different trajectory. It led to my promotion to the first CPO at Headspace.

**Chandra Janakiraman** (00:09:07):
And most interestingly, I had a chance to, while going through it almost in a sort of out of body way, observed the process of like, okay, how did we put this thing together? And what actually went into it? What really started as almost a personal crisis moment of finding this need to create a strategy for a product and a company led to a bigger sort of quest for me, which is I started noticing that there was a certain mystique and aura about product strategy, and there was this perception that some people were intrinsically really good at strategy and others were not. And it was almost as if there was a strategy gene that you needed to be born with to be good at it. And that bothered me a lot, and I sort of wanted to ask myself, is it possible to break that divide between the haves and the have-nots and make this capability widely accessible through a procedural approach. And I have news for you. The answer is yes, anybody can build product strategy through a clear understanding of what it is and through a friendly and repeatable playbook.

**Lenny Rachitsky** (00:10:29):
Amazing. That's exactly what I want to do here. The point you made about the why, I think everyone listening to this that's been in product for long enough has heard that of just, "Your team doesn't understand why we're doing this." I've heard that a number of times, as much as you think you're killing it, there's always that, you forget sometimes to do that or you aren't doing it great. And I love that basically the solution to that is the strategy solves that problem of helping people connect the dots and understand why this is the roadmap, why this is the strategy. Okay, so before we get into it, just one more context question, what's just the best way to think about what you're about to share and also who's it for? Who needs to hear this briefly?

**Chandra Janakiraman** (00:11:11):
So the way to think about this, the substance is that this is not a new framework or theory that I'm going to be talking about. There are plenty of excellent materials on strategy from ancient texts like The Art of War from Sun Tzu, all of Michael Porter's work, Good Strategy Bad Strategy by Richard Rumelt, who's been on your podcast, Playing to Win Lafley and Roger Martin. So there's a ton of extensively researched and well-founded materials out there. So the way I would think about this is, it's more of an operator's interpretation of all the stuff that's out there and are able to package it into something that's friendly and repeatable, particularly for product people who think they are weak at strategy or perhaps have received such feedback.

**Chandra Janakiraman** (00:12:05):
And just in terms of the battle-tested nature of it, I personally use this playbook about five to six times tweaking and optimizing it each time based on what I thought worked, what people thought worked and didn't work, including several times at Meta and usually leading to strong results, both for me achieving senior leadership alignment as well as driving business results for the company.

**Lenny Rachitsky** (00:12:36):
Okay, I'm getting more and more excited. Let's get into it, let's get into this playbook. People hear this word strategy a lot. They're told be more strategic, build a better strategy. What's the simplest way to understand what is a strategy?

**Chandra Janakiraman** (00:12:49):
Yeah, let's start with some basic definitions, which is as you asked, what is product strategy? And if you think back to the Headspace example, this sort of comes to life as well. So product strategy, strategy sits between the mission and vision and the plan. It could be at the company level or at the team level, but it's usually sitting between the mission and vision and the plan. And the plan, you can call the plan the roadmap, which is basically an ordered list of things that you want to get done and the mission and vision is basically sort of the purpose of existence, what does it look like when you achieve your purpose of existence? So it sits between the two and it forces choice to deploy scarce resources to generate maximum impact.

**Chandra Janakiraman** (00:13:40):
And I want to borrow an analogy from the world of physics. There is this concept called resonance, and the concept of resonance is really interesting, and it's actually very close to the concept of strategy. So the concept of resonance works as follows, when you apply a certain frequency to an object and you get pretty close to its natural frequency, you see a disproportionate increase in the amplitude of how that object vibrates. And so it's very interesting, if you apply any other frequency, there's very little effect on the object, but if you get close to its natural frequency, there's this exponential increase in the vibration of the product. So this concept of resonance is interesting.

**Chandra Janakiraman** (00:14:27):
So the way to think about it in the context of strategy is, it is selecting that frequency to achieve resonance between the product and the market. And so when you get close to that frequency, you should see tremendous impact in terms of the product landing well in the market. And so that's how I would think about it. It sits between mission, vision and the plan. It forces choice to deploy scarce resources to generate maximum impact, so using resonance as a sort of an example, and it ideally includes three components. The first is a handful of areas to focus on and I call these strategic pillars and then a whole bunch of areas that are explicitly not the focus. And the third component is why. So why are the focus areas, A, B and C? Why are these whole bunch of areas not the focus? And that's the three components. That's really it in terms of product strategy.

**Lenny Rachitsky** (00:15:32):
I love it. And I love the things we aren't doing as a core part of this, that comes up a lot on this podcast of being clear is what we're not going to do. We know these could be things we could do, but we're deciding we're not doing these things.

**Lenny Rachitsky** (00:15:42):
Okay, so let's just talk through how you go about developing a strategy using this method. Say someone's sitting down, "Okay, I'm going to start developing a strategy for our product." Where do you begin? What does this process look like? Let's start talking through it.

**Chandra Janakiraman** (00:15:59):
So I sort of want to first explain this concept called smallest strategy, and I'll sort of talk about what that is and how it's different from another kind. But the basic sort of strategic process I would say takes about eight to 12 weeks long. It's something that I think people often underestimate how long it takes and eventually end up taking a lot more time. But when they start off, they think, "Oh, I could probably stand this up in a couple of weeks," but usually through iteration it actually ends up taking 8 to 12 weeks anyway. So it's good to start with setting clear expectations that it takes about eight to 12 weeks. And the way to justify the ROI on that is typically a strategy like this can be leveraged for about a couple of years. So relative to that sort of payback period, I think the investment is relatively small. So it's pretty healthy from that sense to manage expectations and say that that's how long it's going to take.

**Chandra Janakiraman** (00:17:00):
So within that period, there are five phases. There's the preparation phase, there's a strategy sprint, the design sprint, the document writing and the rollout. And those are the five phases, which I'll explain how one would go through. And basically each of them has a certain sort of time recommendation. For example, the preparation phase I would say is probably about four weeks. The strategy sprint is up to about one week. The design sprint is another one week, document writing maybe one to two weeks, and the rollout is maybe two to three weeks. So that's how you get to that range of eight to 12 weeks.

**Lenny Rachitsky** (00:17:40):
Essentially it's like a quarter of work to get to a final great strategy. And of these five phases, the biggest bucket is preparation, which to me sounds like it's not like a full-time team thing. It's starting to gather data and user research-

**Chandra Janakiraman** (00:17:56):
Totally.

**Lenny Rachitsky** (00:17:57):
As you talk through this, I'm curious just how much of the team is involved in each of these steps. But I think it's an important point, if you want a really good winning strategy, you need to give a time. You can't just say, "In a month or a week, we need to develop a strategy, go figure it out, write this document." Great. Cool. So let's talk about step one preparation.

**Chandra Janakiraman** (00:18:14):
That's correct, Lenny. And I think you, if I forget, remind me that sometimes there is pressure. There's just business pressure, like the CEO might still want a strategy in two weeks, how do you respond to that? And I think we can find some clever shortcuts there, but I think to the extent possible, the leader should push for [inaudible 00:18:39] to make something really great.

**Lenny Rachitsky** (00:18:40):
And I think part of it is that as a leader, you can start on this before... you know this is coming, so you should get started before it's even asked for.

**Chandra Janakiraman** (00:18:48):
That's correct.

**Lenny Rachitsky** (00:18:49):
Yeah, awesome. Okay, let's get into it.

**Chandra Janakiraman** (00:18:51):
So I think you sort of touched on this in terms of the preparation phase being the longest phase, but also not being sort of like a full-time thing. So that's absolutely right. So the preparation phase is really... The way to start this, which is a little different from other approaches I've seen from people is to actually form a strategy working group. This is an important concept. So the strategy working group is sort of a small team. It typically consists of engineering, product design and data at a minimum. And in certain cases, there's a luxury to have other functions like product marketing, user research, that's also part of the strategy working group. But the minimum quorum, I would recommend is engineering, product design and data because design in some ways represents both sort of product design and user research. So you do get the voice of the user from that perspective. And typically the PM is driving the strategy working group and the process, but that working group is actually the team that's going to collaboratively create the strategy doc.

**Chandra Janakiraman** (00:19:59):
And so in the preparation phase, there's usually a kickoff meeting where the PM pulls the team together, talks about the purpose of the process, lays out the different phases, and gives everybody a feel for what's going to happen in the next 8 to 12 weeks. And basically then creates a list of very discrete action items and deliverables for each of the stakeholders in the working group. So specifically, there's an action item around aggregating all the behavioral insights that the team might have around the product. And usually this is a combination of previous analysis that the team has run on the data side and potentially also feature launches and how they have done. So all kind of analytical analysis.

**Chandra Janakiraman** (00:20:50):
The ask is really to create a meta analysis of all of the analysis. So the data person on the strategy working group has to scan the historical archives at the company and sort of synthesize and condense that into a very sort of digestible macro themes and learnings about users. So that's sort of one preparation phase item.

**Chandra Janakiraman** (00:21:16):
The second is UXR insight. So again, there's probably a lot of soft hard signals about users, not just based on research that's run by user researchers, but also potentially from the customer service team, social channels, and basically a meta analysis of all of that into one very actionable and synthesized deck on all the insights on users. That's usually led by the design person and with support from their research team. And that's sort of the second action item.

**Chandra Janakiraman** (00:21:53):
The third action item is leadership interviews. So I sort of have this fun story of the fruit with leadership strategy reviews, which I want to share. So imagine, this is sort of how sometimes strategy reviews go, which is you bring a fruit to the reviewer and say, "Hey, here's a mango, what do you think?" And the reviewer says, "I actually don't like mangoes." And you're like, "Oh," you're sort of sad. You take it back, you bring an apple and you show, "Hey, what do you think of an apple?" And then the leader says, "I actually stopped eating apples last year." And so you're disappointed again. You go back, you bring a banana. "I hated bananas since I was a kid." And so it's a bit of a silly caricature of reviews, but there's a bit of grain of truth there, which is, imagine how frustrating that is for both the reviewer and the person who's reviewing.

**Chandra Janakiraman** (00:22:54):
And so it could be made so much better if you just engage with your leaders before you actually build a strategy. It's amazing how few people actually do that. And so the fruit story tells you, hey, imagine if you just asked the reviewer, "Do you even like fruits?" How much better the experience would've been for both parties. And so leadership interviews are a very important part of a strategy formulation process. So you can divide and conquer, there's several leaders, like you assign different leaders to different people on the strategy working group and each of them sort talks to that leader. And there are a few questions that I would recommend asking, and it's basically, what does success feel like for the leader? What does failure look like? What is the measure of success? What are principles to keep in mind while going through this process?

**Lenny Rachitsky** (00:23:51):
And these are centered around the product that you're working on?

**Chandra Janakiraman** (00:23:55):
Exactly.

**Lenny Rachitsky** (00:23:55):
For Headspace, it'd be like, what do you think success looks like for Headspace or the specific feature of Headspace? [inaudible 00:24:03].

**Chandra Janakiraman** (00:24:02):
Exactly.

**Lenny Rachitsky** (00:24:03):
Okay, cool.

**Chandra Janakiraman** (00:24:04):
That's right, yeah. And also I think ask them for their favorite or better ideas. It's actually what I've found through this process is a lot of leaders have these better ideas, they just feel shy to share it because they don't want their teams to think of them as micromanage-y. They want their teams to figure out the answer themselves, but then when you ask them, they actually have a pet idea always.

**Chandra Janakiraman** (00:24:29):
And so asking them just takes the mystery out of it. And it also gives them a creative avenue. So some people feel nervous about engaging senior leaders in these conversations in the sense that, hey, is it a waste of their time? What I've found is the exact opposite, senior leaders are, they welcome this because one, it's actually a more fun conversation for them than the other meetings that they have in their day because they're getting a little bit of their creative juices going, and they actually feel happy that somebody actually asked them what they're feeling and thinking about. And so it's actually very, very positive energy when you ask leaders just what they want. And it's also not a sign of weakness. It's actually a sign of strength and humility to ask your leaders what you want. And so keeping that fruit story in mind, I want to just say that this is a very, very positive thing, a very powerful thing.

**Chandra Janakiraman** (00:25:26):
The next area is competitive analysis. So if there is a product marketing person, they can do it for you. If there isn't, the PM should do it themselves. And basically the idea is you try to understand who are the comparables or the competitors in the space, and you sort of build a little bit of a head-to-head and sort a stack chart of where's everybody going and what are of the angles of investment for different people based on the explicit signals, which is what are they releasing?

**Chandra Janakiraman** (00:25:59):
You don't really know what their strategies are, but you can kind of tell when you look at the features they're putting out that, oh, they seem to be focusing on this particular area. And so that's sort of competitive analysis. And bigger companies, there's also another important input, which is adjacent roadmaps. Are the teams adjacent to you and what are they investing in? Because oftentimes that can have a rub-off effect on your team, and if you're not aligning with other key teams, it's going to be important. So adjacent roadmaps and a summary of that.

**Chandra Janakiraman** (00:26:33):
And last but not least, is what I call user observation. So I like the strategy working group to actually either interview a user or watch a video and report key learnings. And the idea is not to action those insights, it's really to build empathy. When you get somebody in the room with a user, it just changes their mind, it softens them a little bit. It gets them out of their own preconceived notions of what to build, what the strategy should be and...

**Chandra Janakiraman** (00:27:00):
... their own preconceived notions of what to build, what the strategy should be, and it humanizes the whole process. That I think is the purpose, but you still give them the homework of writing down what they learned because it's a little bit of a forcing function.

**Chandra Janakiraman** (00:27:16):
The output of all of this is what I call the comprehensive preparation readout. It is a single master deck where you sort of have the behavioral insights meta analysis, you have the UXR insights meta analysis, you have a download of the leadership interviews, you have the competitive stack charts, you have the adjacent roadmaps, and you have a section on user observations. It's a lot of work, but to your point, it can be done in parallel with your day jobs. You can multitask, and that's why you take about four weeks to do that. That sort of concludes the preparation phase. You get the deck, and that rolls into the strategy sprint, which is the next phase. I want to pause and see if you have any questions.

**Lenny Rachitsky** (00:28:06):
Yeah, I have a million questions, but I'm going to keep myself contained. Just to quickly summarize. Basically you kick off, we're going to start developing strategy for this product, and you were going to touch on this, but it may help just to talk about this right now. I know you have big S Strategy, small S strategy. This process, what's an example of the level of product scale that this process is for? Because I know you have another approach for a larger scale of a product, basically for a company. What's the best way to think about small S strategy, which is what we're going through?

**Chandra Janakiraman** (00:28:37):
Yeah, I think it's a good question, Lenny. I would say that the process works well at sort of a growth stage company or in a vertical within sort of a larger company. I think it scales pretty well. The main difference between this small S and big S is the time horizon aspect and the aspirational component, which we'll get to.

**Lenny Rachitsky** (00:29:02):
Okay, so this is a process you can use for entire company strategy of not a large company, not a massively, and then also just a product within a company?

**Chandra Janakiraman** (00:29:11):
Yeah.

**Lenny Rachitsky** (00:29:11):
Okay, perfect. Okay. Hey, we're kicking off strategy for this thing. Let's say VR Chat 2.0. You have this kickoff meeting with your working group. You assign action items and there's six things you ask everyone to do. Gather all the behavioral insights, all the user research insights that you've had, leadership interviews, interview people, ask what they want and see what they're hoping for, what success looks like, competitive analysis, adjacent roadmaps across other teams and user observations, just like watch users, see what's happening. You assign each of these tasks to different people in this working group. You set a deadline. We have to present this in say, four weeks. People go work on it. You meet ongoing as it's coming together, and then the output is a deck that you share. Do you share this deck with just that working group, or who do you share this deck with and read it out to?

**Chandra Janakiraman** (00:30:02):
Yeah. The deck is an output of the first phase, and then it flows into the second phase, which is a strategy sprint. You don't share the deck with anybody yet.

**Lenny Rachitsky** (00:30:10):
Got it. This is our secret.

**Chandra Janakiraman** (00:30:11):
In the strategy sprint, it's like a key thing. Yeah.

**Lenny Rachitsky** (00:30:13):
Awesome. Okay. I love how very concrete and actionable this is with time boxes and action items and the exact output you're looking for. I love this. Okay, cool. Basically what you're trying to do here is gather all of the input that will inform the strategy.

**Chandra Janakiraman** (00:30:27):
That's correct.

**Lenny Rachitsky** (00:30:28):
It's giving yourself time to do this, because if you think about it, the output is determined by the quality of your input. I love, it feels like a core component of this method is spend time creating, gathering all the input, all the best input, actually spend time there. Don't just make it a half-assed last minute thing. Okay. Amazing. Okay, let's talk about step two, which is the strategy sprint.

**Chandra Janakiraman** (00:30:52):
Yeah, so the strategy sprint is the heart of the process. This is sort of where you make the decision. If you recall, the definition of strategy is it forces choice to deploy resources into a few areas for maximum impact. The core of strategy is really picking those areas and the areas you're not going invest in. That happens in the strategy sprint, so really it's the heart of the process.

**Chandra Janakiraman** (00:31:22):
Typically, it's like a three to five day process. The first day is the share out date. Everybody has done some great work. They go in, they share what they've collected, what they've sort of learned. It brings everybody in the working group to the same state of understanding, the same state of consciousness on the state of the union. That's a great process where what I encourage people to do is write down things that you are, as you're listening, write down things that you find are problems for our users, things that are coming in the way of our growth and things that are sort of suboptimal for the business.

**Chandra Janakiraman** (00:32:05):
People take a lot of notes during that time and the people who are presenting are sharing everything they've learned. That's really day one. It's just absorbing a ton of information and writing down a ton of notes so people can kind of understand where all the problems are. It's a very problems focused process, and that's an important point. I'll get back to it when we talk about big S, which is different. Once people have that common awareness, shared knowledge of all the problems, day two is literally the most important day in the entire 8 to 12 weeks because that's where you actually make the choice. The way to flow through it, it's actually really important to flow through it correctly. The first step is really generating a whole bunch of problems, because people have been taking notes the previous day.

**Chandra Janakiraman** (00:32:59):
You start the day with, "Hey, let's collect everybody's thoughts on what the problems are that are holding us back." It's a free flowing session. Everybody throws out their observations on what's holding us back, and you just capture all of that in a Google Sheets, for example. Over even an hour, you start to see these patterns emerging of like, there's these clusters of problems that are really holding us back. The next step is you do a joint clustering of related problems and you create these potentially, typically in my experience, I've seen about 10 to 15 clusters form of very related problems. The beauty of it is each of those bigger cluster, you actually know what the sub problems are within that cluster because you sort of generated it very organically.

**Chandra Janakiraman** (00:33:55):
Then you have, let's say 10 to 15 clusters. What you then do is you, because remember, because it started as a problem generation exercise, each of the clusters also has a name that is a problem, and so the next step is to flip it into an opportunity framing. Let me give you a couple of examples.

**Chandra Janakiraman** (00:34:16):
Let's say there's a bunch of problems around people don't really know where to find different things in our product, and they don't really know where to go, where to find a certain feature or a certain experience. There's let's say a lot of problems in that area, so difficulty finding things becomes the cluster, the problem cluster and discovery becomes the opportunity sort of framing of it. It's sort of the more positive framing of it.

**Chandra Janakiraman** (00:34:46):
Another example could be that people get a lot of content that they don't like. They see a lot of stuff that they don't like, and so they disengage with the product. The opportunity framing of that would be relevance. It's basically stuff that really matters to me. Maybe if it's a social product, then maybe people are finding it difficult to find friends and they are lonely because of that. The opportunity framing would be social connection.

**Chandra Janakiraman** (00:35:20):
Flipping all of those problem clusters into positive framing and opportunity framing is the next step. Then you're in a great spot because now all you have to do is you have to down select from those 10 to 15 opportunity areas into ideally three, maybe five, but I would recommend three because it creates more clarity and focus. The way to do that is really sort of ranking them on, I would say four or five key dimensions or criteria, and the first is expected impact.

**Chandra Janakiraman** (00:35:56):
Let's say you actually tackle that area. What is the expected impact to whatever matters to you as a company, as a business, as a product? The second dimension is certainty of impact. Certainty of impact is basically how concrete is the evidence that this is a problem? Sometimes you have really hard data, sometimes you have more anecdotal evidence, and so the confidence really depends on how big the problem's sizing and frequency is. Expected impact, certainty of impact. The third one is also very important, which is clarity of levers.

**Chandra Janakiraman** (00:36:34):
Do you actually have an idea of how you would solve it? If you don't, it's going to be really difficult to move the needle on it because you should kind of know that, okay, I can imagine these solutions could actually move the needle. I can actually launch this sort of nudge system that can help people find things. I can recommend people. I can recommend friends so that people can form friends quicker on the platform. You should have a sense of how you would solve that particular space. Is there clarity on levers? That's the third dimension.

**Chandra Janakiraman** (00:37:08):
The fourth dimension is super important, which is are the levers unique and differentiated to that particular team or company? Which is that if another team or company could build it better than this particular team or company, then it's probably not that differentiated. It's probably going to be pretty generic once you launch it. It's a combination of sort of like, is there a lot of impact here? How confident are we of the problem? Do we have a sense of the solutions? And basically, are we the team or company that has the capabilities and the skills to uniquely build it where other teams cannot?

**Chandra Janakiraman** (00:37:52):
Once you have that, and sometimes what happens is you don't have too much data, and so it's okay to have qualitative scores on this like high, medium, low, T-shirt scores, whatever that is. The key is you're doing it together as a strategy working group, and you're debating the scores and you're reasoning why it should be higher versus lower. There's a ton of alignment and collision that's happening when you're doing that, which is very healthy for the eventual outcome.

**Chandra Janakiraman** (00:38:22):
Once you do that, basically you can do a simple sort of addition of the scores and a sort, and what you have is basically the top three and you have the remaining 7 or 12 that are basically not the focus. That's the core of the process, is getting those opportunity areas and getting to a shared sense of how do we prioritize them and why?

**Lenny Rachitsky** (00:38:51):
Okay, and this is all done in a week. I know there's more to it. There's a couple more items that help you move from what you just said to the next step, but I love that this is the core of the biggest element of the process, and you do it in a week and you're only able to do it in a week because of the work you did ahead of time. Again, highlighting the importance of that prep step.

**Lenny Rachitsky** (00:39:13):
Just to share what you've shared so far, and then we'll finish the strategy sprint step. It's basically do the share out so everyone's on the same page about all the information that, all the inputs essentially. Enumerate all the problems, like individual small problems and then cluster them into 10 to 15 problem clusters. Flip it from here's the problem to here's an opportunity we have. Rank them based on, basically there's these four attributes you shared, which I'll actually, I wrote these down. Impact potential, confidence that it will have the impact, clarity of levers and are they differentiated unique levers? Is there something different from what other folks are doing? Those are ways to rank these ideas and problem clusters. You essentially come up with, here's three bets basically we potentially should take. Okay, and then I think that's where you stopped. Is that right?

**Chandra Janakiraman** (00:40:03):
That's totally right, Lenny.

**Lenny Rachitsky** (00:40:06):
Okay, cool.

**Chandra Janakiraman** (00:40:06):
Those three that are at the top of that pile are basically our strategic pillars. We've sort of gotten our strategic pillars and they basically hold up the strategy. That's why they're called strategic pillars. The idea is once you have the strategic pillars, we basically translate that into a few how might we's? Let's say it's a relevance thing. How might we find the best content for a particular user? How might we surface it in the right place? There's a few how might we's, and the how might we's are basically intended to help the next phase of the process, which is the design sprint.

**Chandra Janakiraman** (00:40:43):
You generate these areas, these strategic pillars, you generate the how might we's. How might we's are typically pretty straightforward. Once you have the strategic pillars, it takes probably an hour to generate some how might we's for each of them, maybe two or three for each strategic pillar, and then you're done with that stage.

**Lenny Rachitsky** (00:41:00):
I just want to highlight real quick, this phrasing is really important. I used exactly the same phrasing. I had a PM that I worked with and his name was Andrew Chen, but not the Andrew Chen people know about, that's an investor, who had this concept of fertile questions that create ideas and spark ideas and solutions. This phrase, "How might we", is actually really powerful as a way to come up with ideas to solve problems. It's just like, how might we increase discoverability in our app? How might we improve relevance? There's something magical about that phrasing that it opens up your mind to, how might we? Let's think about it. Versus like, how do we improve discovery? That's a different, your brain works differently hearing this. That is a really powerful phrasing. Just wanted to highlight that.

**Chandra Janakiraman** (00:41:47):
Yeah, that's awesome. It's also something that designers are familiar with, Lenny, so it flows really well into sort of a design sprint.

**Lenny Rachitsky** (00:41:56):
Okay, so you have these how might we'd. You have three pillars, maybe three or four or five how might we's to solve these opportunities/problems, and then what happens after that?

**Chandra Janakiraman** (00:42:06):
Exactly. Then the third day is it's good to start fresh. The team's accomplished a lot.

**Lenny Rachitsky** (00:42:13):
This was all the first two days.

**Chandra Janakiraman** (00:42:14):
This is the first two days, yeah.

**Lenny Rachitsky** (00:42:15):
Wow. Okay. So much done in two days.

**Chandra Janakiraman** (00:42:18):
Yeah, exactly. The second day is particularly intense on the team, so it's good to give them a break, because it's a lot of really mental sort of wrestling, and so give the team a bit of a break. Then the third day is when people are a little bit refreshed, we get to winning aspiration. Winning aspiration is super interesting because it's a very creative exercise. You basically imagine in two years, because that's the typical time horizon of a smaller strategy, is like 18 months to 24 months. This is what I tell the team. Imagine two years there's a newspaper, there's a journalist that covers this work, and there's a newspaper article that comes out. I want you to imagine the progress on all these strategic pillars and what the headline of that newspaper article looks like. It's called a newspaper headline approach.

**Chandra Janakiraman** (00:43:09):
Basically, everybody generates a newspaper headline in parallel. It's interesting because you often see there's these common themes that sort of form when people generate these headlines. The forcing function with the headline is also that it has to be somewhat simple and plain speak. It's not too technical. You have to get to a more simple layperson's language and you have to get to the key benefit that ultimately, the impact it has on the world. Those kinds of themes come up often.

**Chandra Janakiraman** (00:43:44):
Then you do a little bit, you put them all into a blender, you put all of those headlines from the team into a blender, and you mash them together and create the winning aspiration, which is ultimately what does progress on the strategy look like in a couple of years time? That comes from the working group, it's not like one person writing.

**Lenny Rachitsky** (00:44:09):
What's an example of an aspiration that you've come up with on a project you worked on, a winning aspiration?

**Chandra Janakiraman** (00:44:14):
We did this process for the privacy team when I was at Meta. Really the strategic pillars were around a lot of features that we would build, but the winning aspiration was really like, could we move consumer trust? The newspaper headline is something like, "Facebook has moved the needle on consumer trust by investing in these areas." That's how you create the bigger impact.

**Lenny Rachitsky** (00:44:47):
That's a great example. Obviously this is similar to the PR Amazon method and what I love about your approach, as you said early on, is you're just pulling together all of these awesome ideas from all these different methods of strategy work to a very methodical step-by-step process. Taking all the best ideas into, and as you described, an operator's playbook for doing this. I love that. Okay. When you say blender, by the way, I was thinking as you were talking, put them into a blender, what I'm inferring is you just take everyone's headlines and you come up with one that kind of covers the gamut of all three pillars being successful.

**Chandra Janakiraman** (00:45:20):
That's exactly right. More tactically what I typically do is I put them all on a slide and start, it's almost like a word cloud, and then you start to see these common words. Then you converge those words as the key elements of the final winning aspiration. Then you try to create a nice statement out of that. It also symbolically, everybody sees their own statement on the deck, and that's how you get to the final model.

**Lenny Rachitsky** (00:45:46):
Awesome. I feel like AI could help with that now.

**Chandra Janakiraman** (00:45:46):
Exactly.

**Lenny Rachitsky** (00:45:48):
Just put in all your headline ideas and it comes up with some suggestions. Okay, cool. Is that the end of the sprint or is there more to the sprint?

**Chandra Janakiraman** (00:45:55):
That's it for the sprint.

**Lenny Rachitsky** (00:45:57):
Okay. At the end of the sprint, what do you have? What are the outputs?

**Chandra Janakiraman** (00:46:00):
Yeah, this is a great progress on strategy because now we have the three strategic pillars. We have the how might we associate it with the strategic pillars. You also have the why. Why did you get to those three strategic pillars and what are you not focused on and what are the reasons for it? You also have the winning aspiration. It's great progress. The team's done a great job. I think now we sort of move on to the design sprint.

**Lenny Rachitsky** (00:46:30):
In your experience, how often are these three the correct three that you end up going with versus you learn something over the course of the rest of the sprint and adjust?

**Chandra Janakiraman** (00:46:38):
It's a very good question. We'll speak about an example that I had in Meta where typically during the strategy sprint, you don't change it once you go through the strategy sprint, but eventually there are a lot of signals you get through execution where you sort of have to course correct. We'll talk about one very interesting example of how that sort of changed things here.

**Lenny Rachitsky** (00:47:01):
Awesome. Okay. Basically, you've developed your strategy at this point, and in your experience it ends up being... Until you hit the market and test, you don't really know what's going on.

**Chandra Janakiraman** (00:47:01):
Correct.

**Lenny Rachitsky** (00:47:10):
But there's kind of an implication that in your experience doing this, I think you said five or six times, it has been correct and as good as doing it any other approach.

**Chandra Janakiraman** (00:47:21):
I would say it's not sort of an empirical study, obviously because of the small sample size, but I would say that it's really opened people's eyes and it's led to really good alignment and eventually good results. Even when it has not, it has led to good organizational buy-in on why and how we are approaching things.

**Lenny Rachitsky** (00:47:42):
This reminds me of, Tomer Cohen was on the podcast, he's CPO of LinkedIn and he has this phrase that always says he says, which is, "We may be wrong, but we're not confused." I love that a core part of this is everyone is completely on the same page, not from the beginning, but once you start this process of here's all the inputs, here's step by step we are working together on narrowing down, so at least everyone understands the why. Which is what sparked your interest in this in the first place, everyone understanding. I know there's a whole rollout at the end too to solve that.

**Lenny Rachitsky** (00:48:14):
Okay, cool, so we have the sprint. We have basically the three pillars that you're going to invest in, this headline of what it might look like if you're to launch and some solution ideas with how might we's. What comes next?

**Chandra Janakiraman** (00:48:26):
Yeah. The next phase is the design sprint, and the design sprint can be led by the design person who's in the strategy working group. If you remember, there's engineering, product design and data, so the design person can lead it and the PM can sort of take a bit of a backseat during this process. The input to the design sprint is sort of the three strategic pillars and the how might we's associated with it.

**Chandra Janakiraman** (00:48:50):
The goal of the design sprint is not to sort of come up with these are the features we should build. That's not the purpose of the design sprint. The design sprint is to generate a lot of illustrative concepts that bring the strategy to life because a picture is worth a thousand words. Oftentimes, even though you might have the right words in your strategy doc, people might still scratch their head like, "What do you exactly mean? What are you going to build?" The illustrative concepts really sort of give people something to latch onto, "Oh, okay, I get it. This is what you're going to build at the end of the day with this strategy."

**Chandra Janakiraman** (00:49:25):
The more generative, the better. Then there's the ability to, sometimes if you do the Google Ventures design sprint, you can even test some of it with users and get a little bit of sharpening of things. The goal here is not to build feature-ready designs, it's more to generate concepts. Once you have that, basically the design sprint, and there's different flavors of design sprints, which I won't go into. You can lean on your design lead to decide what's the appropriate way, but the input and the output is what needs to be very emphasized. The input needs to be the strategic pillars. The output needs to be a ton of illustrative concepts to explain each strategic pillar. You could almost have a section where you talk about each strategic pillar and insert those concepts in that section. That's the idea.

**Lenny Rachitsky** (00:50:16):
We're going to go through a couple examples to make this very real. You mentioned the design sprint method, which we had the authors on the podcast. They actually didn't go through the design sprint method. They have another book called Make Time that's about productivity. We also had their colleague from Google Ventures that designed a bullseye sprint, which is also an interesting sprint. It's a new thing that helps you figure out your ICP and who to focus your product on and how that informs your product.

**Chandra Janakiraman** (00:50:40):
Oh, that's right. Interesting.

**Lenny Rachitsky** (00:50:41):
It's another type of sprint you could use here. The input is here's the three things we're going to invest in, say discovery, relevance, privacy. The output is here's concepts of what this could look like to get people's minds going. Is it just a bunch of mocks basically in a deck at this point?

**Chandra Janakiraman** (00:51:00):
That's correct. Yeah.

**Lenny Rachitsky** (00:51:01):
Okay, cool. Okay, amazing. That's a week. That's another sprint. PM could maybe take a little bit of a break.

**Chandra Janakiraman** (00:51:06):
Exactly.

**Lenny Rachitsky** (00:51:07):
And designers take the lead.

**Chandra Janakiraman** (00:51:09):
Exactly.

**Lenny Rachitsky** (00:51:09):
The engineers are kind of inputs and thought partners in this, I imagine.

**Chandra Janakiraman** (00:51:13):
They're optional. Yeah.

**Lenny Rachitsky** (00:51:15):
Optional, but yeah, ideally they're involved because you want all the best ideas in there.

**Chandra Janakiraman** (00:51:19):
Yep.

**Lenny Rachitsky** (00:51:20):
Okay, cool. What comes next?

**Chandra Janakiraman** (00:51:21):
Then the next step is the document writing. This is sort of a solo activity that the PM should take on, but the great news is the PM is not starting from scratch. There's so much great stuff to write. If you remember, there's a ton of user insights, a ton of behavioral insights, a ton of competitive analysis. There's the three strategic pillars, how might we's, winning aspiration. There's a whole bunch of illustrative concepts.

**Chandra Janakiraman** (00:51:48):
Oftentimes, product leads have this sort of creator's block. That is solved here completely because you have a ton of great material. I have to tell you, it doesn't make the job easier. You still have to weave together a good story. That's why I think it takes a week or two. I think it's really about combining, connecting and editing at this point and telling a cohesive story from all those components, but I think the building blocks are really solid and defensible.

**Lenny Rachitsky** (00:52:17):
Is there a template or kind of sections you like to include in your strategy doc? As someone sits down and is trying to write this out, what do you want to see there as headings?

**Chandra Janakiraman** (00:52:26):
I think you could almost take the building blocks as a little bit of a steer for the template. You start with the broader context where you talk about what the leaders kind of want from this overall effort. Then you get into key insights and analysis where you have user insights, behavioral insights, competitive analysis. Then you get into the strategic pillars and you explain them. You also explain why. In the appendix, you include the full table that you generated on day two of your strategy sprint. You include the full table in the appendix, including the criteria. That's going to be really important because most people are going to ask, "Why did you pick these?" That's basically sort of the defensibility.

**Chandra Janakiraman** (00:53:17):
Then you have the winning aspiration that's very bold. It's a big part of the heart of the deck, and you sort of embed the illustrative concepts into the actual each of the strategic pillars so that it flows well. Then finally, you end with some kind of alignment questions like, "Hey, do these feel right? Are there things we are missing?" So that it creates that framework for alignment in the subsequent meetings.

**Lenny Rachitsky** (00:53:45):
I'm excited to chat with Christina Gilbert, the founder of OneSchema, one of our longtime podcast sponsors. Hi, Christina.

**Christina Gilbert** (00:53:52):
Yes. Thank you for having me on, Lenny.

**Lenny Rachitsky** (00:53:54):
What is the latest with OneSchema? I know you now work with some of my favorite companies, like Ramp, Vanta, Scale and Watershed. I heard that you just launched a new product.

**Lenny Rachitsky** (00:54:00):
... and to scale and watershed, I heard that you just launched a new product to help product teams import CSVs from especially tricky systems like ERPs.

**Christina** (00:54:09):
Yes, so we just launched one scheme of file feeds, which allows you to build an integration with any system in 15 minutes, as long as you can export a CSV to an SFTP folder.

**Christina** (00:54:18):
We see our customers all the time getting stuck with hacks and workarounds, and the product teams that we work with don't have to turn down prospects because their systems are too hard to integrate with. We allow our customers to offer thousands of without involving their engineering team at all.

**Lenny Rachitsky** (00:54:31):
I can tell you that if my team had to build integrations like this, how nice would it be to be able to take this off my roadmap, and instead use something like OneSchema, and not just to build it, but also to maintain it forever?

**Christina** (00:54:43):
Absolutely, Lenny. We've heard so many horror stories of multi-day outages from even just a handful of bad records. We are laser focused on integration reliability to help teams end all of those distractions that come up with integrations.

**Christina** (00:54:55):
We have a built-in validation layer that stops any bad data from entering your system, and OneSchema will notify your team immediately of any data that looks incorrect.

**Lenny Rachitsky** (00:55:02):
I know that importing incorrect data can cause all kinds of pain for your customers and quickly lose their trust. Christina, thank you for joining us, and if you want to learn more, head on over to oneschema.co. That's oneschema.co.

**Lenny Rachitsky** (00:55:17):
I pulled up Playing to Win, which I know you also pull ideas from, and actually Roger Martin was on the podcast talking about this stuff. And one way I think about it as you're describing a way to break up your strategy doc, is he has these five questions that you ask.

**Lenny Rachitsky** (00:55:31):
The first is actually what's your winning aspiration? So I love that you're pulling that in. This is one approach your doc would be. What's your winning aspiration? Where will you play? What market are you going after? How will you win? What capabilities must be in place for you to win? And then what management systems are required?

**Lenny Rachitsky** (00:55:46):
We'll link to this framework just in case people want that as a crutch in their thinking strategy, but is there anything that we could link folks to that describe how you like to think about this doc? And if not, it'd be cool if you make a template that folks could borrow.

**Chandra Janakiraman** (00:56:00):
I'm happy to share of the flowchart of the process and the template as well.

**Lenny Rachitsky** (00:56:05):
Awesome. Okay. So this is how long of a process, this writing of the document?

**Chandra Janakiraman** (00:56:10):
It's about one to two weeks, and I think it's mostly solo work and hopefully by the end of it there's a very tight doc, and that's what we use to roll out.

**Lenny Rachitsky** (00:56:21):
And by solo work, I imagine you're looping in this working team to get their feedback as you're pulling it together, or is it just you sit there in a room and then...

**Chandra Janakiraman** (00:56:27):
I would minimize pulling them in, because they've all contributed so much to the process already. So, I would say that at the end of it, obviously, once you have a draft, it's good to share with them, but I wouldn't tap their cycles too much at this point, because they've given a lot already.

**Lenny Rachitsky** (00:56:47):
I see. Okay.

**Lenny Rachitsky** (00:56:48):
And how many pages do you see this doc being, roughly? What's a heuristic?

**Chandra Janakiraman** (00:56:52):
It's not too long. I would say probably three or four pages and then an appendix has a lot of additional... Like, the table I spoke about and a lot of additional maybe illustrative concepts. Maybe you can only use a few illustrative concepts in the main section so that there could be others there.

**Chandra Janakiraman** (00:57:09):
I think usually there's a desire from leaders to say, "Okay, what are we building next?" And it's important not to include a roadmap as part of a strategy doc, because a strategy doc is meant to be separate from the roadmap. It's meant to be a companion to your roadmap. And even though there's interest, maybe sometimes you can include a illustrative roadmap in the appendix, but I would try to keep it clean and try to keep it focused on just the strategy.

**Lenny Rachitsky** (00:57:40):
So at this point you basically developed your strategy, the next step, I think you call rollout, where you just start actually rolling this out, so let's talk about that. But it's important to note how many weeks in this is. This is like six-ish weeks of work and you've got a strategy for your company/product?

**Chandra Janakiraman** (00:57:58):
That's exactly right. So I think you're probably in the last two to three weeks of the process, and pretty important final step is the rollout, and I would start with what I call gatekeepers. And these are people who are absolutely... You have to get their one-on-one alignment and blessing on this before it moves forward. And it's probably not too many, probably two or three people. So I would pre-flight it with them and get their alignment.

**Lenny Rachitsky** (00:58:28):
So these are one-on-one meetings with these gatekeepers?

**Chandra Janakiraman** (00:58:29):
These are one-on-one meetings. Exactly.

**Chandra Janakiraman** (00:58:31):
And then there's a larger group of what I call key stakeholders, people who are impacted by it, different functional leaders, et cetera. And that can be done either async or through a group review. And then there's probably a rolling of list of team roadshows. There's different ways to do this, but the one that I feel is most effective is the roadshow where you have about eight to 10 people in each session, so people feel more comfortable to ask questions and it's more conversational.

**Chandra Janakiraman** (00:59:06):
So the purpose of this stage is to land it. It's not to seek too much feedback, so it's a delicate balance. At the same time, you don't want to appear just being too inflexible. So it's a very delicate balance. When people ask questions, you can clarify it, and you can add clarifications to the doc, but I wouldn't change... The most important thing, three strategic pillars, I wouldn't change that. I would defend it using the framework, but if people are like, "there's really good arguments about the criteria that led to your ranking," then it's okay to reconsider it. I've not seeing it happen in my five to six attempts, but it's possible, theoretically possible.

**Lenny Rachitsky** (00:59:47):
The core of this approach it sounds like, is you've done a lot of the pre-work where you land in a generally correct place.

**Lenny Rachitsky** (00:59:55):
Okay, so this is the rollout. So what I'm thinking as you talk, say you're like a PM on a team, say you're working on privacy, it's like an IC, your ICPM, working on strategy. How do you think about including, say, your manager because through this process, when do you start to like, "Hey, here's what we're planning, here's what we're thinking," making sure they're on board? Because I could see this working team off to the side working, working, working. "Okay, we're ready to roll it out." Oftentimes there's like, "No, wait. There's all this other stuff happening at the company, we don't have resources." Where do you loop them in? How do you think about that kind of stakeholder?

**Chandra Janakiraman** (01:00:29):
The the more attuned PMs who understand organizational dynamics, probably keep the manager pretty in sync through the process. I think, definitely, the manager becomes a person who you interview as part of the leadership interviews, so you know what the manager's looking for from the effort.

And then once you get to the strategic pillars on day two of the strategy sprint, you probably want to just quickly pre-flight it with the manager and say, "Hey, look, this is how it's trending, and these are the things we are probably not going to do" and any issues with that. And then eventually you actually want your manager to support you in some of these bigger meetings, so you enlist their help. So I would say that it's probably each individual style, but I would keep them pretty aligned through the process. But you don't have to be too heavy [inaudible 01:01:23], just be super light.

**Lenny Rachitsky** (01:01:24):
Great. Makes a lot of sense to me. There's like a...

**Lenny Rachitsky** (01:01:26):
We're not going to solve everyone's problems with this one framework, maybe just quickly touch on resourcing. As you think about this. Like a part of a strategy includes like, oh, we also need these resources. Just any thoughts on how to include that? And then I want to get into some examples of how you've actually implemented this.

**Chandra Janakiraman** (01:01:42):
So I actually don't recommend thinking about resources in the strategy phase, because what you're saying is, "these are the areas of focus" and the resourcing question becomes more relevant from a road mapping standpoint.

**Chandra Janakiraman** (01:01:55):
Because then you say, "Okay, what percentage of our engineering do we put on strategic pillar A, versus B, versus C? And what are the specific things we build?" So it becomes a road-mapping question as opposed to a strategy question.

**Lenny Rachitsky** (01:02:08):
And so by the end of this, the rollout, the part of the rollout is developing the actual roadmap. We're not going to get deep into that. That could be part two, solve every PM's problems, teach them all the steps of the process, but we're going to cut it off at, here's you have a strategy that is starting to be rolled out.

**Chandra Janakiraman** (01:02:24):
Exactly.

**Lenny Rachitsky** (01:02:25):
Cool.

**Lenny Rachitsky** (01:02:25):
Let's go through a couple examples where you actually move into this to make this even more real. I know there's a couple companies you were thinking about sharing, strategies you worked on.

**Chandra Janakiraman** (01:02:35):
I think there's just three quick notes to close off this process and I'll share two examples, Lenny.

**Lenny Rachitsky** (01:02:44):
Perfect.

**Chandra Janakiraman** (01:02:46):
So the reason I think this process works, the first is because there is a ton of alignment built in within team alignment and leadership alignment built in. And it's not seen as, "Oh, this PM went off and wrote this strategy doc and I don't agree with most of it." And part of this is actually very...

**Chandra Janakiraman** (01:03:09):
It goes back to human psychology of just something that comes from you, feels a lot more familiar and easy to accept. So this doc is actually not from the PM, the PM is facilitating it, but it's actually from the strategy working group. And the strategy working group are the leads of the team. And so it's actually... And the leadership inputs have been baked in, so it's actually very team representative. And so hopefully there isn't too much misalignment when you roll it out. And I have seen that, I've seen that work out.

**Chandra Janakiraman** (01:03:43):
The second is there's just better results. You get to better problem articulation, you get a better strategic pillar, because there's just more minds on it than if it was just a product lead.

**Chandra Janakiraman** (01:03:55):
And the third thing is you have clearly defensible criteria and outputs, and if you want to change it, like I said, you have to go back to the criteria and the scoring and then say, "Okay, why do we believe this has to be changed?" And even a change is easy to justify once you do that. So it creates a lot of benefits, but ultimately, and we'll get to this a little bit more, it has to be tested with execution, and that's the most important thing.

**Lenny Rachitsky** (01:04:22):
I think it's important to highlight, no framework playbook [inaudible 01:04:26] method is going to guarantee you have the correct strategy that will win and your company will thrive. It's always just the best effort at plan.

**Lenny Rachitsky** (01:04:35):
As the quote, "no good plan survives first contact with the customer."

**Chandra Janakiraman** (01:04:41):
Exactly.

**Lenny Rachitsky** (01:04:42):
Cool. Let's talk through some examples.

**Chandra Janakiraman** (01:04:45):
so the first example I wanted to talk about is at Zynga. I was at Zynga a very long time ago, and this was the heyday of social gaming on Facebook. The thing that I was extremely impressed about Zynga was the strategic clarity and strategic encoding at the company. And this was not attributable to me, by the way, this was already there when I got there as a entry-level PM, but the kind of strategic clarity there was really, really impressive and it was very evident and observable from all the games.

**Chandra Janakiraman** (01:05:22):
So if you look at all the games, there were three elements that were very common across all the games. The first was viral game loops. There's these game loops that just required you to have a social and active social network to be successful and play, and it was very tightly and fictionally integrated into the core of the gameplay. That was the first one.

**Chandra Janakiraman** (01:05:47):
The second one was there was this idea of paying to complete things. So you're not paying to skip a whole experience, what you're paying is just to complete it. So what happens then is different people, depending on how much time they have, complete different percentages of a progression task, or a part of the gameplay, and you only had to pay the rest to get through the experience. And it was this interesting experience where people put a lot of investment into it, and they didn't mind paying for the last 20% or 30%, which was very interesting, because it created this market for different elasticity of spend, different times that people had in their lives. And it was very, very clever. So that was again, very common across all the games.

**Chandra Janakiraman** (01:06:38):
And then the third was network. So all of Zynga games had this cross-promotional component at the top of the game and they would promote other games, and it's like the Zynga network was the most important thing, and not an individual game. And those were the three in our parlance now strategic pillars, and there were non-focus areas. It wasn't like...

**Chandra Janakiraman** (01:07:05):
The focus wasn't on high-fidelity graphics, it wasn't on complex game mechanics, and this was extremely clear and hard-coded into the company culture and operations. And it was perfectly tuned to the environment at the time. Facebook platform afforded strong support with social graph, and channels, and basically the game studios contributed through different games in a very network accretive way.

**Chandra Janakiraman** (01:07:31):
So for example, I was at Zynga San Diego, and we took the company into net new game genres like action strategy, match three puzzle games, but we stayed true to the strategic pillars of the company, and we had to invent new mechanics so that the strategic pillars would work in the new genres that we introduced to the company. And what was fascinating is the company actually had systems to reinforce these strategic pillars.

**Chandra Janakiraman** (01:07:57):
And, for example, the data infrastructure was very incredible at the time, and really reinforced these strategic pillars. There was this function called Central Product Management, which basically propagated best practices, made sure that games were network accretive and all these things worked in harmony to enhance those strategic pillars and reinforce it, and it worked really well. The company, if I remember right, got to a billion dollars, it was the fastest to get to a billion dollars in the history of companies at the time.

**Lenny Rachitsky** (01:08:32):
In revenue.

**Chandra Janakiraman** (01:08:33):
In revenue.

**Chandra Janakiraman** (01:08:35):
And it worked really well until the environment changed and there was a shift to mobile. And temporarily... Basically, if you go back to that resonance concept, that deep resonance between the product and the market was temporarily lost once that shift happened to mobile.

**Lenny Rachitsky** (01:08:51):
So there's a lot here. I think one interesting note here is, as you said, the strategy work you're doing, which sounds like a lot of time, eight to 12 weeks potentially, this lasted a long time for Zynga. And so I think it's important to remember the work you're doing here, even though...

**Lenny Rachitsky** (01:09:06):
On the one hand it feels like a long time, on the other hand, it feels like very little time to come up with the things that will most help your business grow. In this case, these three elements for Zynga. And by the way, these three pillars, they didn't emerge from this method, but it helped you see the power of being very clear...

**Chandra Janakiraman** (01:09:24):
Clear, exactly.

**Lenny Rachitsky** (01:09:25):
And having everything centered around, we all agree, these are the three ways we win. Okay, cool. And then I think the number three, again, I just want to highlight the power of just very few bets in investments. So it's always three. I've always suggested three as well. Some people are like, "three to five," but I think it's... You just find, in general, three is the right number as much? Try very, very hard to make it three?

**Chandra Janakiraman** (01:09:46):
I agree.

**Lenny Rachitsky** (01:09:47):
Okay, cool.

**Lenny Rachitsky** (01:09:48):
And then I think there's also this element of differentiation being really important. So these three elements you shared for Zynga are just like they're unique and differentiated for Zynga. Network, powers of apps driving other apps, paying to complete, these are things Zynga has figured out, "this is how we" win versus other products in the market.

**Chandra Janakiraman** (01:10:08):
Totally.

**Lenny Rachitsky** (01:10:09):
Sweet. Anything else along those lines with Zynga?

**Chandra Janakiraman** (01:10:13):
That was in hindsight reflection, Lenny, I was pretty naive when I was there, and I realized how... And this was, in fact, what led to my blind spot when I joined Headspace from Zynga, that you even need a strategy. That's the story I said at the beginning.

**Chandra Janakiraman** (01:10:33):
The reason is because Zynga had it figured out so well that I didn't actually have a need to exercise that muscle at Zynga, and we just had to come up with new games that applied that strategy. And so it was a net new muscle I had to develop when I got to Headspace.

**Lenny Rachitsky** (01:10:48):
It's like that parable of the fish swimming in the water where the older fish swims by and he's like, "How's the water?" And they're like, "Hmm?" And then he leaves and he's like, "What's water?" You just don't realize what you're in. You don't realize what you're surrounded by until it's gone.

**Chandra Janakiraman** (01:11:06):
Exactly.

**Lenny Rachitsky** (01:11:07):
Maybe one other thing I just want to highlight while we're on this topic is just the power of focus. In the case of Zynga, just this focus on everything we do needs to get these things, these three things, everything we do, we need to do in order to win. And I think that's a recurring theme in what you're talking about, is just focus. The steps of the process are focus on this one part for now. So I think that's a really interesting thread is just the power of focus.

**Chandra Janakiraman** (01:11:32):
Yeah.

**Lenny Rachitsky** (01:11:33):
Okay, cool.

**Lenny Rachitsky** (01:11:35):
And now you have another example of your time at Meta.

**Chandra Janakiraman** (01:11:37):
At Meta there's this fascinating example, which I think illustrates a different point, which is...

**Chandra Janakiraman** (01:11:46):
Basically I was standing up the product strategy for a couple of product growth teams in Reality Labs, Oculus, which was coming up with the Quest II at the time. This is a standalone headset, and then Portal, which was our video conferencing product. And we stood up these teams to go after product growth, which is basically driving hardware sales through software features. And we went through this process that I just described. We stood up strategic pillars, and the strategic pillars were fairly similar for both Portal, which is our video conferencing product, and Oculus.

**Lenny Rachitsky** (01:12:24):
I love Portal by the way. I was a big fan of... Sad that it's no longer around.

**Chandra Janakiraman** (01:12:31):
And that led to features that are known to everybody, like the Oculus Referrals Program, the Portal Memories Integration, where you see one of your Facebook memories on the Portal, sponsored ad in your Facebook feed. And also we had this section on the Facebook app, which is also from Facebook at the time, and then eventually also from Meta, which is these other products from the Facebook ecosystem. And all those came out of that effort.

**Chandra Janakiraman** (01:13:01):
And about 18 months into it, they actually had very different outcomes. So on one hand the Oculus effort was incredibly successful, and we graduated it into the VR division at Meta, and till today it continues to be an incredibly successful effort. The Portal effort actually didn't move the needle as much as we wanted to, and we sunset it, and we basically redeployed that team to other initiatives. And that's super interesting. Basically it was like the same strategy process. We got the nearly identical strategic pillars, but eventually completely different outcomes.

**Chandra Janakiraman** (01:13:49):
And I think that that illustrates the most important point about strategy, which is intrinsically strategy has no business value. It's basically a document with a few words. And I think it starts accumulating value as you generate business impact and results. And that happens when you actually test strategy with execution.

**Chandra Janakiraman** (01:14:15):
And so ultimately any strategy is only as good as the results it produces. And so one has to have the intellectual honesty, and the humility, and the courage to say when it's working and when it's not. And sometimes what happens is parts of your strategy might work, parts might not, and you have to pivot away from some and double down on some. But I do think there's that evaluation that's really critical and testing strategy with execution.

**Lenny Rachitsky** (01:14:44):
Such a great point. Just a strategy sitting there in a doc is worthless, where the worth comes from, is it actually having impact and being successful. Makes me think about a product manager also, just a PM is not worth anything until they help you drive impact.

**Lenny Rachitsky** (01:15:02):
I think an important, to trickle down from that point, is that you don't want to spend too long just thinking about strategy, you want to spend as little time as possible to get to a strong hypothesis, basically to start learning if this is the right path. And so I love that your approach is like this middle ground between give it real time, but don't spend three, four or five months working on this one document that just maybe one day you'll use.

**Chandra Janakiraman** (01:15:31):
Exactly.

**Chandra Janakiraman** (01:15:31):
And there is room for a six-month process, which we'll get to in a moment. But for small S I wouldn't recommend more than two or three months.

**Lenny Rachitsky** (01:15:39):
And again, small S strategy is for a couple of years out, not like Exactly. It's like a timeframe. Okay, cool.

**Lenny Rachitsky** (01:15:47):
Okay. Anything else along those lines of the examples of Meta or Zynga?

**Chandra Janakiraman** (01:15:53):
Those are pretty good lessons I would say.

**Lenny Rachitsky** (01:15:55):
Okay, sweet.

**Lenny Rachitsky** (01:15:56):
So, let's talk about big-S strategy. When should you approach strategy this way and what are just the steps of it?

**Chandra Janakiraman** (01:16:06):
So everything we've spoken about so far is what I would call small S, and it's very problem focused. It's basically what I call present forward.

**Chandra Janakiraman** (01:16:17):
It's like you have something, you have a product out there, it has a bunch of problems. How do we make it better for... What are the areas we tackle for maximum impact? Typically led by product managers. And there's this interesting quote by Elon Musk, which is, "Life has to be about more than just solving problems." And he says it in the context of the aspiration to become a multi-planetary species, but I think this is true of every company, big or small. And there needs to be an aspirational and cool component to strategy. And I call this big-S strategy.

**Chandra Janakiraman** (01:16:56):
I'll run through this at a higher level because I think this is a little bit more fluid in how you build the big-S strategy. And it typically takes longer, potentially up to about six months. The approach is a bit different from small S. And you start with the company mission and vision, and there is a little bit of groundwork that is done on long-term cultural trends, social trends, competitive trends, technological trends, and those are all the backdrop to trigger ideas. And what you do with that backdrop is you, again, do these leadership interviews, but with a different goal of generating long-term futures.

**Chandra Janakiraman** (01:17:41):
And some of the questions you can ask during these interviews are "What does a day in the life of a user look like in five years? What does the product look like in five to 10 years? Why is the world better in 10 years? And what is the most exciting version of that view?" And basically take all of that input and cluster it into I would say, three cohesive holes.

**Chandra Janakiraman** (01:18:05):
And what I mean by that is three, at least fairly distinct, descriptions of the future. So to give you almost a very simple example, imagine you doing big S for the future of travel. You could have a future that is talking all about fully autonomous travel, where there's very little human involvement in going from A to B. You could talk about another future where there's extreme speed, where you can get from A to B really fast across the world that is really fast. You could talk about a third where there is no travel, it's virtual travel, and you still have the feeling of travel, and you accomplish the same goals.

**Chandra Janakiraman** (01:18:45):
But those are different futures, they have different properties, like different elements to them. And once you generate those distinct futures, you actually generate prototypes with learning goals. And think of these prototypes as concept cars. And the automobile industry uses this notion of concept cars. Concept cars, the interesting thing about concept cars is they're never commercialized. They're often produced for inspiration, and to potentially take some part of them, like maybe a technology or a feature, and that is brought into mainstream production.

**Chandra Janakiraman** (01:19:23):
So think of these prototypes as these concept cars that really drive inspiration and potentially give you some small nuggets that you can run with. And then you start doing research with them with potential users, you answer key questions, and you uncover certain elements that resonate with people. So you eliminate a whole bunch of stuff, you combine a whole bunch of stuff, and you establish some winning components that are interesting.

**Chandra Janakiraman** (01:19:48):
And then what you do is you push stuff that is winning into the product, live product, testing. So this is actually something that you actually want to start testing your way into and understanding if it works from a scalable standpoint. And this whole thing is typically led not by the PM team, but by design and UXR. And intentionally it's a little bit more open-ended and green field. And it actually is a very different mind space that people have when they approach big. So the roadmap is really built through a combination of small-S and big-S work. And, for example, at VRChat, we are doing both small-S and big-S work, Lenny. And they are run as parallel work streams. The product management team is leading the charge of the small-S work, and the design team is leading the charge on the big-S work. And it's really exciting stuff. What's coming out of both work streams is really exciting and really different. And so we are building the bridge from both sides and both work streams ultimately flow into one roadmap. It's almost like two tributaries that ultimately merge into one river. That's how I would think about it.

**Lenny Rachitsky** (01:21:02):
There's so many things here that are so...

**Chandra Janakiraman** (01:21:01):
That's how I would think about it.

**Lenny Rachitsky** (01:21:02):
There's so many things here that are so interesting. One is we've had this conversation on the podcast a couple of times, this point that people think and see the world differently. Some people are very open-minded and creative, and think big, blue sky people. Some people, me, just like, "What are we doing next? How do we move this metric? Let's talk about concrete things that we can do?" For this big S approach, I think what I'm hearing is make sure the people leading it are very open-minded, creative, blue sky type people. If you're the person that's like, "How will this move our metric?" Maybe you shouldn't be leading it and give someone else the reins.

**Lenny Rachitsky** (01:21:40):
Cool. And then with these timelines of doing both at once, it's probably hard to align them fully. So what I'm hearing is it's like these are kind of ongoing sometimes.

**Chandra Janakiraman** (01:21:49):
Correct.

**Lenny Rachitsky** (01:21:50):
And they'll both inform what you're doing. You don't need to have this perfect timeline of-

**Chandra Janakiraman** (01:21:54):
I agree to this. Yeah.

**Lenny Rachitsky** (01:21:56):
Sweet. And the big thinking I love, because sometimes, you may experience this, designers often are like, "Oh, we just work on all this boring, incremental optimization stuff. I want to think bigger." Which is great, and this is a really good lever to allow for that in a contained environment. Cool. Let's just go crazy. Let's think about what this could be in the future. Still, let's make sure we're moving some metrics short-term, but let's give us opportunity to think big. And oftentimes the biggest ideas come out of that. So I love it's giving space for both types of thinking, thinking bigger and thinking long term.

**Lenny Rachitsky** (01:22:30):
Anything else about these two methods that might be interesting to share? Otherwise, I want to talk about AI a little bit, just how that impacts things and a couple other things.

**Chandra Janakiraman** (01:22:39):
Yeah, yeah. I'll end on one short sort of note on how this all feels when you do it, and then we can segue into AI-

**Lenny Rachitsky** (01:22:47):
Great, great.

**Chandra Janakiraman** (01:22:48):
... Lenny. So I think that whether it's big S or small S, these are incredibly satisfying processes to go through in the end. So once you get to the end, they're incredibly deeply satisfying. But with anything in life that is very deeply satisfying at the end, they have a ton of challenge, frustration, and dead ends while you go through it where you kind of get a lot of self-doubt like, "Hey, will you ever reach the end?" And I just want to sort of normalize that. That's actually normal and I want people to expect that as they go through it. And I think that's what makes it even more rewarding when you get to that sort of point where you start to see things connecting.

**Chandra Janakiraman** (01:23:33):
The second thing is I would say that the person who's leading the charge on these efforts has to be really good at connecting diverse viewpoints and keeping them all moving forward. And it's not easy, it's very hard because sometimes you have people who take you in different directions and you have to keep it all hanging together.

**Chandra Janakiraman** (01:23:57):
So in terms of picking people who have that skill, who are good integrators, who are good connect the dots, I think it's critical for these to be successful. And in some ways low ego because it's not like... They do get to introduce their own ideas, but truly, it's about bringing the team together. So that's the skill that you should look for for these leads.

**Chandra Janakiraman** (01:24:22):
And then the third piece is I would say that I would just approach these with a little bit of a lighter touch and more playful sort of approach, because it's an intensive process, it's a long process. And people can get tired, they can find it grindy. So just having a little bit of playfulness along the way goes a long way in making it feel more tolerable for the people going through it. So those are some thoughts about either of these processes.

**Lenny Rachitsky** (01:24:48):
That was really important context, especially this will be frustrating throughout. It sounds really beautiful and great and smooth as you describe it. In reality, there's a lot of pain that goes into... Because you're making hard decisions. A lot of people have opinions, they have perspectives, there's data that's... Rarely is there clear answer from the beginning, so I think that's really important context.

**Lenny Rachitsky** (01:25:08):
I wanted to actually come back to your point you made about how Elon made this point about life shouldn't just be about solving problems. I think part of that quote is just like we should work on things that are awesome and exciting and inspiring, that aren't just immediately pain solvers. And he's very good at that, just inspiring people to what could be, and we should be thinking much bigger than we are.

**Lenny Rachitsky** (01:25:32):
I was actually at an interview with Zuck where he said the same thing. He's probably inspired by Elon. He's like, "I've just gotten to a point in my career where I want to work on awesome stuff, stuff that is just awe inspiring, not just-

**Chandra Janakiraman** (01:25:43):
Exactly.

**Lenny Rachitsky** (01:25:44):
... another social feature." so I think that's a really important point. And there's a lot of power in that. People get really excited about holy moly, I did not imagine this is what Headspace could become, what VRChat could become, what Meta could become. It's just really powerful, getting people really inspired. And I love this process they shared of how to do that.

And I didn't actually summarize it, so I have it right here in front of me. I'll just share the five steps of the biggest process. So it's preparation, mission, you figure [inaudible 01:26:10] mission, vision, trends, interview people about where they think things are going. Then you come up with three distinct futures of what the future might look like if you were to do the things you're thinking. Then you build prototypes of what might this look like? And then you actually test these in your product, which I love, to start de-risking these ideas. That helps you converge to here's actual plan we want to execute, and then you turn that into roadmap.

**Chandra Janakiraman** (01:26:36):
Yeah, yeah. And the first testing of the prototypes is with UXR, so it's more sort of concept testing or prototype testing with a handful of users. And then you use that as synthesizing to a live product test.

**Lenny Rachitsky** (01:26:50):
Great. Basically, it's like big ideas and then de-risking, validating them along the way-

**Chandra Janakiraman** (01:26:51):
Exactly.

**Lenny Rachitsky** (01:26:55):
... in various ways to like, "Okay, maybe this could actually work. Let's try." Sweet. Okay, so just two more things. One is I want to talk about AI briefly, of how that impacts this work. And then two, just maybe a recap of the framework for people that are listening, you're like, "Oh my god, I have so much. I wrote all these notes." Let's give them a summary so they could remember to use it.

**Lenny Rachitsky** (01:27:14):
Okay, cool. So let's talk about AI briefly. How has AI tooling helped you evolve this process? How can people use AI to make this easier?

**Chandra Janakiraman** (01:27:24):
So the first disclaimer is I'm not sort of an AI futurist, but I read all the emerging stuff that's coming out, most of the emerging stuff that's coming out and use some of the tools. And it's fascinating what's happening. And I think it will definitely have an impact on strategy formulation. So what I'll share is what resonates for me in the context of strategy formulation with AI.

**Chandra Janakiraman** (01:27:52):
So right away, I think the basic idea is everybody should be using assistance in the strategy formulation process with the basic tools that we have. And there are two ways to get AI to assist you in the strategy formulation process. The first is to support the preparation phase in terms of research. And this could be competitive analysis and, for example, you could do trend analysis from a vast library of competitors' release notes. And you can sort of say, "Okay, what are the themes of investment of a competitor's release notes?" Or you could do a reviews analysis of a competitor product and sort of understand what's resonating for users, what's not. And you could also ask a tool like ChatGPT to do a head-to-head comparison between a few players on a certain dimension and really give you a heat map of how they all stack up. And you could also ask an open-ended question like, "Hey, why is this new product so successful? Why is it getting so many users?" And there's some good hypothesis that you usually get. So really leverage in sort of the preparation phase from a competitive analysis standpoint.

**Chandra Janakiraman** (01:29:12):
The second one is in this idea called generating mock strategies. And I know Claire Rowe at your summit spoke about this, and I think this is absolutely right and should be a critical input into the strategy process, which is asking these tools for a mock strategy. And I call this a mock strategy because it's kind of almost the answer but not quite the answer. So what I've found is that these mock strategies, like let's say I support VRChat now and I ask it like, "Hey, what should VRChat do? How should we grow?" And it generates a mock strategy.

**Chandra Janakiraman** (01:29:49):
What I've found is that it's, one, surprisingly good. It's incredibly well-informed and well-articulated. I've also found that its biggest strength is also somewhat its weakness, which is these mock strategies tend to be pretty comprehensive and extensive, and there's an investment recommendation in a vast number of areas. So basically, if you remember, the core of strategy is really to be very targeted and to be very focused. So these mock strategies become an interesting input and the burden is still on the team to down-select into the most important areas for investment. So forcing that choice still, I think, is a human element and needs that layer of additional judgment, which is very context specific to the company. So this is sort of right away, I think people should be doing this.

**Chandra Janakiraman** (01:30:44):
I think more medium term and probably not too distant future and likely sooner than we all think, there's probably going to be, sort of the model that resonates with me is sort of the multi-agent model, which is you probably have different components of the strategy workflow automated. So you probably have a strategy agent, you probably have a roadmap or feature agent, you have maybe a engineering agent, and these can communicate amongst each other to cycle through results and iterate. And I think Armand Ruiz from IBM has some good definitional frameworks on some of the stuff. He shares it often on LinkedIn.

**Chandra Janakiraman** (01:31:27):
But let's take a simple example. I can easily imagine something like this for a topic like onboarding. So every company and product team obsesses about their onboarding experience. And today, there are advanced experimentation frameworks. So imagine you're expecting a large surge in traffic. There's these sort of experimental frameworks like multi-armed bandits that can really help you get to the optimal sort of variation very quickly in real time. And there are variations of that, like contextual multi-armed bandits, there's combinatorial bandits.

**Chandra Janakiraman** (01:32:04):
But the interesting thing is they still rely on human design of the variations, the different variations that you test. Even though the experimentation framework is very sophisticated, the variations are still human generated. Now imagine if those variations could actually be generated through generative AI and could be plugged into the advanced experimentation frameworks. The possibilities become infinite. And really, you might be surprised by what you find is the winning onboarding experience that you couldn't even have humanly imagined. And it might be different for every user, different for every sort of territory, etc.

**Lenny Rachitsky** (01:32:44):
That is such a cool idea, I just want to say. This agent that's just running, thinking about ways to optimize your onboarding, coming up with concepts that you probably review, like, "Cool, let's try it." And then it does it, ships an experiment to run it and just is constantly optimizing your onboarding. Holy shit, that's an awesome idea. I think everyone will have this.

**Lenny Rachitsky** (01:33:06):
And then what that makes me think there's going to be some company that's built the best onboarding agent. That's what you're going to be paying for.

**Chandra Janakiraman** (01:33:14):
Totally. Totally.

**Lenny Rachitsky** (01:33:14):
Oh my god, that's so good. Okay. Great. Great idea.

**Chandra Janakiraman** (01:33:16):
Yeah, yeah. So then the question becomes, hey, what is our job? Our job becomes architecting bigger and bigger pieces of the product to take advantage of these agents. And there will be a sequential increase in the complexity of workloads that get automated over time. Onboarding is probably relatively on the easier side in terms of complexity of workload. And you get into deeper experiences, eventually AI will get there.

**Chandra Janakiraman** (01:33:44):
So funnily enough, when that happens, some of the manual processes I describe about will seem archaic. But I think the fundamentals will still have a long shelf life, and as you think about these automated frameworks as well.

**Lenny Rachitsky** (01:33:59):
Man, this could be its own podcast. I have this whole post about how AI, how PMs are the best position, role in tech to thrive in a world of AI, which I'll link to that makes people, I find, feel better when they hear you talk about how much might get done through agents in AI.

**Lenny Rachitsky** (01:34:14):
The other quick thought I had, and I want to get to the wrap up, is I have this ongoing debate with a friend about strategy in AI. My feeling is in theory, an AI tool will be incredibly good at coming up with your strategy for you because it's just, here's all the data, here's everything you need to know, how do we win? And to me, that feels like the ultimate way AI is good, is just here's data, here's what I found as a path to a winning. But my friend's also arguing that's the one thing that AI will be least good at because that's where we need people and context and discussion, all these things.

**Lenny Rachitsky** (01:34:53):
I don't know, it's like one or the other. Either AI will be incredibly good at helping you figure out your strategy or the worst.

**Chandra Janakiraman** (01:34:58):
Interesting.

**Lenny Rachitsky** (01:35:00):
I'm curious. I'm curious, I guess, do you have any quick thoughts on that? Where do you side?

**Chandra Janakiraman** (01:35:05):
I think there is a crossover point where the human judgment will be inferior to something that's able to process multiple signals simultaneously. There's an element of lateral thinking as well here, and I'll sort of share a simple example.

**Chandra Janakiraman** (01:35:24):
For those who drive Tesla cars with the fully self-driving capabilities, when you make a turn, humans have this fairly narrow field of vision. So you have to look both ways and you have to make a decision combining both those signals, right? Whereas the car has six cameras, which is simultaneously processing and it can make a decision not in sequence, but in parallel. And it moves with confidence at turnings because of that. So there is this crossover point where some of these, the ability to hold multiple signals in the head, it's going to be more... It's going to be stronger.

**Lenny Rachitsky** (01:36:09):
That makes total sense. What it makes me think about, there's going to be a strategy agent just sitting around always looking for ways to improve your strategy and point you all in a different direction. Oh my god, future is wild, as I've said before.

**Lenny Rachitsky** (01:36:21):
Okay, let's do a quick wrap up of the process for folks that are taking notes and they're like, "Cool, here's the overview." So let's just do that and then let's go to get to the exciting lightning round.

**Chandra Janakiraman** (01:36:31):
I think the quick recap of everything we've covered is that product strategy definitionally sits between mission vision at the top and plan at the bottom. So it sits between those two either at the company level or at the team level. What it does is it forces choice to deploy scarce resources towards maximum impact. And think about the frequency selection and resonance as an analogy, and it ideally includes three components. A handful of areas to focus on, which we call strategic pillars, and several areas that are explicitly not the focus and why. That's really it in terms of what product strategy is.

**Chandra Janakiraman** (01:37:12):
There is a smallest flavor of it which focuses on solving problems. It's what I call present forward, and it typically operates in a two-year horizon. We use a five-stage process to get there, and it takes about eight to 12 weeks. There is a biggest process that focuses on an aspirational future, is future backward and typically has a three, five, 10-year horizon, also a five-stage process, and can be ongoing up to a six-month period to give it enough space to generate something exciting.

**Chandra Janakiraman** (01:37:47):
And the roadmap for the company or the team is built from a combination of smallest and biggest work, which is like building a bridge from both sides of a river. And ultimately, any strategy is only as good as the results it can produce. So test and iterate through execution and double down on what's working and pivot away from what's not.

**Lenny Rachitsky** (01:38:10):
Incredible. I'm glad we did that. Anything else that you want to share or leave listeners with before we get to our very exciting lightning round?

**Chandra Janakiraman** (01:38:18):
Just all the best with strategy work and ultimately individual and business success.

**Lenny Rachitsky** (01:38:26):
Amazing. And when I asked you at the beginning, before we started recording, what your goal was for this conversation, your answer was just to create a ripple of success across companies and teams, and I feel like we've done that.

**Chandra Janakiraman** (01:38:36):
Thank you.

**Lenny Rachitsky** (01:38:37):
Okay. With that, we've reached our very exciting lightning ground. Chandra, are you ready?

**Chandra Janakiraman** (01:38:41):
Yes.

**Lenny Rachitsky** (01:38:43):
Let's do it. First question, what are two or three books that you've recommended most to other people?

**Chandra Janakiraman** (01:38:47):
Yeah, I love books on creativity and innovation, Lenny, and some of my classic favorites are Walt Disney's biography. And there's several out there, but I think the one that I like the most is the Triumph of the American Imagination. And what's interesting is it talks about how Walt Disney actually loved the theme parks more than the movies. And the reason for that was he could actually tinker with the theme parks. He could make changes to where the rides were, he could change rides in and out, and then he could observe real sort of changes to reactions of people flowing into the theme parks, which he couldn't do with the movies because once it was produced, it was done, it was out of his control. And it was interesting, he was basically A-B testing long before the term was coined. And it's just a fascinating look into how his brain worked, way ahead of its time.

**Chandra Janakiraman** (01:39:46):
The other classic is Ed Catmull's Creativity Inc, and it talks about all the negative forces that eat away at creativity in an organization. How do you, as the leader, have to keep them at bay and make sure the team can innovate?

**Chandra Janakiraman** (01:40:01):
And then there's this probably less popular book, but really good one from Tom Kelly of IDEO, which is The Ten Faces of Innovation. It talks about different archetypes that you need on the team that are essential for creating something special. And you actually need, I think there's the devil's advocate, you need the researcher, you need the ethnographer, you need the stage setter. There's all these interesting personalities that you need to actually make something successful. So I would say those are some of the ones that I usually recommend, yeah.

**Lenny Rachitsky** (01:40:37):
That's a really cool... That third book is really interesting, I haven't heard of that. And Creativity Inc, something I want to highlight real quick is to your point, that a lot of it is how to avoid killing good ideas. And my favorite metaphor from that book is the ugly baby metaphor where every new idea is an ugly baby that people just want to get rid of. Get rid of this ugly baby, we don't want this around here. And just every new idea is ugly when it starts, and you need to protect the ugly baby, basically. Although I don't know who's hurting ugly babies, that's not-

**Chandra Janakiraman** (01:41:06):
Yeah, yeah, it feels a bit... Yeah.

**Lenny Rachitsky** (01:41:08):
Feels aggressive.

**Chandra Janakiraman** (01:41:08):
Harsh.

**Lenny Rachitsky** (01:41:11):
Yeah, harsh. Yeah. But it's memorable because I've never forgotten that.

**Lenny Rachitsky** (01:41:14):
Okay, next question. Is there a favorite recent movie or TV show you really enjoyed?

**Chandra Janakiraman** (01:41:19):
Yeah, yeah. We watch a lot of animated films with the kids. But were a couple of good movies this year. We really liked If, Imaginary Friend, it's about growing up and forgetting these memories of your younger days, is was a really nice movie. And then of course, we enjoyed Dune II as well. It was a fun watch. But most of our time is dominated by the kids sort of animated folks at home or the cinema.

**Lenny Rachitsky** (01:41:53):
There's a new Dune TV show, I don't know if you've been watching it. It just started recently on HBO. It's not the best thing in the world, but it gets you a little Dune fix-

**Chandra Janakiraman** (01:42:01):
Interesting.

**Lenny Rachitsky** (01:42:02):
... in between movies. I think there's a third movie coming out.

**Chandra Janakiraman** (01:42:04):
I see.

**Lenny Rachitsky** (01:42:06):
Third question, do you have a favorite product you recently discovered that you really love?

**Chandra Janakiraman** (01:42:10):
Yeah, yeah. I've been playing this cute little game for the last few days. It's called Capybara Go! It's a tab-based strategy RPG game. It's got really humorous writing and this cute capybara that goes off on an adventure. It's very well-paced and it's a very well-produced game. It's one game I've been playing recently.

**Chandra Janakiraman** (01:42:37):
I've also been poking around Bluesky. Bluesky is interesting. I haven't gotten into a habit yet, but it's super interesting concept of empowering the community with these community-generated feeds. It's very different experience. But it's also interesting, there's a trade-off with the simplicity of the product and what you get with the additional complexity. It'll be interesting to see how it plays out, how it scales.

**Lenny Rachitsky** (01:43:05):
Do you have a favorite life motto that you often come back to, find useful in work or in life?

**Chandra Janakiraman** (01:43:10):
So there was this 1995 interview of Steve Jobs where he talks about this idea of there's a tremendous amount of craftsmanship between a great idea and a great product, and it sort of stuck with me over the years. And the simple way to think about that is it takes a lot of effort to make something special. And conversely, if it's easy, it's probably not that special or not that great. And it's something that I sort of think about when we build product, is there sufficient pain here where it's sort of a proxy for, are we creating something great, is interesting.

**Lenny Rachitsky** (01:43:53):
Wow. It makes me think about founder mode a little bit. Like people talk about founder mode. The reason founder mode I think is important is the people that are most committed and passionate and driven to go through that pain and continue to make it... To take it from just an ugly baby idea to an awesome winning product oftentimes needs to be the founder. And that's why I think it's kind of emerged as a trend is it's important to have that drive, to just-

**Chandra Janakiraman** (01:43:54):
Surely. Yeah.

**Lenny Rachitsky** (01:44:22):
... continue to refine it and not just like, "I had the idea, go build it now."

Okay, final question. You have an amazing background behind you. [inaudible 01:44:31] if people are on YouTube, it's just a beautiful book background, a bunch of objects and books. I'm curious if there's an object on there or a book on there that might be fun to highlight that you're especially into or proud of. And feel free to turn around if you want to look around. Curious if there's one thing that stands out, like, "Oh, here's this thing, [inaudible 01:44:48]."

**Chandra Janakiraman** (01:44:48):
Oh, well, it's the picture of my kids right in the center, which is without doubt the most important and memorable thing there. It's also caught it a unique time when they were incredibly... They were sort of super embracing each other and it's a phase where they've grown out of it. Now there's a lot more, I think rivalry and then sibling tussles. But that moment is my favorite.

**Chandra Janakiraman** (01:45:18):
I also have a bunch of fun stuff, like a big sort of Snoopy fan, and so it is... Also a big Beetles fan, so that picture is actually an interesting blend between Snoopy walking, crossing the road and then the Beetles crossing the other way. And then a bunch of books that are more memorable over the years. You'll see Creativity Inc there, you'll see The Ten Phases of Innovation, all of that.

**Lenny Rachitsky** (01:45:45):
Amazing. Thank you for sharing... Your kids' photo's blocked when you're in the center, so it's like a little Easter egg. Thank you for sharing that. Chandra, this was incredible. It was everything I was hoping it'd be. I think we're going to create that ripple that you were hoping for.

**Lenny Rachitsky** (01:45:57):
Two final questions. Where can folks find you online if they want to reach out? And how can listeners be useful to you?

**Chandra Janakiraman** (01:46:02):
A reasonable choice would be LinkedIn to reach me, and I really, like you and I discussed earlier, I'd be really happy if this creates that kind of ripple effect of successes, both for individuals and for products. And I want everybody to think of this, the stuff I shared as a bit of an open source model. So test some of the concepts, modify it, remix it, and share what worked or did not. And ultimately, that's what makes it all interesting is the community owns it ultimately, and it doesn't belong to one individual. That's what would make me super happy.

**Lenny Rachitsky** (01:46:44):
Amazing. Chandra, thank you so much for being here.

**Chandra Janakiraman** (01:46:48):
Thank you so much for having me, Lenny, and thanks for your service to the product community.

**Lenny Rachitsky** (01:46:52):
Same to you. Bye everyone.

**Lenny Rachitsky** (01:46:57):
Thank you so much for listening. If you found this valuable, you can subscribe to the show on Apple Podcasts, Spotify, or your favorite podcast app. Also, please consider giving us a rating or leaving a review, as that really helps other listeners find the podcast. You can find all past episodes or learn more about the show at LennysPodcast.com. See you in the next episode.

---

## [5/15] Linear’s secret to building beloved B2B products | Nan Yu (Head of Product)

**Lenny Rachitsky** (00:00:00):
I think you see in the team at Linear that a lot of people don't see, which is that there's not actually a trade-off between speed and quality.

**Nan Yu** (00:00:06):
People talk about this as if there were a trade-off because when they think about speed, the thing they over-index on is rushing or being sloppy. What they should be indexing on is being really competent. If you look at people who are at the pinnacle of their craft, you can basically tell how good the output is going to be of their work product by how fast they're going.

**Lenny Rachitsky** (00:00:26):
What does speed look like when you say it can be done quickly and high quality?

**Nan Yu** (00:00:30):
What it really looks like is you have some rough time budget for how long you think something's going to take. By the time 10% of it has passed, after week one, you have something that works that tests some kind of key hypothesis internally.

**Lenny Rachitsky** (00:00:42):
I imagine a criticism you all get. Over time, you'll probably become a bloated piece of software as well.

**Nan Yu** (00:00:47):
When we examine this problem, we look at, "Well, what feature requests can we debate and what kind of feature requests do we absolutely have to say no to?" The stuff that we absolutely have to say no to is the exact kind of thing that leads to this bloatedness that makes ICs hate their lives.

**Lenny Rachitsky** (00:01:02):
Something that your head of sales shared with me is how impressed he is with the way you ask questions on customer calls and just keep digging and digging until you get to something.

**Nan Yu** (00:01:10):
My goal is to feel bad in the same way that customers feel bad.

**Lenny Rachitsky** (00:01:17):
Today, my guest is Nan Yu. Nan is Head of Product at Linear, which is one of the most beloved, most beautifully designed, and also the fastest growing B2B SaaS product out there today. You rarely see the kind of love that people have for Linear for any enterprise B2B SaaS product. So, there is a lot that we can learn from how Linear operates and how they build product. In my conversation with Nan, he shares a system that he uses for being creative and coming up with non-obvious solutions to customer problems, why it's a red flag to him when PMs tell him there's a trade-off between speed and quality, how he talks to customers in order to figure out the emotion that they want to avoid and then figure out the solution to avoiding that emotion, plus some killer advice on how to land a job, including how he landed his job at Linear and his previous role at Mode, and so much more.

**Nan Yu** (00:04:59):
Thanks for having me. I'm a long-time listener and reader, so it's really a treat to be here.

**Lenny Rachitsky** (00:05:05):
I want to share something with you to kick off that I haven't shared with you yet, that I haven't shared with anyone. These results might have come out by the time this podcast comes out, but I'm running a survey right now that I'm calling, "What's in your stack?" Where all my subscribers are asked, "What tools do you use most day to day? What tools do you love most? What tools do you hate?" And one of the questions asked was, what tool do you wish you could switch to if your IT department allowed you to? The number one answer by far is people want to switch from Jira to Linear.

**Nan Yu** (00:05:38):
Wow. I mean, hopefully, that means we're doing a good job.

**Lenny Rachitsky** (00:05:41):
I think that's exactly what that means. I'll read a couple quotes to give you a sense of what people are saying about Linear. I doubt these are surprising to you, but this gives people a sense of why you're here and why I'm excited to extract as much wisdom as I can from you. So, a couple quotes here. "Linear is a joy to use as I interact with my engineering teams, and I find inspiration in its design." "Linear is simple to use, yet powerful." "Linear's design is obviously an industry benchmark, but moreover, the performance and speed is a massive productivity boost."

**Nan Yu** (00:06:12):
I mean, it's really good to hear that because in a lot of ways, that's what we're trying to do. If you think about the entire impetus behind why Linear was started, it's because Karri was sitting at Coinbase and Airbnb and these places and just watching everyone around him struggle using the tools that they had available and always incumbent tools and just seeing that it made people hate their day-to-day a little bit, and we all got into technology and design and engineering, all this kind of stuff because it was fun. All of us started off building stupid MySpace pages and all of these side projects when we were young, and it started off as this fun thing that we do, and we're like, "Wow, we get to do this for a career," and then to have all of this kind of stuff put these big speed bumps into our day-to-day workflow, it just was really sad. So, that's why we started Linear. This really bust through all of that.

**Lenny Rachitsky** (00:07:11):
What I love about Linear, I feel like it's an inspirational business because many people want to, "I'm going to build just a much better version of something," and often that doesn't actually work out. Often nobody cares enough. There's all these barriers and reasons. People don't switch to something that's better, and Linear is an amazing example of building an excellent product and actually succeeding, and there's a lot more to it maybe than just building an awesome product. So, that's what I'm excited to dig into and understand how you all operate, and I guess just based on these results, to me, this is the ultimate sign of product market fit. People being sad they can't use a product in B2B enterprise software especially, so let's get into it.

**Lenny Rachitsky** (00:07:52):
First question I want to get into is something that I think you see and the team at Linear sees that a lot of people don't see, which is that there's not actually a trade-off between speed and quality. I think a lot of people think this is just an innate fact and something I've heard you talk about is that's not actually true. I actually saw Patrick Collison tweet this exact point that I'll read after you... I want to hear your thoughts, but talk about what you've learned about how there's maybe not actually this trade-off between speed and quality.

**Nan Yu** (00:08:20):
People talk about this as if there were a trade-off almost in a naive way because when they think about speed, the thing they over index on is rushing or being sloppy, and what they should be indexing on is being really competent or being like an expert. So, if you look at people who are at the pinnacle of their craft, it could be anything. It could be like a chef or a programmer or someone building houses or something. You can basically tell how good the output is going to be of their work product by how fast they're going. If they're going really fast, and they're obviously not being sloppy and then leaving a mess all over the place, it's like, "Yeah. Well, they got there because this is just second nature to them," and they're able to go at a really rapid pace and try stuff. And when we're building software, that's such a big component of how good the product is on the other side of it, which is like, "How many iterations were you able to do?" So, the only way you're going to get a bunch of iterations done and try different things and really feel out these different variations is by just going very fast.

**Lenny Rachitsky** (00:09:25):
In terms of speed, is the speed there moving quickly on each of iterations? Like what does speed look like when you say, "It can be done quickly and high quality"? What does speed look like?

**Nan Yu** (00:09:36):
Speed... What it really looks like is you have some rough time budget for how long you think something's going to take, and by the time 10% of it has passed, you have a workable solution. It's not like, "Oh, at the halfway point, we have something that is maybe a candidate that we can play around with." It's like, no, no, no. After week one you have something that works that tests some kind of key hypothesis internally so that you can feel like is this thing actually panning out the way we expect it to or did we have some crazy incorrect assumption? And you don't want to wait until you're 80% done to be able to make that kind of judgment because then it's just too late. Then you're pushing deadlines out, and you're making your marketing team very sad.

**Lenny Rachitsky** (00:10:18):
Amazing. Okay, so the way you think is, "We're going to spend a month on this feature. Let's get something workable. We can start testing with potential users even internally in the first few days, essentially in the first week"?

**Nan Yu** (00:10:30):
Yes. Yeah.

**Lenny Rachitsky** (00:10:32):
Yeah. I guess how can you do that? Because most teams can't do that. Most teams need to research, design, build. "Okay, cool. We have something," and once a month later, what allows you to do that?

**Nan Yu** (00:10:43):
Yeah, I mean, there's a lot of components of it. I think having really good talent really helps. Having engineers who don't get blocked by every single little design choice, they're happy to just make something workable. Even if they don't feel comfortable with that particular solution, they'll just bust through it and make something happen there. Part of it is intent. We don't have any expectation that the first version of it is going to be great. That is not in the cards. Look, the first version of it is our best guess in the general direction of what we want to actually ship in the end, and sometimes it works out. Sometimes, it's like, "Wow, this first version was pretty good. Let's make some minor adjustments, and we're good to go," but there's no expectation there. So, no one feels like they have to be a perfectionist and get everything, like all sanded down and really in tip-top shape. It just has to work and get the job done and validate or invalidate our major assumptions.

**Lenny Rachitsky** (00:11:38):
I'll read this quote from Patrick Collison. He tweeted this today as I was preparing for this interview, and he's the CEO and founder of Stripe, if you're not familiar. His tweet is, "I increasingly believe that 'good, cheap, fast -- choose two' maxim is devious misinformation spread by the slow. In my experience, slow and expensive usually go together."

**Nan Yu** (00:11:57):
Yeah, exactly. I mean, use the contractor kind of example. Like If someone's making modifications to their house, and it's taking forever, one, you're in a hotel and also the bills are adding up.

**Lenny Rachitsky** (00:12:09):
The other example you used when we were chatting about this earlier is chess players. I'm thinking of Magnus Carlsen, watching him. I think he was number one in speed chess in addition to just regular chess and what a microcosm of this point.

**Nan Yu** (00:12:22):
Yeah, I think that's the case and Magnus and Hikaru and all those guys who are at the top of their game, they can go unbelievably fast. In fact, that's the usual... I mean, I don't want to get too out of my depth with chess, but the usual way you try to make the game fair is you give them much, much less time than someone who's not quite as strong of a player, and they'll still win a lot of time, too.

**Lenny Rachitsky** (00:12:43):
So, maybe just to close out this point and give someone something concrete they can do with this information, say they want to start moving faster while not cutting quality, what do you think they can do? What's one thing they can start trying to work on and improving in the way they operate?

**Nan Yu** (00:12:58):
I think it's really that sort of attitude and point of view question to understand and take the almost controlled risk that the first version of this is not going to be perfect. So, it actually makes it a lot cheaper in many ways. It means you don't need a pixel perfect design. It means you don't need to make sure that all of the little UI bugs and stuff like that are solved because none of that really matters. What matters is you have working software that you can interact with, and you can see if it feels good. Does it actually solve the core problem that is facing our users? You can take it back to users. You can even let them into an early beta or something like that and get real validation there and to really focus on getting the smallest, shippable element, like not shippable in the sense of, "I can actually put on the production," but in the sense of like, "I can start learning from here."

**Lenny Rachitsky** (00:13:50):
Just a question I imagine is in everyone's mind is what do you do with this first very ugly V1... not ugly, not fully ready, first version. Is this something you're using internally to see if it's something? Is it something you have beta design partners with?

**Nan Yu** (00:14:04):
We have a gradually increasing sort of circle of users that use every single feature. So, by the time it hits GA, by the time it gets released, it's been used by a lot of different users up to that point. So, the first circle is just internal users. We use Linear every single day to write software and do our own work, so we have that kind of advantage and then once we feel like it's good enough, we'll put it into some beta customer group, and again, as early as we can in the process. We have to make sure that we don't end up corrupting people's data, and it doesn't look hideous and that kind of stuff, but as long as it reaches that level of quality, we can release it to early access customers who can give us good feedback and also just try to solve their problems with it. If no one engages with it, if no one's using it, then that's a good signal that we didn't really hit the mark, and then we have a couple of different beta audiences that we grow and then the ultimate release obviously is for GA where everyone gets it.

**Lenny Rachitsky** (00:14:59):
That's an amazing answer. Okay, so secret number one to Linear success, I'm going to take some notes here, is get new feature, product ideas out to people as early as possible, say, in the first 10% of the amount of time you've allotted, and then release it increasingly to more and more people to get feedback. I think the implication here is just most wasted time is on building things nobody actually ends up wanting or using. So, the sooner you at least get directional sense of are you heading in a good direction, the faster it all go?

**Nan Yu** (00:15:30):
Yeah, totally.

**Lenny Rachitsky** (00:15:31):
I imagine a criticism you all get. People are like, "Yes, Linear is so great, so beautiful, so much better than what's been out there for decades," but over time you'll probably become a bloated piece of software as well. That's just the fate of enterprise software. You have to check all these checkboxes. IT teams need all these features. So, there's always this like, "Oh, yeah, sure, you guys can operate this way for now. You have an amazing product for now, but it'll get ugly and bloated." How do you think about avoiding that? I know it's something you spent a lot of time thinking about. Maybe give us a glimpse into some of the conversations you have internally when there's these feature requests like, "Oh, I need single sign-on with this thing and this button here." How do you think about what to add, what not to add, and how to add these features to not make it bloated?

**Nan Yu** (00:16:14):
This question actually comes to us a lot from candidates that are interviewing with us. When you go like, "Hey, do you have any questions for us?" This is the question that we're going to get. So, we hear it quite a lot, and it's very sensible for them to ask it because they see history being littered with the corpses of startups trying to compete in this space and not making it, and I think when we examine this problem, we look at, "Well, what kind of feature requests can we debate and what kind of feature requests do we absolutely have to say no to?" And the stuff that we absolutely have to say no to is also the exact kind of thing that leads to this bloatedness that makes ICs hate their lives, and it's very specific. It's customization features requested by middle managers in order to make reporting a little bit easier at the cost of making IC workflows worse.

**Nan Yu** (00:17:16):
It's like if it fits that description, we're just saying, "No." There's no debate because we've already thought about it and this is the thing that we can't take a single step down this path. So, I think that's honestly one of the core promises of Linear is that we will not make this particular trade-off. So, when you see people saying like, "Wow, Linear is so much faster. It's so much easier to use and it makes my work so much more enjoyable." This is the reason because we have not taken a single step in this direction. It's very easy for a PM to say yes to this kind of request because often they're talking with buyers. Any kind of B2B type of space, they're talking with whoever the gatekeeper is and sales is putting pressure on them, and they're saying like, "Hey, we really want this one feature. It's going to make our reporting nicer." 

**Nan Yu** (00:18:02):
So, the director's going to be really excited by this, and we'll definitely make a buying decision based off of this, and we have to convince them that this is a false trade-off. The whole premise is wrong because the moment you start going down this path, and you make the IC user experience worse, they're just going to disengage. No one has to do this. If I'm an engineer, I get paid to write code. My performance review is based on my code contribution. It's not based on like, "Did I fill in all the tickets right?" So, I'm just not going to do that part, or I'm going to do it very sporadically, and then I'm going to just focus on my actual job. 

**Nan Yu** (00:18:38):
And then all your reporting is wrong because all the data is wrong, and it's sparse, and you get situations where people will... They'll say like, "Well, here's a dropdown field that someone put in here that's required." There's nine choices. I don't know what any of them meet, so I'm just going to pick one at random. I'm still going to pick the first one. Also, I'm going to pray that my boss is not actually using this data to do any kind of reporting and that has consequence because the data can't possibly be correct. So, I think for us, it's a very easy decision when it comes to that particular category of feature request.

**Lenny Rachitsky** (00:19:12):
I love how simple and clear that is. Basically, you all have a policy. We'll prioritize ICs over middle managers. Especially, like I love that it's around reporting. Almost always it sounds like, "I want to track what's happening."

**Nan Yu** (00:19:23):
Yeah, exactly. It's always, "I want to track what's happening." Well, what do you want to track? Well, I want to track which version of the product this thing's tied to based on some field information. It's like, okay, how is the person working on this supposed to even know that information? Well, it takes like a five-minute scavenger hunt every single time. It's like, "I don't think they're going to do that, man."

**Lenny Rachitsky** (00:19:43):
What I imagine happens, and I think why this is hard for most companies is there's an implication that you're turning down deals. You're not adding that one feature that will close a massive million-dollar sale, very difficult to do. I imagine it helps a lot that... I imagine the COO is very bought into this and there's this, "We will win long-term holding the line on this." Is that right?

**Nan Yu** (00:20:05):
So, it is, but I also think that there's not as much pressure as you would expect to do these kinds of things. There are basic scaling things, like we had to make SAML and SCIM and that kind of stuff. It's like, "Yeah, sure, we're going to do those sorts of, like keep the lights on type of work," but when it comes to work that's related to the actual business logic of the app's value proposition, what buyers care about is, is this going to make their team more effective? That's the reason that they're making this buying decision in the first place is that they're like, "Well, the current situation we're in... " And especially with large companies, right? The current situation we're in is a mess, and if we can convince them that these types of things are actually the reason that it's a mess, then we can really navigate them out of wanting them in the first place.

**Lenny Rachitsky** (00:20:57):
Got it. So, there's an element of you think you need this, but it turns out you'll be more successful and get everything you want, not getting this?

**Nan Yu** (00:21:04):
Yeah, and the thing is, it's not everything you want, right? Because people come with a laundry list, and it's like laundry list. Here's 10 things I want. You're like, "Do you want all of those 10 things equally?" They're like, "No, actually I don't." The first three are the things that really matter to us. If we solve the first three, then the other stuff, we can negotiate on. So, our job is to solve the first three-way better than anybody else that if they got through the first three through some kind of visual programming, customization type of thing, that it's never going to get to the quality level and the depth that we're able to offer by offering those as native features.

**Lenny Rachitsky** (00:21:37):
It's interesting thinking back to that survey I shared where the tool people want to switch to if IT allowed them was Linear, and on the one hand you could argue, "Well, okay, IT is not letting them use Linear for all these reasons.  On the other hand, you guys are growing really quickly within enterprise, like you're a new business. You started, I think, mid-market startups, and now you're working way up. So, I think it's not fair to say it's not going to work in enterprise. It's clearly working really well. I don't know if there's any stats you can share anything of that, but it seems to be going well, expanding up market.

**Nan Yu** (00:22:11):
Yeah, I mean, growth has been good. Growth in enterprise has been leading the other segments because I think this year, especially we reached a tipping point where I think with software, so much of the buying decision is based on almost like a brand thing, like is this for us? A lot of times people pick "enterprise software." It's like, "Why? You know everyone doesn't want this," and they're like, "Yeah, but it's for us." 

**Lenny Rachitsky** (00:22:36):
You won't get fired for buying Microsoft or whatever.

**Nan Yu** (00:22:39):
Yeah, exactly, and I think that we're starting to have enough brand penetration amongst enterprises where people can have that feeling, right? They're like, "Hey, Linear is for us. Who are we? Well, we are a large company that wants to act like a startup." It's like, "Who doesn't want that? Who doesn't want to go fast?"

**Lenny Rachitsky** (00:22:58):
Yeah. I had Jeffrey Moore on the podcast, and this is exactly what crossing the chasm looks like. He talked about basically you need someone that's across the chasm like a later adopter that isn't the person that's, "I love new stuff, and I'm an early adopter kind of evangelist." You need someone that's like traditional old school, takes their time to start to adopt it for you to be like, "Oh, okay. Now, maybe I should really take it seriously."

**Nan Yu** (00:23:21):
I also think that with this particular category of tool, and with a lot of other B2B software, not... Like no means not now, right? Not right now because it doesn't fit our budget. It doesn't fit our change management situation. "Oh, we have this exec that's really wedded to this other tool," but those things change, right? So, we keep in contact with them. They're in our CRM where we make sure we follow up, and we've had a lot of these where we've been said no to, like two years ago, and now we have some new features, and then go like, "Oh, yeah, it seems like you're ready for our scale," or whatever.

**Lenny Rachitsky** (00:23:59):
You mentioned that when you have these debates and questions that come out, you have features that a big company wants. There's this category of, "We know we will not build things for middle managers that want reporting and custom stuff just to track what's happening," versus something an IC wants to be more productive and successful, Linear. Give us a little sense of some of the more complicated debates that aren't necessarily in that bucket.

**Nan Yu** (00:24:22):
I think the complicated debates are often when we do add a new native feature, do we extend an existing feature and make it more powerful or do we add a new sort of service? And a big part of that is trying to figure out exactly who's going to use it, what are the actual real life use cases that we know about? Like that I know that Bob from Company X has this workflow and this is how it would work for him. Here are the different variations where it would work. So, tying it all the way back to real people is-

**Lenny Rachitsky** (00:24:52):
Like a specific person?

**Nan Yu** (00:24:53):
Yeah, specific person. Yeah. Yeah, exactly. Not a hypothetical person. Not one that you made up like Alice, Bob, or whatever. It's like, "No, here's the first name, last name. Here's their email. You can ask them," and I think that being able to tie it all the way back to reality in that way is a big part of how we really think about and discuss these things.

**Lenny Rachitsky** (00:25:13):
This connects the way I think about my newsletter is I always try to answer the question a very specific, like a person actually asked, not a general sense of something people may be interested in, and that very specific question, like it implies there's a need. Like not implies, it proves there's at least one person who needs this thing versus you have this idea of somebody that may want this thing. 

**Nan Yu** (00:25:36):
Yeah. I think a trap that a lot of times PMs will fall into is they'll make something, and they'll make some choices in it because maybe it's beautiful or it's elegant, but they don't go the step of like, "Is reality also beautiful and elegant?" Because reality is ugly sometimes, and if you have a beautiful and elegant solution that doesn't match with reality, it doesn't really matter. People can look at it, and they can ooh and ah, but if they don't use it to get their work done, it's never going to have long-term staying power.

**Lenny Rachitsky** (00:26:01):
Do you have a heuristic of how often you need to hear something for you to... could be just convinced, this is worth investing in? People may hear this, "Oh, one Bob. Bob wants this featured." That doesn't make sense. It's just one guy. How do you know when it's like, "Okay, we should really invest in this"?

**Nan Yu** (00:26:17):
Part of it is you hear something, and you're like, "Gosh, that actually is... " Not only is that true. It means that the way we thought about this was a little bit wrong, and I call this process... I don't know if it's the right way to describe it. I call it a kneeling where you have a thing, and it's not quite the right shape, and you put it out into the wild. So, this happens way in the first bit of the life of a particular feature. You release a thing, and then you start getting feedback about it, about hey, it doesn't quite fit reality, and then you ask yourself like, "Did we test that aspect of it? Did we actually match that part to reality?" And if you didn't, then it's like that's the part where you don't actually need that many pieces of feedback against it. It's not really a volume thing. It's like, "Did we think about this right or wrong?" That's one sort of category.

**Nan Yu** (00:27:01):
Another category is just you're getting a request for maybe a very big feature or a feature set from a lot of different people, but then you dig in, and you try to say like, "Okay. Well, tell me about how you're trying to use this," and there's 100 different use cases. So, you have choices here. You can either build the big feature that covers all the long tail of use cases or you can try to see if there's really concentrated pools of use cases for this that really make a lot of sense to adopt as a first order type of feature. So, I think those are the two sort of strategies that we employ the most. It's like, "Did we think about this wrong? And now we're just learning something about how it matches reality or for this big general feature that people are asking for, are there actually more specific use cases that we should be solving, and we should be solving really, really well?"

**Lenny Rachitsky** (00:27:52):
A thread that's coming through so far across a lot of these examples is getting to the specific person using the thing and making them happy and making sure the ask is going to solve their actual problem. In the case of looking at the IC versus the middle manager, in this case, it's like, "Let's talk to the person actually asking for this thing," not, "There's like 100 people generally asking for this thing and let's build what we think is a general solution."

**Nan Yu** (00:28:18):
Yeah. I'll give you an example of all of these things, which we just launched a feature called Customer Requests, and basically what this does, it adds a new concept of Linear, which is a customer. For B2B companies, this is very relevant, and the reason we did this is because we kept getting this request for fully customized fields, and we would be like, "Well, what is it that you want with your custom fields?" Because the problem is you add 100 custom fields and all your ICs start hating it. So, we don't want to go down that path, but what is it actually you're trying to do? And 40% of them were because, "Well, I have a customer," like Walmart or whatever, right? Like, "Walmart asked for this feature, and it's really important. I need everyone to know that Walmart needs this. I need to track it. I need to see how have we report... " 

**Nan Yu** (00:29:09):
We can report on what have we done for Walmart over the past year so that when my CSM has a one-on-one conversation with a rep, they can have some kind of evidence that we've been doing stuff for them, like all this kind of stuff. We're like, "Okay. Cool." That sounds like a very useful and powerful thing you want to do. How do you expect people to tag these things? Well, manually, because that's how we did it in our spreadsheets. It's like, "Okay, instead of that, we're going to hook up with your customer support tools. We're going to hook up with your CRNs. We're going to automatically bring in feedback from these companies. We're going to analyze the emails where they're from, and then if someone requests a feature that gets escalated into engineering, it'll just be tagged with whoever asked for it. You don't have to do anything, but you will know, and you can still report on this stuff, but there's nothing about this that makes ICs lives harder.

**Nan Yu** (00:29:54):
In fact, it makes them feel more confident because when they're building the thing, they actually understand who's asking for it and exactly what the email said. So, when they're doing the design or the details, they can actually see the real-life use cases that are present and solve for those directly.

**Lenny Rachitsky** (00:30:09):
As I'm hearing this, it's like, "Okay, obviously, this seems like an obvious solution. Of course, 40% of people telling me they have customers." In reality, most of the time, if you hear from a bunch of your customers, "Hey, I need this custom field," and sometimes you hear one thing, sometimes you hear another. Most of the time you're going to build this custom field. Something that your head of sales shared with me is how impressed he is with the way you ask questions on customer calls and just keep digging and digging until you get to something that is an insight for you, and then you start to try to solve the problem for them and think about what the product might be, and I think this is such an important and underappreciated skill for PMs. Is there any advice you could share of just how you approach this, how you ask questions, how you think about these customer calls to get to, "Okay, now, I see what we need to build versus let's just build what they're asking for"?

**Nan Yu** (00:30:59):
It's funny because I think from the outside, I'm on these sales calls and then the AE or someone's watching me ask these questions, and I think often they're like, "What are you doing? You're just asking questions from angles that I don't even know what your goal is here," and my goal is to feel bad in the same way that customers feel bad. They come to us with a request, "Hey, we want X," and it's like there's something motivating it and you can do the normal analytical thing and be like, "Ask five whys," and try to figure out like, "Well, what are your goals?" "And as a persona X, I want to achieve this outcome." You can do it that way, but you might miss the reason that they actually feel bad for not having this thing like, "I can't accomplish this goal. So what?" "So, I'm not going to get promoted at work."

**Nan Yu** (00:31:44):
Okay, great. I understand the severity of your problem at this point. What is the actual emotional valence that is motivating whatever you're telling me? And it takes a little while to get there. You can ask people directly like, "How do you feel?" And they're not necessarily going to tell you, but if you have a long enough and deep enough conversation with them, you start to level with them, and you're starting to see stuff from their perspective, and the more you see it from their perspective and the more they know that, the more they're willing to open up to you and tell you like, "Okay, honestly, I had this thing happen where I marked the ship date of this project as December 30th because it's a Q4 project, and I wanted to put it at the very end, and then my marketing team lost their mind because they're like, 'We can't ship something on December 30th. Everyone's on vacation,'" and you're like... And then they're like, "Yeah, this has made me feel really bad." 

**Nan Yu** (00:32:36):
So, I don't ever want to put dates on things ever again. So, like, "Okay, cool. We can help you deal with that. If that's what you're feeling, then I can start building stuff to make sure that you never have to have that bad feeling again."

**Lenny Rachitsky** (00:32:50):
People talk about empathy like, "You need to have empathy as a PM. You need to build empathy the best product leaders, have empathy in this." I think it's such a succinct and powerful way of describing what empathy actually looks like as a product leader, which is I want to feel as bad as they feel in hearing the story they tell, and it sounds like the way you do that is you keep asking questions to understand the moment they felt bad about something. In this case, the deadline.

**Nan Yu** (00:33:17):
And if you ask somebody in that last story, like what kind of issue do you have? You're like, "Oh, marketing and I would just never align on anything." It's like that doesn't really tell you what's going on. What it tells you is you had this terrible moment of communication that it's all miscommunicated, and you're like, "It's just going to keep happening over and over again." So, the thing that we did specifically to solve this was on projects in Linear, you can just specify a target date at whatever level of granularity you want. You can say it's a December project. You can say it's a Q4 project. You can say it's a second half of 2024 project. Like whatever you're happy promising, you can just put it on there and that way you never feel like you have to give this sense of false precision so that it ends up with a whole bunch of miscommunication down the line.

**Lenny Rachitsky** (00:34:04):
I could see why people love Linear is it just makes them feel less bad less often. There's a lot of connection here. I know this idea of emotions and feeling bad is a core part of how you think about building product, looking for moments. People feel bad. Is there anything more you could share there to share how you think about this idea of emotional hooks, emotional moments, and how you decide what to build?

**Nan Yu** (00:34:27):
So, to set the background of this, I've worked in very, very competitive industries. I worked at Everlane, which was a direct-to-consumer clothing brand. I worked in Mode, which is like BI tools and there's so many BI tools out there, and then obviously, Linear. We're project management. There's a lot of project management tools, and I think the more competitive your industry is, the more the low-hanging goal-oriented stuff is already picked because every PM from every one of these companies has been asking like, "Well, what's your goal? What is your job to be done," and all this kind of stuff. So, you have to look at things from an angle that other people might not have seen and for me, and for us, it's the angle of where are the emotional hooks that you're experiencing as you go through your work day, as you use our product, as you use competitors' products?

**Nan Yu** (00:35:21):
I think it's probably underexplored because... I don't know. I feel like PMs and engineers, we're like very thinky people. We avoid the touchy-feely stuff. So, I think that's the opportunity. You can see where are you feeling bad throughout your day where you don't even know? You might think, "I hate Mondays." "Why do you hate Mondays?" "Well, on Mondays, I have to go out and gather a whole bunch of stuff to write this report that it's really annoying." "Oh, so if I gave you a button that made the report, would that help?" It's like, "Oh, yeah, then I might not hate Monday so much." So, I think Paul Graham has a word for this. He calls it schlep blindness, right? It's like I'm schlepping through life, and I'm just completely blind to it, and it's true. You have to have an outsider come in and see what the rhythm of your feelings are throughout the day, throughout the week, and you note the spots where you could really use a lot of improvement.

**Lenny Rachitsky** (00:36:14):
Is there an example? I've shared a couple, but just where you've noticed this in someone using maybe a competitor or even Linear that you solved. I know you gave an example of the dates. I guess is there anything else?

**Nan Yu** (00:36:26):
A big feature that people love about Linear is we have this thing called Triage Management, and what it does is it systemizes this thing where if I put an issue into a different team, if I'm asking them to do something or I'm reporting a bug to them, it sticks in a special zone where it'll notify the right people. They're on a rotation and people will be able to respond to it in an organized manner, and I think this kind of automation, this feature, it came out of two different fields people were having. One, people were trying to implement this stuff by hand, and it was just a lot of touches, and they were doing it, but they felt like, "Oh, I'm totally underwater." "Why are you under water?" "Well, I have to throw all these tickets around and route them correctly and stuff like that," and they didn't see this as an opportunity to have a tool specialize in managing their triage queue. 

**Nan Yu** (00:37:23):
Because they were managing by hand.... They were on top of it, but it just felt really bad because they just had to spend so much attention doing this and then there's the folks who didn't do that. The feeling was just like, "Well, it's totally out of control. People are just throwing tickets over the wall, and I don't know what to do with them. I don't know where they are. They end up in all these holes and then the people on the other side are like, "I throw tickets over the wall. I have no idea what happens to them. I have no expectation that people are ever going to respond to them." So, there's all of these bad feelings that people are having. They all have the same root cause, which is like there wasn't a very automated organized way to deal with your triage queue.

**Lenny Rachitsky** (00:37:54):
Marketers, I know that you love TLDRs. So, let me get right to the point. Wix Studio gives you everything you need to cater to any client at any scale, all in one place. Here's how your workflow could look. Scale content with dynamic pages and reusable assets effortlessly. Fast-track projects with built-in marketing integrations like Meta, CAPI, Zapier, Google Ads, and more. A-B test landing pages in days, not weeks with intuitive design tools. Connect to tracking and analytics tools like Google Analytics and Semrush, and capture key business events without the hassle of manual setup. Manage all your client's social media and communications from a unified dashboard, then create schedule and post content across all their channels. If you're on content-rich sites, Wix Studio's no-code CMS lets you build and manage without touching the design. And when you're ready for more, Wix Studio grows with you. Add your own code, create custom integrations with Wix-made APIs, or leverage robust native business solutions. Drive real client growth with Wix Studio. Go to wixstudio.com.

**Lenny Rachitsky** (00:38:55):
I'm going to try to summarize some of the secrets of Linear's success so far. So, the first is get something out as quickly as possible, say, in the first 10% of the time that you have to build this thing and get it out to internal users and then maybe a growing list of beta users and people that are aware of they're using early stuff. Two is prioritize the IC and the user, basically, versus the buyer or the middle manager that wants reporting and all these custom features. So, it's basically focused on the user, which I think you hear a lot, but I love this very specific example. Three is when you hear asks for features and requests, get to the specific person using the thing, not just general, "Okay, cool. I've heard it 100 times." Find the person that actually needs this thing and understand what's going on, and then four is look for people feeling bad in a moment working in the product. Is there anything else that I'm missing that's important or any nuance you want to add?

**Nan Yu** (00:39:54):
The part where you said, like focus on the user, I think it's maybe a little bit more subtle than that. There's a nuance which is find where the incentives are really misaligned amongst your user base. There's a middle manager that wants really detailed reporting and there's a IC who just really doesn't want to go through all those extra steps, and the incentives for what they want are just very... They're just very misaligned, and you have to find those situations and be pretty judicious about how you make those trade-offs and where you can really find win-win outcomes there.

**Lenny Rachitsky** (00:40:30):
That's a really important nuance. Something else that's come through a couple of times as you've been talking is also something Patrick Collison tweeted once that has stuck with me, which is this idea of having a mental model in your head of the user. So, the way he described it and the way you've described it is oftentimes people are like, "Cool. We're going to figure out what to build. We're going to do a bunch of research, talk to users. That'll inform what we build, and we build it, versus what you've been saying and what he said is you do a bunch of research, look at data, talk to people. That informs your mental model of what the customer needs in their life, and then that informs what you build. So, that anytime you do more research, talk to customers, it's informing your view of the person, and then you're like, "Oh, this was different from what I imagined," or, "Oh wow. This is exactly what we've been thinking and let's build that." Anything along those lines that you might want to share?

**Nan Yu** (00:41:19):
Yeah, I mean, I can tell you a little bit about how we manage our backlog, which I think actually ties directly into this. At any given moment, we have probably 20 or 30 opportunities that we could possibly explore, just product opportunities, like problems to solve, areas to improve for our users, but they're not ready yet. We don't have enough conviction around how we might approach it. So, we just accumulate understanding of this stuff and periodically, we accumulate some more stuff, and then we reevaluate, "Okay, what is our current understanding of how we might best approach this thing?" And I think something that people struggle with is that they might have this model in their head. Like a PM might have this model in their head about how a user behaves, but it's just very hard to share that with someone else. You have to telepathically throw it into their brain, which is hard. So, what we try to do is identify areas that we might attack with a product, but also keep an up-to-date analysis of each of those areas so that everyone can engage with it and also contribute.

**Lenny Rachitsky** (00:42:22):
Is there an example of something that's sitting in your roadmap? I don't know if you could share these sort of things that's just sitting in the backlog of just like, "We're not quite ready to tackle this yet, but here's something we're inkling on."

**Nan Yu** (00:42:31):
Yeah, sure. Capacity planning is a thing that's been sitting in our backlog, and it's something that we see managers struggle with all the time, which is like I have a limited amount of personnel and resources, and I need to deploy them in such a way where we can theoretically accomplish our roadmap, but also we don't get blocked by some bottleneck that we don't end up blocking all of the projects because this one engineer is stuck on some info thing, and that's a thing people struggle with all the time. All the solutions out there are bad. The best solution is a very, very custom spreadsheet that someone would make, and it's a lot of upkeep. So, we have some ideas about how we might automate this, how we might use existing data within Linear to really help out with this problem, but I don't think we've quite cracked it yet. 

**Nan Yu** (00:43:18):
I think there's some nuances that we have to really explore a little bit further. So, we're continuously developing this, and as we hear from hear from users that are struggling with this problem, we will get on a call with them and sit down with them and talk through it.

**Lenny Rachitsky** (00:43:31):
And the idea there is keep informing this mental model, keep informing what this could be until you get to a place of like, "Okay. Cool. I think we figured out what will really solve this problem in an elegant way"?

**Nan Yu** (00:43:42):
Yeah, and I want to really stress a nuance here, which is it's not that we want to solve the entire problem. The entire problem is quite big, but there's something that's really right for Linear to do that would help people have a good starting point for them to reason about it. So, I think a lot of building conviction around stuff is not even like do we have a workable solution? It's like how much of the problem should we actually take on? Because if we take on too much of the problem, then we'll end up overpromising and not being able to deliver on it.

**Lenny Rachitsky** (00:44:13):
I think what's also useful here is you all keep your team very small intentionally and being constrained keeps you from taking on these things too early because you don't have the engineers to build their designers.

**Nan Yu** (00:44:24):
Yeah, that's true. I actually hadn't really put that part together, but I think some of the reason we've done it this way is because we don't have the bandwidth to action everything. So, we have this backlog that we maintain to make sure that when we do take it on, we're pretty set up for success.

**Lenny Rachitsky** (00:44:41):
Yeah, it's interesting. I think a lot of companies are starting to realize that they can build better products and move faster with fewer teams. I want to move in a different direction and talk a bit about how you actually think about building new products. Something that I've heard from you is that you have a systemized way of being creative, which I think is a dream for a lot of people's. It's like how do I be more creative? How do I think of new innovative concepts? You have a really interesting process for how you do this. Can you talk about it?

**Nan Yu** (00:45:09):
Yeah, totally. I think when people talk about being creative, a lot of times what they have a problem with is extrapolating. They can see the stuff that's right in front of them, but what about two or three steps down the line? And then it's just like, "Well, there's just so much possibility. I don't know what direction to go." So, the way that we try to do it is we ask a question which is like, "Okay, how extreme can you take it? You're designing a product. You're trying to come up with a solution. What's the most outrageous version of this along some trait?" I don't know if you guys did this at Airbnb, but I think Brian Chesky talks about like, "What's the 11-star experience?" Is that a thing you guys did?

**Lenny Rachitsky** (00:45:51):
It was a thing he talked about. Yeah, there's always a push of what's the 10X version of some idea.

**Nan Yu** (00:45:57):
When you think in that way, when you're saying like, "Hey, what's the 11-star experience?" What you're really asking is like, "Hey, what's the most luxurious version of this hotel stay? Or what's the most unforgettable kind of experience we can give people?" And you throw away things, I don't know, like cost. You throw away things like practicality because that's not what's interesting. What's interesting is I want to actually explore the possibility space, and I think this is really important to do because the goal is to get you to see beyond your defaults. We have all of these constraints that we're operating under that we psychically have in the back of our heads that we just don't even realize we have them. So, just break past all of them, and then you can really see what your options are because we talk about product decisions. It's like, "Oh, yeah, you have these choices. What are you going to decide?" There's all this decision-making kind of theory.

**Nan Yu** (00:46:52):
But the biggest risk is you didn't see the right choice to begin with. You have these three choices and none of them were right. It's this fourth one that was over in this corner, but you didn't look in that corner, so you never found it. So, I think the whole goal of this is to try to expand the search space of what you're trying to do.

**Lenny Rachitsky** (00:47:09):
So, what you're saying is people often don't think out of the box enough by not thinking too radically enough. So, the choices they're deciding between are just meh options and there's this process of breaking out of that, and I think you could hear this and be like, "Yeah, sure." I could spend 10 minutes being like, "Oh, hey, what's the craziest [inaudible 00:47:35]- "

**Nan Yu** (00:47:34):
Yeah.

**Lenny Rachitsky** (00:47:35):
But you're saying that actually is what you do and that actually works really well?

**Nan Yu** (00:47:39):
Yeah, and you actually build it. You can think of a very extreme version of a product and you can say, "Hey, let's actually... " For the first version, we talked about, like the first version, you know it's not really the right answer. Sometimes, you know it's so hard because you know this is the most extreme version of the answer. So, let's build that as fast as we can and see how it feels, and then we're going to learn so much about what the right actual answer is because we have seen this area of the product space and really felt it.

**Lenny Rachitsky** (00:48:05):
Awesome. Let's talk about an example of this because this feels awesome.

**Nan Yu** (00:48:09):
Yeah, I can talk to an example. Actually, is it okay if I demo something?

**Lenny Rachitsky** (00:48:13):
Absolutely. Let's do it. Show and tell.

**Nan Yu** (00:48:15):
Yeah, let me do that right now.

**Lenny Rachitsky** (00:48:16):
Here we go. We're going to share screen.

**Nan Yu** (00:48:18):
All right. So, this is just like a demo space instead of Linear. So, the feature where we did this that I remember very clearly, because it was recent, is we built this feature to save drafts for your issues. So, Linear, as hard as an issue tracker, if I make a new issue and let's say I'm trying to report a bug or something, so it's like I make a bug report, then I might start thinking through like, "Okay, what are the repro steps?" And then I start typing them, and this happens all the time. When you're at work, you're doing this and someone distracts you. If someone pings you on Slack or you have to go to a meeting or something like that, you're like, "I got to put this away for a second. I'll come back to it later." Note to self, figure out the actual repro steps and do it.

**Nan Yu** (00:48:56):
So, what can you do? Well, you want to save it as a draft. So, we're like, "Okay, this is the problem," and the first version of this, we're like, "What do we want to do? Linear is about being fast." So, we don't want to get in your way. We want to say like, "What is the fastest draft saving experience possible?" So, if you save it as draft, you can save it as draft. If you decide to not... you want to throw it away, you don't want it, just hit the X button, and it'll just throw it away. We're not going to interrupt you with a popup that says like, "Do you want to save your changes," or any of that kind of stuff. We'll just absolutely get out of your way fast as possible. So, we're like, "What's the risk here?" Well, it might feel really unsafe.

**Nan Yu** (00:49:31):
If you close this, and we don't ask you if you want to save change, you might feel like, "Oh, I just lost my changes on accident." We knew that going in. We built this anyway, and it felt super unsafe. It turns out that sort of inkling that we had was true, and we really felt exactly how unsafe it was. So, then we were like, "Okay, well, what's the safest thing we could possibly do?" The safest thing is just auto save everything. So, you start a new issue, and then you start typing some stuff, and it's just like auto saving as soon as you type a single character and that did feel quite safe. So, cool, but it also ended up leaving behind a whole bunch of like a paper trail of things you change your mind about. You've probably had this happen in document tools where you have a whole bunch of things in your space called like Untitled Document or New Document and stuff like that. It's just like-

**Lenny Rachitsky** (00:50:24):
So many untitled folders.

**Nan Yu** (00:50:25):
Yeah, so many untitled folders because the moment you say new folder, it starts saving it, and then you don't actually mean for that to happen. So, we had those two sorts of variations that we built, and we fell through and where we ended up was a balance between those two. So, what happens is if I'm creating a new issue, like I am here, and I close it out, it'll interrupt me, like we have to interrupt you, otherwise it feels too unsafe. So, I can save the draft, I can go to my drafts, and then if I'm in this draft I've already made, and I go in there, and I start to say, "Okay, I'm going to keep working on it," but then I get interrupted again, then I'm just going to auto-save it for you. There's no point. I'm not going to ask you again.

**Nan Yu** (00:51:06):
I'm always going to auto save it because I'm not going to create a new object. I'm just making modifications in place. So, we made this very specific choice of on a brand new issue, we will interrupt you, and then on an existing draft that you're messing around with, we're just going to auto save everything and someone doing a analysis. If they did a detailed teardown of these decisions, they might say like, "Wow, they made very specific choices here," but the path to get there is to do something totally extreme in one direction and then totally extreme in another direction and then find where they really meet up.

**Lenny Rachitsky** (00:51:39):
Such a good example, the way that you described it is you went like here's the safest route. Here's the fastest version. Where did you come up with these list of options? And for folks that are trying to do this for their company, are these like... Because these are Linear principles, we're going to be very fast. Is this the way you think most companies should operate these sorts of attributes? Do you think it's specific to what makes their product different? How do you think about that?

**Nan Yu** (00:52:04):
I think for a lot of companies, you have to ask, "What is the promise that your product or your business is making people?" It might be you always have a car available if you need it, and if you do that, then maybe we're going to have to implement search pricing to make that happen. It's always going to be available. So, here's the trade-off that we have to make. It's a very extreme point of view to do that. Or you might say the price is always predictable, but sometimes you can't have a car in the first place. Those are all choices that you get to make, and you have to sort decide, like where in that spectrum does it make sense based on the promise of your company?

**Lenny Rachitsky** (00:52:40):
A lot of people talk about this idea of working backwards. Brian Chesky in Airbnb has a big concept of working backwards from the ideal. Let's design the best possible scenario and work backwards. I love that this is even more tactical, which is just pick the extreme version of very specific attributes. Probably not that ideal, but it'll give us insight into a version of the ideal and an element that works well and then what doesn't. Yeah, exactly. I did this a lot actually at Airbnb, just like testing the extreme. So, it super resonates, this idea, and when you say test, so was it like you build it and play with it? Do you roll it out to some of these circles of users or is it often just internal, and then you learn and then iterate?

**Nan Yu** (00:53:23):
Yeah, we rolled out some of these versions to people.

**Lenny Rachitsky** (00:53:25):
Oh, wow. Okay.

**Nan Yu** (00:53:27):
So, the super-fast version that was unsafe, that only went interna, and everyone felt it was too unsafe, but then we thought, "Okay, let's go to the super-safe version," and then we rolled that out and everyone started having a whole bunch of... Like how many drafts are people making? I'm like, "This is too many." The people are leaving behind this crazy paper trail. Okay, we got to figure out some difference here.

**Lenny Rachitsky** (00:53:46):
Awesome. So, this very much connects to your first point of get things out really quick, and in this case, it's like extreme versions. You're probably not going to work long term, but it will teach you.

**Nan Yu** (00:53:56):
Yeah, exactly.

**Lenny Rachitsky** (00:53:58):
Amazing. Okay, and seeing it in action, I'm like, "Okay, obviously, this is the solution," and that's how the way this should feel, and to your point, it was not an obvious solution when you started thinking about it.

**Nan Yu** (00:54:08):
Yeah. I mean, the best solutions are always obvious in hindsight, and it's just like you have to develop a process internally that to eventually find your way there.

**Lenny Rachitsky** (00:54:16):
Something else that you've mentioned when we were chatting that connects to some of the things we've been talking about is you have this perspective that B2B software isn't just solving people's problems, it's also teaching them how to work, and it's this accumulation of information. Can you talk about that? Because I thought that was really fascinating.

**Nan Yu** (00:54:38):
If you think about how a lot of B2B software gets created, it's because there was some person in the middle of some giant company who implemented some kind of process, and they're like, "Wow, this process is really working for us. Maybe we should make it easier," and they build a little tool internally and then all of their colleagues can now press on buttons and good things happen, and then they turn that process and that tool. They spin it off into a startup, and they make a startup. This process repeats thousands of times. So, when you adopt that tool, you're not just adopting the actual software, you're adopting the idea that this is a practice that you ought to be doing in the first place. So, if you're a marketing person, and you adopt some marketing software, you're not just saying, "Okay, now, I can write emails and send them to people."

**Nan Yu** (00:55:24):
There's all sorts of process around that. You're organizing stuff into campaigns. You're measuring click-through rates. You're calculating cost of acquisition and all that stuff probably comes equipped with a tool because those are the right practices to do when you're doing this sort of marketing exercise. And whether you knew about it before or you learned it from the tool, like as a buyer for this kind of product, what I'm doing is I'm saying like, "Hey, I'm going to bring in this baseline level of marketing competency into my organization, that this is the worst we can do is whatever the tool defaults are."

**Lenny Rachitsky** (00:55:58):
Interesting. So, you're basically buying into a way of working when you're adopting a piece of software, not just have this problem I need solved.

**Nan Yu** (00:56:06):
Yeah, exactly, and I think the most salient example of this is if you've ever seen like a company adopt an ERP product, it's the most painful thing you can imagine. It's doing deep surgery. They have to redo all of their internal processes and the way they manage inventory and all this kind of stuff, but they're willing to do it because they know that this is a battle-tested way of making sure that you're actually doing good management of resources. So, they're like, "We're growing up now. It's time for us to adopt these best practices. In order to do that, we have to adopt this tool, and we will conform to whatever the tool is best is to do."

**Lenny Rachitsky** (00:56:44):
This connects to a couple things I know about Linear, one is what you've shared of just avoiding these customizations requests from people. Do you have a very opinionated way of here's how you should operate in order to build a great functioning product, org, and company in general? I'm just connecting threads here. One is like we're going to avoid letting people customize too much because we know they'll have a bad time, and then two is just this idea of we are opinionated about the way you should work in Linear, and it's like you have a Linear method, I think it's called, of just like here's how product team should operate based on everything we've seen be successful.

**Nan Yu** (00:57:19):
Yeah. Yeah. It's definitely connected in a way, and I think sometimes when people talk about... You mentioned like being opinionated, and I think sometimes when people talk about being opinionated, it can feel like they're almost saying like, "Hey, this is arbitrary," like your opinion and my opinion, they're just too opinions, man. Neither is right or wrong. What we try to do is find where there's actual consensus amongst a lot of different high performing teams, and then we can take those practices and say like, "Okay, for a team that isn't already practicing this, can we give them a button so that they can start practicing this?"

**Nan Yu** (00:57:56):
When we see companies doing a really good job of managing their triage queue, but it's very manual, we're like, "Okay, can we automate this? And then for this other company that really needs it that they don't know this is what they need, can we just give them a button to activate this?" And now they have the practice within their org, too.

**Lenny Rachitsky** (00:58:10):
So, I think the takeaway here is when you choose a tool, recognize it's going to change the way you operate and be thoughtful about is this the way we want to work versus just we just have a problem we want solved?

**Nan Yu** (00:58:21):
Yeah, exactly.

**Lenny Rachitsky** (00:58:22):
I want to come back to something, a thread that's come up a couple of times in our chat is the way you collaborate internally. It feels like there's a pretty unique way. You said you were on all the sales calls. Is there anything that you can share about how you collaborate internally, how the different functions collaborate that may be unlike how other companies operate that might be helpful for them to learn from?

**Nan Yu** (00:58:44):
Yes. Something that's worked really, really well for us is we think of product management as partially like a go-to-market discipline in the same way that sales and marketing are, right? When you talk to people and like, "Hey, tell me how product management works in your company," they'll probably say something about like, "Well, there's engineering product and design. They work in this triad, and here's how they interact and collaborate," and we all understand why that's useful, why that's helpful, but this other form of collaboration between product management, sales and marketing, I think it's something that's probably really underexamined and often I feel like in organizations, you actually see some antagonism between product and sales and marketing, and I think that's a shame because when we come together, the way we think about the way that we think about selling is a matter of like... especially because we sell to very expert practitioners, and they have a very sensitive BS detector.

**Nan Yu** (00:59:51):
So, a big part of what we try to do is we try to help our marketing team pick exactly the right word and the right phrasing to make us sound native to the language that our customers speak and also-

**Lenny Rachitsky** (01:00:04):
You're talking about engineers is my sense, right?

**Nan Yu** (01:00:07):
Yeah. Engineers is a big one, but even product managers, right?

**Lenny Rachitsky** (01:00:08):
Mm-hmm.

**Nan Yu** (01:00:10):
Like product managers know when... They know what the job is like. So, when you come in, you say the wrong words, people give you stink eye.

**Lenny Rachitsky** (01:00:17):
Don't call them project managers.

**Nan Yu** (01:00:19):
Yeah, exactly, for example. So, I think that's a big part of what we have to do. So, on our PM team, we actually have a full-time product marketer, and her job is to... Tactically, it's like all the change logs come from her, all the release notes, and also she's always crafting the language for whatever upcoming release that we're building and working directly with the teams and trying to figure out how to talk about it, and then once we go out and build the campaigns, build assets and things like that, that's where a lot of the language is coming from. It's coming from the work that she's doing and then with sales, they're validating all that message in the field. They're saying the words to customers directly and telling you if it's sticking or not, and then you can have a really good feedback cycle between those three disciplines.

**Lenny Rachitsky** (01:01:05):
What I've seen you refer to this way of working as is a double triangle, which is I think a complement to the PM, engineer, designer. Talk about that and give us a visual of what that looks like.

**Nan Yu** (01:01:18):
Yeah, I think PMs, like product managers, we often have a tough time trying to explain like, "What is your job?" It's a little bit of everything. I think the job that I do that we see it as is you're taking the building side of the organization and the selling side of the organization and bringing it together. You're taking all of the commercial motivations and goals of the company and making sure that what you build actually solves for those goals, and you're tempering that with what's possible and where the opportunities are to actually build stuff. So, to me, it's the PM in the middle, and then you have engineering, product design, and then sales, marketing, product management on the other side.

**Lenny Rachitsky** (01:02:03):
PM is always in the middle-

**Nan Yu** (01:02:05):
Indeed.

**Lenny Rachitsky** (01:02:06):
... but I think that's true from the perspective of PM, and I love this visual of just the PM is connecting the builders to the sellers, and you're involved in both worlds. This connects very directly to Brian Chesky's whole thing about how PMs should be doing marketing. So, the way they changed it, every PM is also PMM, and there's no more... They're product marketers now. That's their title and that's like the extreme version of what you're describing. 

**Nan Yu** (01:02:33):
Yeah. Yeah, and I think Apple's been doing that way for forever, too.

**Lenny Rachitsky** (01:02:37):
Got it. So, the advice here is if you're a PM at a B2B business, lean into the sales and marketing side of it, lean into the go-to-market.

**Nan Yu** (01:02:45):
Yeah, and in fact, if you're leaving something on the table in terms of the kind of impact that you are having at your job, that's probably the thing that you're leaving on the table. You're probably already doing a good job of collaborating with engineering and design. It's probably the sort of sell side that there's an opportunity for you to have more impact.

**Lenny Rachitsky** (01:03:05):
Just to make it even more concrete for PMs that are like, "Okay, I want to do this. I want to do what Linear's doing. I'm going to get more salesy." What does it look like when someone is more is in this double triangle working more closely with sales? You talked about being on sales calls. What else there can you share of just like, "Here, try these things"?

**Nan Yu** (01:03:20):
I think originate the message that you send to your audience. There's a lot of things that marketing does, which you are never going to necessarily touch. There's always demand gen and figuring out channel strategy and all this kind of stuff, like sure. That's a peer marketing concern, but actually picking the words and where the emphasis is, like you should understand the customer at a pretty deep level, probably deeper than any other group at the company because of the kinds of requirements gathering, discovery that you're doing. So, you're going to know the native language that your customers speak a lot better and help your marketing team originate those words.

**Lenny Rachitsky** (01:03:58):
Got it. So, basically be really involved in the product marketing, the writing, the emails, the headlines, the website?

**Nan Yu** (01:04:06):
Yeah, yeah, exactly. I know the word product marketing is also so overloaded. They do so many different things, but it's that sort of content creation piece that you really have an opportunity to contributes to.

**Lenny Rachitsky** (01:04:16):
Yeah, I love how concrete that is. It's like don't think about this concept, product marketing. Just think about the words that your potential customers and customers see. Okay, final area I want to spend a lot of time on is totally different. It's around getting a job. 

**Nan Yu** (01:04:31):
Oh, yeah. Okay.

**Lenny Rachitsky** (01:04:32):
You have a pretty unique approach to finding a gig. I heard from the founder of Mode about the very unique way you approached getting a job there. I imagine Linear is a similar boat. What advice can you share with folks that are looking for a job, maybe struggling, that work for you when you were looking for your next gig?

**Nan Yu** (01:04:51):
Project management is a unique role. Because we do just about everything, you don't really get pigeonholed into being compared along a single dimension with everyone else, and everyone who's hiring PMs, just like when they're hiring execs, they're hoping that they bring them on to solve some burning problem that they have. So, it's your job when you're in the interview process to figure out what that burning problem is. So, put on your discovery hat and go figure out what is the actual job to be done of the hiring manager when they're bringing on a new PM onto their team? And if you can do that and then make a good case that you are the person to solve that problem, then hiring you becomes a binary choice between do I hire the solution to my problem or do I hire someone else?

**Nan Yu** (01:05:48):
And I think what ends up happening a lot is when you're in a interview process, you're just trying to put your best foot forward, trying to say that you're great at everything. You have very few weaknesses. Maybe you tried too hard, like whatever, but everyone's going to say that. So, you're just one of end people, and you want to make yourself a little bit of just you versus the field. You're the solution to a problem and then everyone else is like a roll of the dice. 

**Lenny Rachitsky** (01:06:15):
So, the way you're describing it is the company has a job to be done, say it's drive growth of some feature. In this case, it's like for Linear, just build a killer or successful B2B product. I don't know. That's a broad one. Usually, you're not interviewing for head of product role, so that's maybe too broad. So, it's like what is this PM role's job to be done at the company and then help convince them you are the best person to do that job and solve this problem for them.

**Nan Yu** (01:06:42):
Yeah, and a lot of times when you take that approach, it'll feel like you already work there, and the way that I did this, like I got advice from a friend. He said like, "I was interviewing for this job at Mode that you referenced." I'm like, "How should I approach it?" He's like, "Just act like you already worked there. What would you do?" And then it's like, "Okay, I could do that." So, then when you're in this interview process and someone's asking you questions. He goes, "Do you have any questions for me?" You can ask them like, "What are your OKRs this quarter? How can someone help you achieve those?" You can be that specific about it, and they're like, "Oh, yeah, sure. I can tell you about the exact thing that I'm doing this quarter, and then you'll have some level of intelligence about what people are actually trying to solve because I think often we just get stuck in these very high level general types of questions like, "What's the company goals sand all that kind of stuff, and it's like, no, you can get really specific. If you were collaborating with that person in your job, what would you say to them?

**Lenny Rachitsky** (01:07:39):
I love how actionable this advice is. There's obviously an element of this takes work and time. A lot of people are interviewing at a lot of companies, trying to find a job, is part of your advice. Pick the ones you're most excited about and invest a lot of time in this way of interviewing.

**Nan Yu** (01:07:58):
You can invest a lot in the ones where you know that you're going to be able to over deliver on. If you understand what they're actually trying to solve, then you know where you're going to have both the highest chance of success of getting hired, but also doing a really great job on the other end of it.

**Lenny Rachitsky** (01:08:13):
And you talk about how you're like pretending you have the job, pretending you actually have this job as part of the interview process. Oftentimes, as an outsider, you don't have enough information to have a really good thought on what the solution is, and maybe part of it is going to be so wrong because you're like, "I don't actually know. I don't have the data." Do you actually try to reach out to the engineers and designers on the team to try to understand things? How far do you go to try to solve these problems and show them what you can do?

**Nan Yu** (01:08:37):
Yeah, I mean, you're in the interview loop. These are people that you're going to be working closely with. So, start there. Do your discovery questions, and if there's an area that you think you want to dig, you can ask. There's no harm asking, "Hey, can you put me in touch with an engineering manager who's working on the same problem?" And if no one else is asking, again, you're going to have an extra piece of feedback from that eng manager. So, yeah, like this guy asks really good questions, and it seems like they're really with it. No one else is going to have that piece of feedback. So, during the debrief process.

**Lenny Rachitsky** (01:09:08):
And just asking that question alone will show them how deeply you're thinking about this already?

**Nan Yu** (01:09:14):
Yeah.

**Lenny Rachitsky** (01:09:15):
Amazing. Nan, is there anything else that we have not covered that you want to touch on or share or you think might be helpful to listeners before we get to a very exciting lightning round?

**Nan Yu** (01:09:30):
I have a very specific point of view on deadlines. I don't know if that's [inaudible 01:09:34] you care.

**Lenny Rachitsky** (01:09:34):
Let's do it. Fire away.

**Nan Yu** (01:09:38):
I think what often happens is people get depressed about deadlines. It's like, "Hey, here's the ship date," and then you never make it. I don't know if you've had this feeling before.

**Lenny Rachitsky** (01:09:47):
Absolutely, with some deadlines.

**Nan Yu** (01:09:49):
You were an engineer before too, right? So, it's just like engineers is basically like, "Oh, yeah. Yeah, deadlines, they're complete fabrications," and the only way to make deadlines real is to take them so seriously that they are basically like a P0 problem, and everything else has to not matter in comparison to the deadline because that's the only way you're going to be able to signal to the team and also to all the stakeholders that you're actually taking it seriously. So, my feeling on deadlines is don't have too many of them, and when you do, it's a P0. So, the engineer is working on it. They don't get to work on anything else.

**Nan Yu** (01:10:28):
It's like, "Oh, I need them for this," like nope. Nope. You're not pulling them off of anything. We're doing this. As a PM, your job is to just cut as much scope as possible to make it possible to hit that deadline. Like what are the things actually blocking us from doing it? Because what you want to do is at the moment where you have to make the go, no-go call on whether to ship, you want to be able to actually have a product that you can say yes to. It might not have all the features you had wanted or whatever, and you can say no. You can make that choice, but you want to set yourself up to be in a position where you can actually say yes or no to something, because what often happens is like we want this thing. Well, it's not even close to being done yet, so there's no possible way we can say yes. I can't ship it. It's half broken. It's like, "No, no, no. You want to get to a point where it works. It might not be the product that you want, but it is an actual real product that you can conceivably ship."

**Lenny Rachitsky** (01:11:19):
So, you said that don't have too many deadlines, but when you do, make sure you... Everyone understands these are actual deadlines. When do you decide it's worth having a deadline? Is it like a marketing launch sort of thing? What's worthy of a deadline in your experience?

**Nan Yu** (01:11:32):
Yeah, it's usually having to do with some kind of external marketing type of exercise that you're try to hit.

**Lenny Rachitsky** (01:11:39):
Got it.

**Nan Yu** (01:11:39):
And I think that that's the other thing that I think. As builders, we can often look at launch dates and stuff like that. It's like, "Oh, who cares if it's a little bit later or we skip this change log," or whatever it is, and I think that that's really a... I don't know. It makes me go crazy when I hear people say that in all honesty. With marketing and communication with customers, you basically have a limited amount of opportunities to do so. A year is 365 days. There are 12 months. Each of those months has about four weeks. There's some rhythm where you get to have 50-ish weeks to say something to your audience once a week, or you get to have 12 months to say something really big or four quarters to say something huge. If you miss one of those opportunities, you don't get it back again. You can't time travel back and say like, "Okay, actually, let's redo first quarter and say this message that we wish we could have gotten into the field."

**Lenny Rachitsky** (01:12:35):
That is such a powerful point. I could see the sales marketing, go-to-market element of your job coming out there. I imagine everyone that's in that field's like, "Yes, this is exactly right." Maybe just the last question along this line. So, I love this idea of taking deadlines very seriously when you commit to a deadline. At the same time, as you pointed out, it creates a lot of stress knowing there's a deadline we have to hit. So, one lever you've mentioned is cutting scope. Another is just people spending more time estimating to have more accurate deadlines. You invest in that. How do you think about just for an engineering team to come into a deadline, how much to spend on de-risking and estimating versus just, "Let's just do our best and then we'll cut and adjust"?

**Nan Yu** (01:13:18):
This might be my hot take, but we do almost no estimating in order to hit deadlines. What we do is we ship as early as we can. The thing we talked about earlier where if by the time that 10% of the time has elapsed, you have a working thing, you can now spend the rest of the time deciding whether or not you want to do another iteration or you want to polish that thing and get it to be a shippable state. So, you're setting up your future self to be able to make that decision. So, none of this is... You can't go into this at the very last moment and say like, "Okay, now, we have to take the deadline seriously." You have to do it from the beginning and commit to the process of going very fast, iterating early, and then putting yourself in a position where you can say yes or no to a product.

**Lenny Rachitsky** (01:14:03):
So interesting and so different from the way most companies operate. Nan, this was everything I was hoping it'd be. I think this is going to help a lot of people build much better product, which would be good for the world if more products are like Linear. With that, we reached our very exciting lightning round. Are you ready?

**Nan Yu** (01:14:20):
Yeah, let's do it.

**Lenny Rachitsky** (01:14:20):
Okay, let's do it. Okay, first question. What are two or three books that you have recommended most to other people?

**Nan Yu** (01:14:29):
I think the one book that I recommend the most is The Design of Everyday Things by Don Norman. I read it originally in college for an HCI class I was taking, and I think of everything I've ever read, it's the thing that caused me to see the world from the perspective of everything you interact with as a product. Every pencil that you use, every door that you open is a product that somebody designed.

**Lenny Rachitsky** (01:14:55):
And is that the big takeaway from that book? Because it comes up a lot, and it's such an old book. So, I guess for someone that hasn't read or maybe doesn't have time to read, it is the big takeaway for you. Someone designed everything and there's a reason things aren't great, and they can be improved.

**Nan Yu** (01:15:10):
Yeah. I mean, I saw this the other day. I was at a café in my neighborhood, and I saw a kid rip a handle off a door, like of the café. He pulled it so hard, it came right off because it was a push door, but it had a handle that looked like you could pull it, and that's one of the canonical examples of the book because [inaudible 01:15:25] are just mysteries. Yeah.

**Lenny Rachitsky** (01:15:28):
Awesome. Next question. Do you have a favorite recent movie or TV show you've really enjoyed?

**Nan Yu** (01:15:33):
I watched The Diplomat on Netflix. I think it was terrific. It's really fun, easy watch. It has some West Wing vibes if you were into that back in the day.

**Lenny Rachitsky** (01:15:44):
Yeah, have you seen the second season?

**Nan Yu** (01:15:46):
Yeah, I finished the second season. Yeah.

**Lenny Rachitsky** (01:15:48):
I wasn't as excited about the second season, just to put that out there. The first season was really good and then just went off a little like, "Okay. I guess it's cool," but stuff like that.

**Nan Yu** (01:15:55):
Yeah, it got a little like spy thrillery, I think.

**Lenny Rachitsky** (01:16:00):
Okay, cool, but still really good and on Netflix. Okay, cool. Do you have a favorite product you recently discovered that you really like?

**Nan Yu** (01:16:06):
I didn't discover it, but I discovered a version of it that was really interesting. There's a pen. Actually, I have one on my desk. It's called the Sakura Micron. I don't know if you use these. It's like a felt tip pen. It's really great. It was originally invented in Japan for artists to draw comic books and stuff, and you can use it for anything. I use it for journaling or whatever, but I was on Amazon. I was trying to buy more, and I found a package that said like, "Bible Study Kit." I was like, "Why is this labeled Bible Study Kit?" And it was literally just the pen in four different colors, and it was because the thing doesn't bleed through pages. So, if you have a Bible, which they often have these really flimsy newsprint pages. It's not going to bleed through.

**Nan Yu** (01:16:51):
And it's just really interesting to me that someone marketed a normal package of these pens as a Bible study kit and for people who were looking for that keyword, and it was official, too. It was not something hacked together. It was actually an official packaging of this.

**Lenny Rachitsky** (01:17:04):
Amazing. What a unique pen choice. Two more questions. Do you have a favorite life motto that you often come back to and find useful in work or in life?

**Nan Yu** (01:17:15):
The correct amount is too much minus one, and I think this ties into the try the extreme version of it of a thing where... I don't know, like a stupid example, like how much pizza do you want to eat? It's like, well, five slices was too many. I feel bad. Then four was probably the right number, and then if you want to find the right number, sometimes you just have to really shoot for the edge and then find out what's too much, and then you'll find out exactly what the right amount is.

**Lenny Rachitsky** (01:17:41):
I love how tactical that is, makes me think about Elon Musk's thing about cutting things. Like one of his formulas for just getting stuff done, one of them is just cut stuff before trying to optimize it and automate it, and his advice is if you don't bring back 10% of things, you cut, you're not cutting enough.

**Nan Yu** (01:17:59):
Yeah, exactly. 

**Lenny Rachitsky** (01:18:01):
Final question. You worked at Everlane for a number of years, and you shared the rough idea of a story around a shirt, maybe a bestseller that they have now, and how you helped create a bestselling women's shirt. Can you share that story?

**Nan Yu** (01:18:19):
Yeah. So, I mean, to be clear, I witnessed the creation. I don't think I had a direct hand in it, but yeah. So, I saw this advertisement the other day on Instagram for... It's called the Women's Box-Cut Tee, and it's a wide and short for women, and I looked, and it had 20 colors of it, and it sells super well, and I remember when we created this thing, and it was because there was a batch of defective men's t-shirts. They all came in an inch and a half too short. So, we couldn't sell them. You would have your belly button sticking out. No one wants to wear of that. So, what we did was like, well, we have to salvage the inventory because we were a very small company, and we had to make cash flow, and we couldn't just damage it out.

**Nan Yu** (01:19:06):
So, the design team and the marketing team came together, and they said, "Okay. Here's what we're going to do. We're going to cut another two inches off of this and make it really cropped and market it towards women as like a cropped boxed-tee silhouette, and we did that. We're like, "Okay, hopefully, we can salvage this inventory and not have to take a write-down." It sold out in a week, and we're like, "Oh, okay. I guess we just made a hit product," and it's one of these things where it's very hard to know what this was. Was this a marketing thing? Was this a design thing? I don't know, but you just come together, and you find the right product market fit in the weirdest way.

**Lenny Rachitsky** (01:19:43):
I love that it's still going.

**Nan Yu** (01:19:43):
Yeah, it's still going. Originally, it was just white. Now, there's like 20 colors.

**Lenny Rachitsky** (01:19:48):
Oh, man. I love how many industries you have worked in: fashion, data analytics, project management. I don't know what's next. There's more, I imagine. Nan, this was incredible. I really appreciate making time for this. Like I said, I think we're going to have helped a lot of people build better products. Two final questions, where can folks find you online if they want to reach out and learn more? And how can listeners be useful to you?

**Nan Yu** (01:20:08):
Yeah, I'm on X/Twitter as the thenanyu. It's T-H-E and then my name, and if they have any feedback about Linear, we're very happy to take it, especially for people who use it in their day-to-day. We really want to hear from users.

**Lenny Rachitsky** (01:20:26):
What's the best way for them to share that? Is it tweet at you? Is it go to the website? What do you recommend?

**Nan Yu** (01:20:31):
Oh, yeah. You can tweet at us. You can DM me on Twitter. My DMs are open, so it's all good.

**Lenny Rachitsky** (01:20:36):
Amazing. Nan, thank you so much for being here. 

**Nan Yu** (01:20:39):
Yeah, of course. Thanks, Lenny.

**Lenny Rachitsky** (01:20:40):
Bye, everyone.

**Lenny Rachitsky** (01:20:43):
Thank you so much for listening. If you found this valuable, you can subscribe to the show on Apple Podcasts, Spotify, or your favorite podcast app. Also, please consider giving us a rating or leaving a review as that really helps other listeners find the podcast. You can find all past episodes or learn more about the show at lennyspodcast.com. See you in the next episode.

---

## [6/15] Tobi Lutke

**Tobi Lütke** (00:00:00):
Your podcast is a podcast by a builder for other builders. Here's the most interesting question I think people can ask builders, what is your energy source? My energy source is dissatisfaction with status quo. There are so many books are about this ... Technology leading to dystopia. Like no one who really thinks about this would want to be born into a world 20 years before today. I think today is the dystopia of the future. It behooves us to try to build the kinds of products that leads in ... Towards progress.

**Lenny Rachitsky** (00:00:28):
There's a couple quotes along these lines I've seen that describe the way you think about this stuff. "If most people are doing it a certain way I by default don't want to do it that way."

**Tobi Lütke** (00:00:34):
There's an aesthetic in the world that exists which is that business people dress in suit and tie, they are speaking much more sophisticated than I do, usually without an accent. They usually have a stick and show dramatically at the chart that is behind them. How much is that aesthetic overlapped with outperformance? Pessimism sounds extremely sophisticated. Optimism always sounds dumb or at least naive. The most powerful unquantifiable things in the word of business are fun and delight.

**Lenny Rachitsky** (00:01:04):
I don't know of any other company that operates where the founder has this 100-year vision of where the product needs to go and working backwards from that.

**Tobi Lütke** (00:01:11):
I talk about look in the future and then think backwards a lot, right? It's like what would we want to have done 20 years ago on this? We have very long-term plans. At 100 years you can't talk about this software project but you can talk about the mission itself, whatever things that will survive for 80 years that are left on this particular timeframe. Entrepreneurship is just precious. Shopify exists, basically, to make entrepreneurship more common.

**Lenny Rachitsky** (00:01:35):
Is there anything you want to leave listeners with?

**Tobi Lütke** (00:01:37):
I really, really, really think that there is not a single person on this planet who is even close to being at their maximum potential. Reminding people of their own potential constantly is actually a wonderful thing to do.

**Lenny Rachitsky** (00:01:54):
**Tobi Lütke** (00:04:24):
I'm glad to be here. I'm excited for our conversation.

**Lenny Rachitsky** (00:04:26):
I've listened to so many of your other interviews, I've talked to a bunch of people that work for you. I want to try to do something a little different. There's basically two themes that emerged over and over and over as I've listened to you share advice, and interview, and in talking to people that work for you. One is thinking from first principles, the other is maximizing human potential. I'm just going to plant these seeds for now. I'm going to not ask about these directly, I'm going to come at these from the side with the many questions that I have for you. Something that I've heard people describe at Shopify called the Tobi tornado.

**Tobi Lütke** (00:05:01):
Oh, wow. Okay, that's a start. I like it.

**Lenny Rachitsky** (00:05:06):
What is the Tobi tornado?

**Tobi Lütke** (00:05:08):
A Tobi tornado, I would say, is a whole lot of change management or a conversation or conflict or real talk compressed into a very short timeframe. I see something it doesn't ... It's not good, I have a conversation. I have learned something very quickly about hey, I need to update my priors, or cool, let's do it differently. At which point a project might be stopped and we get back together in a room and then we start a new version of a product. And everyone who's currently on the team, of that particular project, is no longer on the project but they are the founders of the next version which is built differently. And that might be a bit whiplashy for people. I mean, I certainly hope that's true, it's certainly what people tell me. It's also what they appreciate about the company. It's like what is best ends up mattering a lot.

**Lenny Rachitsky** (00:06:08):
Basically, it's you going into a chat room being like "Hey, this" ... "We're going to end this project we're" ... "Let's try something else" because you've discovered or realized this is a bad idea. Some people complain about this practice of like "Oh, Tobi kills the we've been working on." My lens is, you realize we're just wasting time on this thing that is not going to work and we shouldn't do it. Is there anything there along those lines?

**Tobi Lütke** (00:06:32):
No. This is everything. Again, once I imagine something might be not the right thing to work on I'm either incorrect at which point this is super important that I understand why or I'm correct at which point it's super unfair for letting people work on something that isn't going to make it. There's a third wave which is I could also ignore it but that's an abjection of my CEO and founder responsibility that I'm absolutely not willing to make so that's just not a path forward that I see valid. I understand that's what a lot of people choose to do. So yeah, compressing time is important. We have a fairly limited time in our careers, right? Our careers are not that long. If you're lucky you have 40 years in the industry. Most people spent more time in school and then maybe leave later if they're so lucky that they can. So it's not even that.

**Lenny Rachitsky** (00:07:33):
I think you want to do the maximum amount of things you can be proud of at the end of your career. When you look back you want to be saying, "Hey, holy shit, we shipped this thing which was absolutely an incredible contribution to a mission I cared about at a company that was full of other people who cared as much as I did." And also are very proud of working. And maybe even thinking of working with me and being really, really glad that we spent time on projects together. Yeah, none of this happens if everyone's sort of optimizing a thing that probably shouldn't be there, right? And, therefore, I think it's the better thing to do.

**Lenny Rachitsky** (00:08:16):
And, by the way, I love this way of describing ... Of compressing it in time to just make a decision, not focus on making it come across the most kind, nicest, sweetest way. When you give feedback to people when something is not the way that you think it could be or as good as you could be it's often very direct and often hard to hear. To me this comes across as you're trying to maximize their potential, you're trying to push them to do something better. Is there anything there that you think is a way of approaching feedback?

**Tobi Lütke** (00:08:49):
I really, really, really think that there is not a single person on this planet who is even close to being at their maximum potential. I just think everyone is way, way, way, way, way better than they think. And the reason why we're not performing at this level is a series of ideas, maybe certain approaches for cultivating our skills and our crafts that have not yet been discovered and, therefore, we could not take advantage of them. It's an environment that just narrows the focus on fairly unambitious things at which point you get stuck competing with literally everyone else in the world because everyone's unambitious.

**Lenny Rachitsky** (00:09:33):
I have found that reminding people of their own potential constantly is actually a wonderful thing to do. And I have a history of being right about people's potential more than they are themselves. Now, in a way, this dooms me fairly often to be disappointed, right, in myself. By the way, I'm talking about myself here too. I think I have way more potential than what I bring to bear and I hate that so I'm trying to cultivate the skills that I need for tomorrow and constantly challenge myself. I'm harder on myself than on anyone else. By some discount rate act equally to the people around me, especially the ones who just so, obviously, are brilliant.

**Lenny Rachitsky** (00:10:23):
Spending time and longer time in careers with people and then holding them to a high standard means that they accomplish very often things that just they didn't imagine they could. To me this is the most wonderful thing to see. And frankly this is a throughline for all of my career because this ... My product is that. I want my product to cause people to be more successful than they thought they could. And, in fact, become more ambitious about what they are building with their online stores and their businesses than they are actually initially set out to do.

**Lenny Rachitsky** (00:10:56):
Because something like this happened to me, right? I started a snowboard store at some point. I didn't set out to build Shopify. If you are committed to following your curiosity as to the next step, and optimize for maximum amount of learning when you choose these steps, it takes you from one place to another and you actually realize the world's full of lies about human potential and progress. Maybe people are not malicious about it but they're definitely confused about it. School teaches you that you have to learn this particular piece of math in this 12 month period, and it doesn't matter how much you understand it. It's like the outcome can be variable and will grade you on a variable outcome for a fixed amount of time which has nothing to do with anything I've ever seen or learned or see ... Or witnessed about how to actually learn things. You follow that thread and you just find that there is no speed limit for personal growth. In a way Shopify has been a wonderful experimental lab for this sort of conviction. I've just seen this to come to be true. And, of course, hearing from someone that you respect that "Hey, I think you had it in you to do this thing significantly better because I think you probably saw fairly early in the project this sort of path A path B. You chose path B potentially out of convenience even though you knew that wasn't the right thing. And I actually expected better of you and I expected" ... "I think the next time this happens in your career you should go path A because" ... "Based on your conviction." And, therefore, that's hard to hear, right, because it's right. But it's also extremely valuable, right? What I love is an environment of people who are holding each other accountable to the actual potential rather than sort of their current level plus or ... Plus some, I don't know, a little bit extra.

**Lenny Rachitsky** (00:13:19):
There's so many threads I want to follow here. The example of the school is such a pertinent one. To me right now we're looking at preschools for our son and I'm ... They're describing their education philosophy I'm like, I don't know why I should believe this is the right approach. And it makes me just want to spend all this time researching what education approach works. I know it's just preschool and maybe not as critical yet.

**Tobi Lütke** (00:13:38):
Did you have a bet yet? I have three kids too and this is sort of a decision that every parent faces, right? So many of your listeners are probably product managers of machine learning products that maybe this resonates. So there's a funny thing about machine learning which you're just like ... You train on a lot of data, and, hopefully, you get something that predicts the thing you want it to predict correctly out of it. The biggest problem of this is overfitting, right? What does good look like, a loss function? Which is a heuristic because it's not the actual task that the thing will do in the future it's something that proxies to the task that you want the thing to do in the future. Predict fraud, predict the next word, whatever. So overfitting is basically model learning how to cheat on the benchmark or on the fitness function.

**Lenny Rachitsky** (00:14:39):
So there's a business analogy of this which is that ... It's called Goodhart's law. It's literally the same thing as overfitting just for businesses. Goodhart's law just says, "Any metric that becomes a goal ceases to be a good metric." Same exact thing. The universal truths are things that almost any competitive field will invent for itself by different terminology often. And I think this is also, by the way, why it's so interesting to focus on personal growth and learning a lot about a lot because you end up finding these sort of hidden harmonies behind things, the things that are clearly enduring correct insights. So overfitting, Goodhart's law are the same thing. School optimizes for what? Marks supposedly, right? In fact, overfitting in school is literally the kids cheating to get marks, right? You get another analogy. What is however the right loss function for children? Have you made a decision yet?

**Lenny Rachitsky** (00:15:53):
No, I have not we just started down this path. You told me-

**Tobi Lütke** (00:15:55):
It's the kind of thing [inaudible 00:15:57]. You have to actually go fairly deep in philosophy to figure this out. And then again afterwards you can build, you can find the schools that you like. For us it was just maintaining curiosity. This is a completely different goal from being good at marks. But I just think everyone's born extremely curious and school has a habit of getting it out of kids. Literally, there's a foundation model of a child, and you fine-tune it at school, and it just loses the neurons of curiosity because it's actually discouraged to meander into other topics and explore them just because they're interesting. I don't know. This is sort of not the beat of a podcast but I just think about this a lot. It's funny how these things just recur constantly.

**Lenny Rachitsky** (00:16:48):
So when Archie was on the podcast, he's the head of growth at Shopify ... I don't know if that's his official role. Basically drives a lot of the growth. He talked about how the core product team, outside of the growth team, operates without KPIs, without specific goals. And decisions are driven by taste and intuition primarily you, and Glen, and some other leaders. And a lot of people heard that and they're like "I" ... "First, I don't believe that. Second of all, how does one operate that way when there's no data to tell us exactly what is right and good?" So the question I have is just how does one operate in that way successfully? What does it take for a company to work that way because a lot of people will try it and fail?

**Tobi Lütke** (00:17:25):
This is very close to what I just talked about before with [inaudible 00:17:28]. Goodhart's law is real. The moment a metric becomes a goal it's no longer a useful metric, right? I think that's more or less a precise wording. Why? Because no metric by itself is a complete heuristic for a complex business because business are complex. There's a million of different tensions in a company and you can't all keep them in harmony by optimizing for one fix. It's true that we don't have KPIs and we don't have at least OKRs in the Silicon Valley sense but we are extremely data-informed. We have invested enormous amounts of money and time into systems that give us basically everything at our fingertips. I sent this demo to other founders and they're completely bowled over by the way we can dig into basically every constituent atomic bit part that makes up the cohort that just got formed 15 minutes ago by the end of a ... At the end of a quarter or month or week.

**Lenny Rachitsky** (00:18:39):
In a lot of different places, this is one of them but also in its products, it's just not overfitting for the quantifiable. Everyone competes for everything but it's highly quantifiable because it's ... It's fun, it's like a game. You tweak a number and 0.1 more is better than 0.1 less. That's an immediate gratification thing. But I just think the overlap of most valuable things you can do with a product, and for things that happen to be fully quantifiable, it's like maybe 20% which leaves 80% of a value space unaddressable by the people who will only look at quantifiable things. Shopify is comfortable with the unquantifiable things such as tastes, quality, passion, love, hate. It's with the strong emotions that people have.

**Lenny Rachitsky** (00:19:49):
The sort of deep satisfaction that a craftsperson feels when they've done a job well is actually better proxy if you allow it to be then the ... Do a unit test pass. A unit tests might not pass. And the unit tests will pass 15 minutes later because we already fix them or adjust the one or two things so they support us. We have systems that tell us exactly if something goes the wrong way. There's an extremely sophisticated rollout system in Shopify that forever holdouts and correlates everything with everything for ... In every experiment and so on and so on and so on. But if you think about it as a cockpit for a pilot. The decisions are still made by pilots and we think this leads to better results. It's just the same with our product. There's plenty of A/B testing tools and all these things for commerce, and it's, of course, really important to figure out what your conversion rates are. But are you representing your brand is an unquantifiable question? Are you proud of the thing that you have built? Do you feel it's your own, right?

**Lenny Rachitsky** (00:21:03):
And so I think there needs to be more acceptance in businesses or for unquantifiable things. The most powerful, not unquantifiable things in the world of business are fun and delight. If people have fun when they're doing something that is just upstream from so ... Sorry, downstream from so many other things. I think that if all the metrics are pointing down but everyone says, "My God, I'm having so much more fun," I think that the very next thing that will happen with some time delay is all metrics will start going up. And if that doesn't happen then we adjust course.

The reason why we specifically don't have OKRs and these things is because ... If you want to hold the unquantifiable as things that are stable and exist ... That people actually do really defer to them and really actually learn to be okay with someone just saying, "Hey, this is actually just really great and they're shipping this" then you need to make certain edits to a business that don't remind everyone too much of the companies they might have come from which the only way to get promoted is by driving the metric up. It's a bit of a [inaudible 00:22:24] conversation I suppose. Good on the fortune cookies saying, "Shopify doesn't do OKRs or doesn't do metrics" and so on. But it's actually just because the metrics take us [inaudible 00:22:37] function where we often defer to just more ... Sometimes a little bit emotional but generally less quantifiable things.

**Lenny Rachitsky** (00:22:48):
I imagine if someone were to hear you describe this of focus on joy, and fun, and love, and delight, maybe ... It's easy to dismiss that.

**Tobi Lütke** (00:22:58):
It sounds completely idiotic, right? Again, there's an aesthetic in the world that exists which is that business people dress in suit and tie, they are speaking much more sophisticated than I do usually without an accent, have a full head of hair. They talk about metrics, they are in front of PowerPoint presentations, they usually have a stick and show dramatically at the pie chart that is behind them. And highly charismatic, highly ... So that's our aesthetic. How much is that aesthetic overlapped with our performance? I don't know but some of them ... Some people pull it off who are like this. I think the world is sort of stacked to lead us astray based on our stories about what optimal looks like are just so incorrect in so many ways. Optimism always sounds dumb, or at least naive. Pessimism sounds extremely sophisticated. Metrics driven sounds extremely sophisticated. Talking about fun sounds like naive.

**Lenny Rachitsky** (00:24:14):
Well, first of all, I've always ignored what people think generally. That came pretty natively to me somehow which I'm very lucky about. But I've now actually learned that almost all of the alpha in the world is now in the ... Exactly the things that are unobvious but true. And the things that people dismiss as naive or so. The most successful business person on planet Earth is Elon and he conforms to no idea of what the most sophisticated business person ought to be like in any which way you can imagine. I think we live in a world where the counterfactuals are winning because our ... Because aesthetics are just leading us astray.

**Lenny Rachitsky** (00:25:02):
This is an awesome segue to the other theme that I wanted to spend some time on which is thinking from first principles, Elon is the classic example of that. Honestly, I think you're the other most classic example of that these days. And we'll keep talking about all the ways you operate very differently from other companies which are examples of this. But I want to read a quote from Glen Coates, he shared with me, of how he sees you that gives an interesting lens into your first principles thinking. So here's what he said about you. "Tobi is at his heart a true futurist, he's obsessed with the way things should be in the future. Being data-driven is innately being anchored in the way users and technology are behaving today. He's never really said this to me explicitly, but knowing him I think any design that is drawn primarily from the way things are or were is one that he sees as inferior to one that is skating to the puck of the way things could or should be." Does that resonate?

**Tobi Lütke** (00:25:55):
Yeah. I mean, I think that's correct. That's actually really interesting. Your podcast is a podcast by a-

**Tobi Lütke** (00:26:01):
Your podcast is a podcast by a builder for other builders. Here's the most interesting question I think people can ask builders is like, "What is your energy source? Where are you getting energy from?" I think fundamentally the world exists at room temperature. Almost all companies are running at that, humming along, doesn't do anything. There are certain individuals who can inject heat into businesses. Founders do this very well. All the startups anyone's ever heard of have people who are injecting heat because if no one would inject heat into the business, at room temperature, you cannot outperform anyone else. You can't be hotter than everyone else if no one's injecting heat into the concern. So fundamentally, there is a injection of energy into companies that comes from founders and the best leaders, like all the people you've had on the podcast from Shopify have a perfect set of cast of characters of people who are just exothermic. They are just like wellsprings of energy that leads to all the amazing results that we get to enjoy.

**Lenny Rachitsky** (00:27:21):
So the question is where does energy comes from? And that's another one of our discussions which very quickly goes into the emotions. Actually, there's a really... So I watched The Last Dance, the Netflix special of Michael Jordan a while ago, of course, but there was one scene where he just, I'm sure this is a super famous story and he just made up an insight that someone told him so that he would then go and just want to destroy them afterwards, which he then of course proceeded to do because it's hard to imagine anyone more exothermic than him. So we know what his energy source is. It's rivalry. It's potentially it's insight or it's anger, something like this.

**Lenny Rachitsky** (00:28:06):
I am not... My energy source is dissatisfaction with status quo. My fundamental belief is all this talk about technology where all... So many books are about just technology leading to dystopia. You know what dystopia is? Today, compared to what it will be in 20 years ago or any. And you can play this for any part of human history. I'm not making a future statement. I'm making a almost totalical statement about the experience on planet Earth. No one who really thinks about this would want to be born into a world 20 years before today rather than today. And so I think today is the dystopia of future, and I think it behooves us to try to build the kinds of products that lead in towards progress in a small way or a big way. But yes, I think if someone comes to me and says, "Hey, let's go do this thing. And we've looked around and here's how people solve this problem, let's make a good version of that," I'm like, "That was not the job." Because everything that you encounter, that every solution, every product, everything that exists is path-dependent, highly, highly, highly path-dependent, and often path-dependent based on having to make compromises, based on things that were true at the time a decision was made but are no longer true.

**Lenny Rachitsky** (00:29:38):
The entire field of... What was it? I forgot the name of a field. Chomsky's field, the linguistic research field. It's cool. We now have autoregressive models that are just like, we don't actually need to set up a complete... We don't need to research the structure of grammar to be able to make machines also engage in the spoken word. We actually can just train on the internet, it turns out. So that was not possible back then because you didn't have the right architecture for this, but now it is. So I think what you have to do is to actually have, when you come up with a new product or you discuss a new product, you have to derive it from first principles. You have to say, "How would we solve this problem given every fundamental building block that we have available right now?"

**Lenny Rachitsky** (00:30:25):
For that, to do that, you actually have to understand the power and the composability of all the building blocks that exist right now, which is a tall order and no one is perfect at this. But so this way, you go ahead and say, "Okay, cool, so this is how we are implementing this thing. This is how it will be implemented today." And now we can talk ourselves in taking shortcuts. "Maybe we should actually start up doing it the way everyone else does. Maybe we derived exactly what everyone else does as the correct thing to do."

**Lenny Rachitsky** (00:30:56):
Sometimes there was a lot more wisdom encoded in the status quo than you expect, which is I think is super delightful. Then you figure that out and so when you act on it. But what isn't okay is skipping the exercise and doing the same thing everyone else does because that is again a abdication of product leadership. And so yeah, I would say I become extremely suspicious if I get a pitch to do a good version of the same thing everyone else does because I just find that in our space specifically very rarely to be the best solution.

**Lenny Rachitsky** (00:31:33):
There's a couple quotes along these lines I've seen that describe the way you think about this stuff. "If most people are doing it a certain way, I, by default, don't want to do it that way. And if you want to do something world-class, you can't do it like everyone else."

**Tobi Lütke** (00:31:46):
Yeah. I don't even think that's an opinion that I hold. I think that's actually... We are basically in axiom territory here. If you want to do something better than what exists, you have to do it differently. That does not make a statement about if it will be better in the end after you do it. It could also be worse. But you can't get something better done if you do the same thing. It's like axiomatically not possible to do. It's fails Archimedean logic, yet it's something... You would be amazed how many business plans are actually failing Archimedean logic in this way like, "Let's do a good version of this thing that we've already been doing and we will capture 1% of the market," and just like this stuff. It's like, "Trust me." I find it kind of cute.

**Lenny Rachitsky** (00:32:36):
Is there an example that you can share of you approaching the problem this way? I imagine it's constantly happening. You also mentioned you're born this way. So I think it's hard for someone to just sit down, learn, think the way Tobi thinks, but I'd love to help people start to approach problems this way, so maybe an example might help.

**Tobi Lütke** (00:32:54):
I think this is entirely learnable, I think. And so I encourage people to just have a practice of think step by step essentially and just do it. It'll become a habit pretty quickly because it just, it also just outperforms.

**Lenny Rachitsky** (00:33:09):
Examples. The very first example is Shopify itself. It's cool. So there was lots of e-commerce software and it was all the way it was because of path dependence because everyone who want it in 2004, 2005, e-commerce was an existing retailer and therefore they had complex businesses that needed to be ported online, including all of their somewhat Byzantine business logic. "I wanted to make e-commerce software that would do very well on the internet of the future and I believe that we can make it easier to start new businesses online than it is in a physical world because the physical world is encumbered by a lot of regulations and also upfront costs for leases and so on. So let's optimize for that case and build something that is so intuitive to use that frustrated people in dead-end careers can spend their lunch breaks making progress towards building their own business, which then eventually allows them to do it on their own way."

**Lenny Rachitsky** (00:34:25):
And so being fortune was biased, eventually it turns out to be a much better prep for also solving all the enterprise cases because no one had to... Enterprise software is overfit to the sales process, which is that it wins the RFPs because it has every feature ever or at least a way of putting a checkbox next to every RFP line ever. But they're not good, right? An RFP is a great example of overfitting in the world of procurement because it tells you nothing about the quality of software behind it. But honestly, this happens all the time. It is just like here, we are setting the stage for much more, much higher quality retrieval for the products on Shopify and across Shopify. I think we've been in a local maxima on search and we think that, especially with the advances of the new models, certain things are now possible to do that could not have been done yet because again of these layers and layers of path dependence and no one coming and saying, "Is that the best way to do and maybe we should rebuild this component periodically?" We can now do a better job if search that will leads to much more delightful experiences. And so this is a fun project that this happening right now in.

**Lenny Rachitsky** (00:35:50):
I think this is really interesting because what I'm looking for is like the Tobi algorithm of first principles thinking. Elon's got these famous ways of thinking he shared. One is start with the cost of metal to help you understand how much a rocket should cost. And then he's got this five-step. First, decide, do we need this thing? Then figure out how to optimize it, then automate it. What I'm hearing so far, and I'm curious if you've thought about this? And if not, this feels like a really good blog post in your futures, the Tobi first principles algorithm, but I'll share a couple of things I've heard so far as you've described it.

**Lenny Rachitsky** (00:36:24):
One is analyze the path that existing solution has relied on, almost like the assumptions that were true for it to be built back and it was built. And the other is this overfit, what is it overfit for? What is it over solving that maybe isn't necessary? Is there anything along those lines of just how you approach problems?

**Tobi Lütke** (00:36:43):
This is probably too nerdy technical. You are right, but I should figure myself out a little bit. My brain runs on a meter language but isn't directly something I can translate into words, I suppose. I think more about things like this in terms of programming constructs, pure functions, overstate. And I think in any moment, the best decision to do is I think the perfect product lead is almost like a thermostat for high quality product. It's like you're setting, saying, "I would like to build something really, really great and I'm going to go through a series which is much, much more complex than what a thermostat does," which basically checks the temperature and then makes a decision of air con or heating. You make you re-derive literally every decision that is valuable, every foundational assumption, every foundational ABC direction. And you want to see the observation you've made in the meantime since you last derived the next step. Re-running the entire function over the state that is now updated, the higher fidelity information, would you come to the very same thing?

**Lenny Rachitsky** (00:38:08):
Sometimes fairly early in the construct in the tree of foundational assumptions, change is made. An example we all had was beginning of COVID when we suddenly had shelter in place. So Shopify has that incredibly good office spaces and we were very in-person company and we've... Our floor plans because I think we really added something to the understanding of how to put great... I got a lot of founder energy from my co-founder, Daniel, to build great collaborative spaces for creative work with lots of happy accidents, people running into each other and so on.

**Lenny Rachitsky** (00:38:46):
Anyway, we were very, very, very, very determined on doing that. But somewhere in this construct of nesting functions, you have to rerun and foundational assumptions. It's in the stack is the fairly basic Boolean of are people allowed to leave a house? Which was yes. The moment that flips, it's not that just like, "Okay, over here let's do the best what we can do." It's actually that the entire tree now moves into a different place. And it can be a very far place, different, that you land because you will make the same quality of decision on every step, but you need to rerun something that takes you to a completely different landing zone. And so then this is also easy for us to say, "Cool, we are going to be remote only forever? Let's go." And because we realized that the temporal shelter in place would cause a series of events that would make that for optimal best set of trade-offs for the company. And so I think this is... Can I put this into a piffy five-step? I wouldn't know, but maybe it resonates with someone who can help me figure out the coded English language for this because it has been nerdy way the I explained it, even to me.

**Lenny Rachitsky** (00:40:05):
This remote work example is something I definitely wanted to touch on, which is I'm glad you got there. So in this decision, is there anything more you could share about how you got to that place of like, "Oh this Boolean changed so we should rethink this?" Because I know you probably, you weren't... I don't know. Were you in the shower and just like, "Oh wow, we should really go remote because of this"? How did that actually come about? And then I wanted to ask on another part-

**Tobi Lütke** (00:40:27):
Yeah, pretty much actually yes because again, I rerun over all the inputs and figure out what happens. And here's a couple of other things that were starting to fray on the decision to be in person as well. We started in Ottawa, Canada, which is a city of a million people, and it has some tech heritage and good universities. But at a million people, it's just not population dense enough and has a depth of talent pool that it can support a company that is going to 10,000 people. So anyone who runs this optimization function I'm talking about here about location strategy for a new startup will come to the conclusion if they do it right that everyone should be sitting around the same table. Only if you can't do that, do you say, "Okay, let's be all in the same couple of rooms." Only if that doesn't work anymore, you say, "Okay, let's be spread over one floor." If that doesn't work, in the same building. If that doesn't work, in the same city. If that doesn't work, at least stay in the same time zones or make sure that there's good hub connections between.

**Lenny Rachitsky** (00:41:39):
This is how it works. It's just like some assumption somewhere along the line is invalidated. You end up in a different side of a decision tree. But what was happening to us was we already were in four or five cities. We were adhering to the same time zone at this point. But my experience was I went through, sometimes at the Ottawa office, through an entire day where I had 10 hours of meetings, which is fairly normal, but I do every single... I don't think I had a single other person sometimes with me because every one of them was with people in another office. Some people are dialing in remote and so on. So there was an awkward hybrid-ness which we ended up in making only good decisions along the way.

**Lenny Rachitsky** (00:42:29):
This is actually the most dangerous thing. Most of the time you end up in a bad part of a tree, in a local maxima of a path-dependent environment, by only making good choices. People think that making a good choice inoculates you from making mistakes or that the presence of a downside of an idea ends up disqualifying the idea. Both of those things are incorrect. So what you need to do then is... So that was an overlay to decision.

**Lenny Rachitsky** (00:42:59):
The moment COVID started, then we also had this thing of shoppers exploding because we were actually an asset to people during COVID, e-commerce for the local businesses. And we took that very, very seriously trying to make more businesses survive COVID, small businesses survive this particular calamity than otherwise won't which is important because small businesses tend to be wiped out first anytime the times turn fragile. And so we needed to staff up. And so there's a very weird question about where to staff up. So clearly the better input there would be if you could hire people everywhere. And so once that decision flipped, you can see how this is actually now super easy to say, "We are going to remote," because there is no turning back. There is no... Even with this decision, it's a better set of trade-offs for the future is to accept the vast additional difficulty of building a remote company. It's way harder.

**Lenny Rachitsky** (00:44:03):
It's not something you should recommend to anyone to do this because it's the same as trying to run a world record marathon run in Aspen, Colorado. It's like there's not enough oxygen up there to do that. But if you end up pulling it off, you're a real Chad. That's very cool. So I find difficulty itself interesting. And again, I spent enough of my teenage years on the internet to know that there's amazing cultures that can be put together purely remote like, I don't know, Wikipedia, World of Warcraft, Raiding Guilds can at least have excellent cultures. And so we are like, "Okay, cool. Let's come to a new stable part." That stable part was don't port the office online. Let's port the internet into a company and then we are like, "Let's go."

**Lenny Rachitsky** (00:45:00):
**Tobi Lütke** (00:46:46):
It's so funny that that's the way the story goes. It's right, but I don't know why that is surprising. That's my happy place, being able to clear out free days of my calendar and being there till midnight with all these remarkable people who join us on journey, just building stuff. It's like, I do my job so that that is the jobs that exist for other people. That's a job I actually want. That's the one I couldn't find because no one did a company for me. But at least back in most days, I think still not really.

**Lenny Rachitsky** (00:47:26):
And so I just love coding. It's one of the greatest, I don't know, hobbies and pursuits. I've done it for a very long time. I came across it very, very early. It fits my brain like a glove. I appreciate so much of the craft behind coding. I am a trained apprentice in... Sorry, I've apprenticed as a programmer in Germany, which has a dual education system that you can do such things. So I've been professionally in companies spending all day. And I really mean it, programming ever since I just turned 16. I think first day was just before I turned 16 when I started my apprenticeship.

And so I love it. I have a view on my screen left of you is a cursor right now, which is opened to a Juniper notebook where I'm working on some projection stuff that I'm playing with. I try to sanity-check as much as I can of what I get. I try to find... It's a game for me to find new insights in data. And I just think, I don't know, it was cool. Maybe this is the picture you talk about. There is a picture that someone captured, which is really fun where I'm on my laptop with a bunch of engineers and then Harley, so president, is on stage DJing. This was taken at 1:00 AM or something like this. It's a picture which actually really precious to two of us because it perfectly summarizes our relationship in a way which these happy accidents are wonderful and these long journeys, I've been doing Shopify for 20 years now, so I appreciate these artifacts. But yeah, so...

**Lenny Rachitsky** (00:49:32):
To be clear, the reason this is unusual is, and why someone told me this story, is most CEOs do not do this, don't just sit there and code along with the team. And the reason I thought this story was important is and the reason I think your first principle approach works is you're actually in the engineering details, similar to Elon if you think about it, of just in the weeds doing the thing, understanding how the thing works, not just coming up with ideas out of pontifications. And so I guess is there anything there of just how important it is, if you want to approach thinking from first principles, it is to be really close to the metal, to the bare metal?

**Tobi Lütke** (00:50:11):
Well, yeah. First principles thinking starts from, I think Elon puts it as physics, which I think is a little bit atoms-coded. I think just from the atomic building blocks really is a right starting point. Atomic building blocks are the computers we are using. Computers are our instruments. We use them to create the music that when people appreciate to receive in the form of software. So you've got to understand how they work, at least to work the way I do.

**Lenny Rachitsky** (00:50:48):
Now, is the way I work optimal? Of course not. This is my point about the aesthetics of how to behave, how to work. I probably don't conform to the traditional view of what the most of public companies should be like or even should spend their time, but I think I'm successful because I don't try to conform to anything other than what I've learned works. And so what I learned works is be in as many details as you can. Really, really, really just understand the stuff that we are making decisions about and be willing to don't put too much stake into the sunk cost fallacy. Try to inoculate business or your parts of a company from a sunk cost fallacy as much as possible because that allows you to just see better solutions and so on.

**Lenny Rachitsky** (00:51:45):
And I think what we were doing, if I remember right, is at hack dates is we were working through pros or cons of just merging all of Shopify into one huge monorepo and we are sketching out directory structures and tooling that we would need, and...

**Tobi Lütke** (00:52:00):
... tooling that we would need and... Again, monorepo, now for companies, it's a very much one of those door A, door B kind of things. It's a very consequential choice that is incorrect to go say yes to at a certain size, and then it becomes very correct in my mind to say yes to, but at that point it's an enormous amount of effort. So, it's a kind of thing that actually is something I'm uniquely positioned to be involved with because it's actually a business strategy thing as well. That's an investment, a very real investment to say, "Hey, let's change the way we are building system. Let's figure out what the best way is."

**Lenny Rachitsky** (00:52:47):
There's going to be change management. There's going to be some people that have very strong opinions on yes or no. And frankly, "Tobi said so.", helps. It also compresses a lot of time, because everyone knows the way I book. Everyone can come to me with better ideas about anything, and if you're right, I will change my mind. But I will hold my opinions very strongly until the point of being convinced that they're not the correct ideas. Maybe that's another aspect of this Tobi Tornado thing you talked about earlier, people do fight.

**Lenny Rachitsky** (00:53:27):
Again, the aesthetics of our times put a lot of stake into consistency. I remember various politicians losing campaigns because they were called flip-floppers at times, which I... Although, I'm more of a Maynard Keynes school of... "When the facts change, I change my opinion. What do you do, sir?" It sounds like a better OS to go by.

**Lenny Rachitsky** (00:53:52):
It's awesome, because I'm extracting more of your algorithm of how you think about stuff. A few things you just shared are, ( 1.), you need to be in the details. This is thinking about to be successful as a first principles thinker is (B.), you need to be in the details in your case code. In Elon's case, it's like, build the thing and be at the factory, sleeping on the floor. As you said, he's very Adams based. You're more digitally native. And then also, don't be so reliant on some costs. Like you said in this Tobi Tornado case, "I know you've been working on this project, we're going to kill it, because it's not going to work better if we do that now versus this. Just keep going, because we've been going." And then, coming back to the stuff you've shared previously is, analyze the path that you've been on, that the previous products and solutions have assumed the path dependence of them, and then don't look at what they've overfit potentially, that isn't correct. Love it. Okay. This actually is a good segue to something else that someone that you work with suggested I ask you, Farhan, your head of engineering. I asked him, "What's the best way to get a glimpse into Tobi's mind?" And it's actually along the lines of what you just described, which is, the question is just, what's the best way to disagree with you?

**Tobi Lütke** (00:55:02):
I think just disagree with me. I immediately love it, honestly. I really crave it. It's very funny. I know how this really surprises people. And I actually appreciate it even more because it requires courage, and frankly, I actually do find that... But it also makes me immediately trust the person more, partly because I, first of all, think they will do what they think is right rather than what is convenient. Agreeing with a group tends to be much more convenient. But more than that actually, that they're courageous enough to do it right then. I think courage is really, really rare. I found a lot more high IQ in industry than courage. I found a lot more maybe even genius than courage. I like that. I wish it wouldn't require that. This is why I try to be very inviting of it.

**Lenny Rachitsky** (00:56:08):
When someone disagrees with me, I tend to immediately stop and say, "Cool, let's figure out why there's disagreement." And it's almost never, I find, in the, "I just feel like we should do this differently." What I'm looking for is offer unstated foundational assumptions. What is our divergence point? Because you might be right. In fact, people often are, when it gets to this point, I found. Sometimes it's an unstated foundational assumption that I hold that is incorrect. People tell me and then I'm so glad we talked about it, because I will stop forever nagging on this thing because now I know we can't do it because of Sarbanes-Oxley or something like this. It's just like, "Well, cool." My mental model of Sarbanes-Oxley, which is regulation for public companies, is not perfect. I have not in-depth studied all the details for it, so I will not consult this in my mind when I am saying, hey, we should solve the task at hand in a certain way. So, this is very good.

**Lenny Rachitsky** (00:57:24):
How to disagree? I really like debate. I will play devil's advocate actively if everyone agrees on something. Again, especially if a proposal is something that feels like I could have predicted would be a proposal before it got into the meeting and there was nothing surprising in it, I will make myself an end boss of a level and just say, "I'm going to say this is just not that good. I think we could do way better." I want people to then argue more in-depth for the veracity of the decisions, and that leads to a form of disagreement. I think all these things end up building trust.

**Lenny Rachitsky** (00:58:08):
I like that that touches also on this idea of, again, maximizing potential, the potential of the teams, the potential of the employees

**Tobi Lütke** (00:58:14):
Because of course, the really important decisions, we don't talk about. This is the most important thing. Shopify probably makes... What?... millions of decisions every day, like, write this code this way, yes, I'm going to add this unit test, maybe I'm skipping a unit test. Might be the difference between a future production audit. Millions and millions of these tiny decisions. So, you're not hiring engineers primarily or accountants, you're hiring people who make excellent decisions, given their specialization and areas they're overseeing. And given that, decision making as a concept is actually really understudied, I find.

I think these are instances where we can just learn to make decision making together. Because I think while a lot of decisions are made independently, we are a product company, we are on a mission, and we want our product to feel like something that a single person made, in the same way how any author tries to write a book that clearly reads that it came from one mind. Because people can see and spot this, if this doesn't happen. You end up with something that looks like a television remote, where there's a Netflix button over here... It's like you can reverse [inaudible 00:59:41] from the remote control. That's important.

**Lenny Rachitsky** (00:59:48):
Using these moments and also bringing decision making inwards to go and say, hey, let's have very efficient opportunities for as many people that can possibly get together, make decisions together, not as a democracy but with clear who needs to convince who. That because otherwise this ends up just taking way too long. Give a opportunity for everyone to change my mind over Glenn's mind, and so on, is an excellent practice, I find. I think this is a more optimal way of going about it.

**Lenny Rachitsky** (01:00:22):
Speaking of disagreeing, it reminded me of a story I heard, about the Tobi Tornado actually, the way you operate that. I love it. It ends up being this interesting microcosm of your first principle's way of thinking. The story is, I think you've said at one point that the best way to get people to give you insights is to say something they disagree with on the internet.

**Tobi Lütke** (01:00:42):
Yeah.

**Lenny Rachitsky** (01:00:44):
That's often the way you approach this is, you post in a Slack group, "Hey, I don't think this product is going to work, and here's why." And that ends up creating the most information for you. Is there anything along those lines that might be helpful to share?

**Tobi Lütke** (01:00:56):
I don't think that's one of my better ways of doing that. I have to be somewhat careful. The people I worked with a lot, I will do this, just because it's funny. But even I can see that that would seriously stress out the interns.

**Lenny Rachitsky** (01:01:16):
You guys have a lot of interns. I think you have 1,000 interns this coming year, is what Farhan shared.

**Tobi Lütke** (01:01:20):
Yeah, we just started. I was in the office yesterday and it's absolutely full of interns. It's great.

**Lenny Rachitsky** (01:01:25):
Oh, man. Love interns, so much energy. This idea that you shared about building one product, building towards one vision, this actually reminds me of something that came up basically every time I had someone from Shopify on, which is this idea of a 100-year vision that you keep. I don't know of any other company that operates in this way, where the founder has this 100-year vision of where the product needs to go and working backwards from that.

**Lenny Rachitsky** (01:01:48):
Can you just speak to that, of that way of operating, why you find that helpful, how that actually works?

**Tobi Lütke** (01:01:54):
I talk about look in the future and then think backwards a lot. What would we want to have done 20 years ago on this, or 10 or 5 years ago? What's the decision our future selves would want us to make, is useful. I find future casting to be generally extremely valuable. I definitely did not build Shopify to flip. I had lots of opportunities to sell Shopify to various people. I didn't consider it, because it's just too interesting of a journey and I think that's too... If a endpoint [inaudible 01:02:34] all companies is a convergence on a set of four or five people who can afford them, that just creates too much of a monoculture, I think, in thinking.

**Lenny Rachitsky** (01:02:43):
I like that Shopify is different. I think this is good. I think it's a terrible place to work for many, many, many people. It's the best place in the world to work for some people. That's so good. That's what we want, I think. We want more of this. We want people to be able to window-shop for a place where they can be enormously successful, because the place's set of beliefs just fits you like a glove.

**Lenny Rachitsky** (01:03:14):
100 years. We have very long-term plans. 100 years, you can't talk about this software product, but you can talk about the mission itself, whatever things that will survive for 80 years that are left on this particular timeframe. Entrepreneurship is just precious. Shopify exists, basically you make entrepreneurship more common. That is the thing we wanted to cause in the world. We have had, I think, success doing this already, but again, there's no speed limit and no stopping point for this. As I keep saying, the word's unbelievably path dependent. And therefore, if we are part of a path, we can cause it to be more in adherence to the things that we value and would like to see. This is the wonderful thing about company building. They can have lasting impact.

We are making decisions based on all the things that we can do. What are the things that will be most long-term valuable for the long-term pursuits we are on, and how can we normalize entrepreneurship? It's a thing that's really important to me, just because people don't spend enough time on it. It's all economics. Our [inaudible 01:04:44] living entirely depends on businesses. We are all part and tag of an environment that is extremely entrepreneurial and actually celebrates entrepreneurship as a courageous act and a glorious act even.

But that's not true in most of the world. In fact, most people never encounter anyone who engages in doing that. It's not something people see as one of their options. I want us to make these choices. And I think the long-term focus matters. It's also really powerful for decision making. And I know this is potentially intuitive to most, but also rarely practiced. You've had a lot of people from Stripe on here. Stripe and Shopify have had a very long-term partnership. Stripe and Shopify had potentially the most valuable partnership, or at least one of the top ones in history, of technology, because we were both very small companies and we decided, hey, let's work on an assumption that both of us are going to win our markets, and work together. So, Stripe's [inaudible 01:06:02] for payments and we're allowed a part of Stripe and so on.

When you're in a partnership like this, you play basically iterated prisoner's dilemma. Every instance, every turn you can make a choice: coordinate or defect, defecting if the other person collaborates, collaborator effect. If both collaborate, everyone gets a point. If one defects and the other one collaborates, you get a lot of points, all at one moment. Being a good partner in business is like this corporate marshmallow test that companies tend to fail in a very funny way.

**Lenny Rachitsky** (01:06:44):
If you see the videos of kids doing actual actual marshmallow tests, the smart ones actually turn the chair around, look away from the marshmallow, and sit on their hands, and just go vibrating. Because they know it's the right thing to wait for getting two marshmallows in the future, but man, if they just look at a marshmallow, they'll just eat it. Most CEOs, most companies can't even successfully do that and just engage in pulling future profits forward at a discount or even defecting on partnerships that would be much more long-term valuable.

If you're talking about long timeframes, like 100 years, there is no question. But clearly, the correct way to play iterated prisoner's dilemma is coordinate for both sides. It's way more valuable [inaudible 01:07:27] doing that over long periods of time than any momentary defection could possibly be convenient at a moment. It's a huge amount of product decisions. When we are deciding roadmap, end up being very influenced by this, because I get pitches for things we should do and why don't we do this kind of change to the system, there's so much money in it, and all these kind of things. I'm like, cool, but almost always come in a form of putting future profits forward at a discount.

**Lenny Rachitsky** (01:08:08):
But we have a long time horizon. Let's not take the discount, let's... Also, you often compromise the entire business after a while because you end up just... Your customers notice if you're going into value extraction. The best companies, I believe, at least from a perspective of a leader, resemble some kind of cockpit, or maybe a room full of dials or levers for monetization. My job is adding as many of these levers as possible to the room, and then not pulling any of them. Because I think we do best if we are on the same side of a table of our customers and we had them become entrepreneurs, which is a mission, and we want to support them to be more successful, which is the business model we are in. We get a very, very small stake of the sales in our monetization system. And therefore, we are incentivized to make everyone as successful as possible.

**Lenny Rachitsky** (01:09:11):
Again, when I get a PowerPoint from investment bankers or so, where they tell me I have enormous pricing power and I could massively change the prices, I'm like, "Yeah, but I'd leave Shopify if that happens to me, so how is that good?" He's just like, "Why?" It just goes like that.

**Lenny Rachitsky** (01:09:30):
There's a quote that you wrote somewhere that, to me, identifies this point so succinctly, "On a long enough timeline, playing positive-sum games with your customers is the ultimate growth hack." Yeah, beautiful.

**Tobi Lütke** (01:09:46):
I think that's pretty piffy and it's also just... Try to argue with it.

**Lenny Rachitsky** (01:09:53):
That's true.

**Tobi Lütke** (01:09:54):
Positive-sum games have incredible returns, especially in the worlds of software, where you can experience exponentials very easily. I see company building as... I wish there would be better analogy than chess, because chess is a game of perfect information, which is totally incorrect. But one thing which I like about chess as an analogy for business is that it basically is two games, that you have to be good at at both. There's a positional game that you learn: Develop your pieces, gain influence by your pieces over the board. That's really, really important. And then there's tactics, which you learn tactic training. You go get a puzzle trainer and you drill tactics and you learn the intuition to sports tactics, sense them. Both of them are actually super independent of each other.

**Lenny Rachitsky** (01:10:53):
The business world only talks about tactics. It only talks about the conversion of we did this thing. We did an A/B test and we changed the color of blue, and conversion red went up. We lionize the easy hack. I don't think that's important, honestly. You need to be good enough at tactics to not go out of business. You can't get margin code, sure. But the sum total of all the value of a potential tactics that you could employ stays with you if you are actually are doing the positional game.

**Lenny Rachitsky** (01:11:36):
The positional game is like, what is the territory on the map that you are taking? What role do you play? How much trust do you have of merchants? Do merchants want more from you or less? Are you the kind of thing they're trying to optimize out of their software spend, or the one that they ask to subsume all other software spend? Do they rely on you? Are you part of a team, or are you used as a tool in the toolbox, usually forgotten, sometimes coming out when certain task is being done? What does your product cover? What industries are you addressing? And so on, so on, so on. This is the positional game. How well do pieces fit together? Do people like relying even deeper on you?

**Lenny Rachitsky** (01:12:25):
If you do that well, the tactics are yours and you can hire a lot of people who are extraordinarily good at spotting tactics and using them. But if you do it too much, you end up extracting through tactics the entirety of the value that you have created and that is yours to take through a positional game. And if you do that, you have nothing left in the tank. That's for companies that we all see, that just got to a point and then just fade. Those are the companies that got tacticked out of their position.

**Lenny Rachitsky** (01:13:01):
The skill, obviously, is finding the balance between these two things of short-term, not get margin calls, as you described, but also think ahead.

**Lenny Rachitsky** (01:13:09):
Is there any kind of heuristic for founders listening to this who are like, "How do I do this? How much should I be thinking about the future versus now?", I guess? How do you try to create this pie chart of driving goals immediately and show investors we're killing it, while also thinking ahead?

**Tobi Lütke** (01:13:22):
Is there a good heuristic? Objective number one is: Don't die. Again, in the end, for companies that become historic companies are the ones which did not die. It also sounds very basic, but actually it's more actionable than it might seem. Past that point, I would argue: Just focus on the positional game. I think this is where a bit of a discrepancy of the founder and the rest of the company sometimes lies. I think the founders create finite, winnable games for people, that are very much serving the infinite game of developing, that the mission implies, the quality of position on the board.

**Lenny Rachitsky** (01:14:14):
By the way, it's an infinite board. Just picture the chess board as the initial terrain, but there's, I don't know, fog on the board and it's a terrain that's much larger that you will explore over time. I think that's the most prettiest way of mentally thinking about the experience of building a company. It's an exploration and a collaborative inquiry into a question that is implied by the mission. And you will get to explore how correct your mission is and how good your decision making is along the way. And you get to learn a lot, and this is why I think it's a valuable thing to do.

But they often, the founders, and the mission of a company are in alignment, but they are, again, non-quantifiable things. They are pursuits. Maybe as time horizons that go past, all of our time horizons of our careers, therefore that makes it hard for people to care deeply yet. But they do care about the instances, the games along the way, the [inaudible 01:15:24] that have to be done to explore a part of a map, the tactics that have to be executed.

**Lenny Rachitsky** (01:15:29):
I'm not going to take us here, but I know you're a big fan of this book, Finite and Infinite Games. I actually had it once at a book club and it blew my mind open. I'm going to link to it. People should check it out. You've talked about this on another podcast. I'm not going to [inaudible 01:15:40].

**Tobi Lütke** (01:15:40):
Yeah, it's a lovely book. James Carse did an incredible, underappreciated... I think they essentially wrote one of the best business books. They is trying to write a hard philosophy book. I don't know if it works for its intended purpose, but I think there's more value in what he actually accomplished. Maybe, sadly, it may be that he didn't know before he died, which would be very sad.

**Lenny Rachitsky** (01:16:08):
It's a little hard to read for people...

**Tobi Lütke** (01:16:08):
Yes.

**Lenny Rachitsky** (01:16:09):
... so just stick with it and just try to wrap your head around what he's trying to say, is my advice.

**Tobi Lütke** (01:16:14):
I think reading... I don't know how many it is, but... the first couple of chapters really helps you get your arms around the story, his insight. The rest of the book are examples of his ideas applied. My takeaway from reading his examples is that he did not fully appreciate the insight of his own idea. It just ends up being very locally limited and very narrow-focused. It's a much grander idea than I think he really... This is why I say I don't think he fully appreciated quality of his own idea, which...

**Lenny Rachitsky** (01:16:58):
I think we need a Tobi version of this book, with a foreword, and...

**Tobi Lütke** (01:17:02):
I heard Simon Sinek...

**Lenny Rachitsky** (01:17:04):
Uh-huh.

**Tobi Lütke** (01:17:04):
... wrote a book of a similar title. I've actually never gotten around to read it, but I think he would be very good at interpreting this if that's what he did.

**Lenny Rachitsky** (01:17:11):
That's what it sounds like. I feel like his book is that book, written in a different way.

**Lenny Rachitsky** (01:17:15):
I want to come back to something you talked about, which is focusing on entrepreneurship and the merchants that you all work with. To me, this is another example of maximizing human potential. The way I think about this, people always talk about Y Combinator and Stanford and all these places that create all these companies and founders. If you think about it, Shopify does this orders of magnitude beyond, and I don't think you guys get enough credit for that, the amount of businesses you create, the amount of lives you change.

**Tobi Lütke** (01:17:43):
That's because we don't want the credit. That's the point.

**Lenny Rachitsky** (01:17:46):
Is there a platform?

**Tobi Lütke** (01:17:47):
We don't want the credit.

**Lenny Rachitsky** (01:17:47):
No.

**Tobi Lütke** (01:17:48):
I mean, maybe if you wanted it, you could also not get it, but we don't need to explore that because we actually, literally don't want the credit. Shopify is a company that pushes from behind. We don't want to be written into the story. We want to just...

**Tobi Lütke** (01:18:00):
... want to be written into the story, we want to just kind of have you do your thing. There's just too much grandeur in wanting to front it all. This is also what caused us to go and help people build their own things rather than start with a marketplace that everyone gets to lease a little component of at prices that corresponds to magically exactly your margin, as this usually goes in marketplaces. And so yeah, I think we don't want it. I'm proud of it, but that's intrinsic. I don't need very extrinsic appreciation for it, I suppose. There's millions, millions and millions of people that use Shopify daily and that represents the business and it causes major amount of employment around the world actually, definitely in North America specifically and Europe. And so I think that's really gratifying. I'm sure there's a way of saying that many of us businesses might actually have existed without Shopify.

**Lenny Rachitsky** (01:19:18):
I also think you can make a case that a good number exists because of Shopify, because what we've observed is that this was actually, it's my favorite thing in the entire Shopify journey that we sort of rigorously determined, which is that because I really, really, really prioritize good UX, legible interfaces. You can tame enormous complexity with great UX in a way that makes sense to people. I think this is actually almost a moral obligation to do for software because when software goes bad, it makes people feel dumb and machines do not get to negatively influence people in my mind. That is the inversion of the priority of the hierarchy of how machines are tools wielded by people to be more powerful at the things that they're already great at than they might imagine. That's when this all needs to fit together, but beyond the moral point, which is 10 years, and you could definitely take the other side of what we have determined is that every single time we make a complex thing simpler. It is actually that more businesses will exist on a platform.

**Lenny Rachitsky** (01:20:38):
So I think this is intuitive or at least directional, but sometimes people are not like, "Well, don't you just need to have all these features?" This is the RFP view of a world of software, right? "Don't you need to just have all the features then people can implement their plan?" Well, no. Think about the mental state and entrepreneurship entrepreneurs, they are, I mean especially if you're first entrepreneurs, they're unsophisticated. They don't know what to do. They're kind of scared. By the way, engaging in building something again is an act of pure courage and usually one which is very hard to hide from others. The people around you will know that you're doing this thing. Ask any entrepreneur. The amount of people who tell you to not do this is actually stunning. People do not want people to just step out of a box that exists that they sort of explain to themselves around them and have a chance of reaching higher because that would sort of invalidate their own life story in some meaningful way.

**Lenny Rachitsky** (01:21:38):
So there's actually a whole lot of nay saying that discourages people already, which is not super helpful. So now you have a situation where people are telling you to quit potentially, you don't have a lot of money presumably, you're asking yourself, are you even doing the right thing? Hopefully there's some elements of encouragement. Hopefully there's a passion. Hopefully you're trying to create a thing. Then comes to a situation that stuns you in some way, somewhere where suddenly the software starts talking about APIs where this is something you've never encountered in your life before or that just the option that you think it should have on the text configuration screen doesn't make sense given what we know about local taxes and so on. Here's the thing that happens. So the beautiful thing with Shopify is there's basically no tax configuration screen. It's just correct. The amazing thing about software is we can actually just overtake this piece of complexity.

**Lenny Rachitsky** (01:22:53):
We know what the taxes are everywhere and we are just doing it for you. You don't have to think about taxes. In this case we have someone encountered something that stunned them, that stopped them in their tracks. On the wrong day, what that means, it's like they are going to close a browser and they say, "You know what? Fuck it. I'm not cut out for this." Or, "This piece of software I use sucks and it's too dumb for understanding the thing I have in my head," or something else. Everyone has a different rendering of the same thing, but just like progress stops. It is not just the businesses that shouldn't exist anyway that stop in this way. In fact, it's actually, and this is again what the data says, so many businesses get very close to this point many, many times. I can tell you Shopify did many times and I have more tools on entrepreneurial journey than most. I'm a computer programmer and I love those things. You tell me I can spend 14 hours programming for the next couple of years. I'm like, "Holy shit, let's go."

**Lenny Rachitsky** (01:24:01):
Most people aren't like that. So I've had unfair advantages that allowed me to overcome very technical climbs on this learning curve, but I know that's like most people can't have these advantages and therefore will churn out of a test. And so in other words, lowering complexity, making good UX, creating software that just autopilots taxes or payments or any of these kinds of things, fraud, actually causes more entrepreneurship. That is the best thing, the best answer to the most important question of my life that I've encountered because up to this point it was an intuition that doing the Shopify thing would be valuable in absolute sense, in a hundred-year sense. Afterwards I knew, right? And so that also just means now seeing things that are overly complex or shouldn't even be there, it's just physically painful because I know, it's like all these business that are trying to get somewhere and could exist and could now employ people and delight customers just died along the way because of something we did wrong or poorly and so on. And so this is a great source of energy for me to keep going.

**Lenny Rachitsky** (01:25:26):
I feel that, I feel that pain of the idea of anytime a button's broken or it's not as... the button's wrong color and more people would be converting means that you're not creating as many businesses, many entrepreneurs. The feeling you described of being scared to launch something, I exactly felt that when I started the newsletter. And my solution to that as I launched it is, "I'm just launching this as an experiment. We'll see where it goes, just don't worry about it. I might blog once in a while." And that really helped just lowering the stakes and I could see how just things like that and advice that can help someone get over that hump.

**Tobi Lütke** (01:26:01):
Yes, and I think this is an under explored thing in the word of UX. Random example, I was absolutely delighted by, maybe TikTok has the same thing, I don't know, but I was posting a video to Instagram and it allowed me to run a try run of the reel. It just seems like a new feature and just saying it shows it to a couple hundred people out of network and if they like it then it's going to post to my profile. I'm like, "I don't need that." But I thought that was one of the more... I honestly just closed Instagram and was like, "Holy shit, this is probably one of the most profound insightful pieces of software I've encountered. And as a connoisseur of good ideas, I've never had this as a valuable thing." It feels like a step function upgrade to the traditional A-B test as a concept because it's so understandable and I'm like, "What else in the world should have try runs that run out of network?" Basically everything, isn't it?

**Lenny Rachitsky** (01:27:03):
This is an amazing thing. Like I said, the rarest thing in the world, it's not even creativity or genius, it's courage. So let's lower the net amount of courage needed. Honestly, that's one of those things you can probably run an entire career or reinvent an industry on. These things unfurl into incredible amounts of value if you really pursue them and come at them from first principles rather than how can I do the things I want to do anyway slightly better because of insight. That's good and people should always do it, but not usually where most of the value ends up becoming, like manifests when a new idea comes around. So I talk radically about an Instagram feature. That's a weird thing to do, but I was actually just really to delighted because again, it just came up a few days ago.

**Lenny Rachitsky** (01:27:54):
I love this thread of courage. The example you gave reminds me also of I feel like going on a dating app at a different city for the first time is another example of this where it's a dry run, nobody knows you.

**Tobi Lütke** (01:28:06):
But even then, shouldn't there be like, "Hey, post my profile to a different city for telling me if it's any good."

**Lenny Rachitsky** (01:28:14):
SO people like it.

**Tobi Lütke** (01:28:14):
I mean I've never dated basically so I mean at least not in the last 25 years.

**Lenny Rachitsky** (01:28:19):
Yeah, of the podcast.

**Tobi Lütke** (01:28:20):
So I'm pre swipe left and right dating, so my conception is very low of this. Maybe I should look more at dating apps. I'm pretty sure there's some amazing UX in most things.

**Lenny Rachitsky** (01:28:34):
I imagine. I feel like dating is going offline more and more. I'm feeling like that's what people are trying to do. They're tired of the swipe, but anyway, going to go in that direction. Maybe a couple more questions while I have you here. One is this idea of the talent stack. You've mentioned this concept before and this is the term that I've seen you describe it as of this idea of the power of focusing on your unique talents and curiosity and that leading you to the biggest opportunities, especially early in your career. Can you just talk about what your insight there.

**Tobi Lütke** (01:29:04):
Yeah, again, I only lived one life, so I can't Monte Carlo all the decisions I make and just figure out which ones ended up being load baring, right? I'm at this point pretty amused by the following thing which keeps happening. But I am getting curious about some absolutely random thing and it really is fairly far-flung stuff and like magic it becomes the way that ends up allowing me to make a very important choice a year later. It's ends up being a better analogy that I've learned or a different way to see something or another idea that I found being represented in this area of expertise that is actually just again, another repackaging of another foundational idea, which allows me to go and look for more examples.

**Lenny Rachitsky** (01:29:57):
And it's just funny this way. So even early in career just I followed my curiosity. I love programming and I love computers and I loved the internet when it came along and I just like, "Cool, I'm going to find a task that I find valuable," which I was always engaged in retail and I have a lot to say about how retail should be brought by the internet to people. But this is a beautiful intersection of all the things that I find interesting. And then on top of it, I found Ruby, which I loved as a technology and then just now I was highly motivated to tinker, explore a space that was clearly emerging but just felt very obviously of value in the future.

**Lenny Rachitsky** (01:30:44):
But I didn't do it because I was following money. I did it because I like learning by doing stuff and I like tinkering the things and so this was a way of financing my tinkering by selling snowboards and then it led to other things. And I've kind of been doing that thing all along. It's great fun, great fun for me. I don't know how much of a advice that is, but maybe a proof point that can work really well. I sometimes worry though that things like this end up being a little bit like all you need to do is buy a lottery ticket, set for lottery ticket winner, right? So it's, who knows?

**Lenny Rachitsky** (01:31:30):
It does resonate. It reminds me of something Brian Armstrong once shared, which is the reason that he-

**Tobi Lütke** (01:31:37):
He's also, by the way, you have to add him to the set of first principle thinkers if you're starting to cultivate one. And I think there's plenty of startup success, successful founders, CEOs, especially for public companies are surprisingly alike in this projecting slightly different, maybe coming from different backgrounds, but Brian is extraordinarily strong at this [inaudible 01:32:06] by the way.

**Lenny Rachitsky** (01:32:06):
Okay, here's to the next nominated for the podcast, Brian. We'll get on it. But interestingly, he had basically exact same advice for what allowed him to create Coinbase is his background was economics, coding and cryptography or something along those lines. And it's the Venn diagram, is like, "This is the thing I'm uniquely strong at and have an opportunity to win it."

**Tobi Lütke** (01:32:29):
Yeah. So this is what we very actively tell our customers, right? There's a 2005 essay by Kevin Kelly saying, "In the future on the internet you just need a thousand true fans." It's what the internet and I think what Shopify celebrates, that instead of trying to create toothpaste, which is of course a huge term but hard to differentiate. It's much better to figure out a triplicate intersection of three different things and nail it. We took a winter vacation with my boys and my family in the Caribbean and I like playing poker with them. I like playing poker and it's great limited information game. I think this were valuable for kids and so I found this, I might have mentioned this in some place, but I found this pirate gold I think it's called.

**Lenny Rachitsky** (01:33:32):
It's just amazing poker set, poker chips that look like they come right out of the treasure chest or in Pirates of the Caribbean. Think about just the intersection, so this is online Texas Hold'em, quality poker set instead of... because you can play with just anything, right? People were in the market for spending money on something like this and pirate themed. You're like, " But I'm such a big fan, I'm telling you about it and people might actually purchase it because I'm enthusiastic about it." And because enthusiasm is actually the best marketing and they have at least a thousand true fans, and it's just better to make amazing things for some people than make something that everyone wants maybe or would tolerate. And I think that's really, really good. So I think that works for your career too because by the way, we are all entrepreneurs so we're ready to think about any career.

**Lenny Rachitsky** (01:34:36):
If you actually ask me for career advice is we are products, we are engaged in entrepreneurship, we are basically software as a service if you want, or talent as a service. We are selling a subscription in the form of our employment we call it, but it's actually just rebranded the subscription frankly because it comes some extra protection, which is very good, but it's still the same thing, it's a double opt-in company. Like you would say, I can do these things and I will be of significantly more value to the company than you will have to spend on me. There's like a positive ROI ideally for a company and a company either says, "I agree or I don't agree with you." And that's actually what employment is. So I think people should think about themselves as a product. Your career is not based on mentors you find or getting a promotion or not.

**Lenny Rachitsky** (01:35:34):
It's about what are you too good to ignore in? How good can you be? Can you be absolutely work? Can you be excellent? I mean you don't have to be work class and stuff, but you have to... Ideally you find a thing where you are just in this sort of tiny space of five intersections or so versus we just know more than everyone else and then everyone calls you. Like I said, we use Ruby. Ruby was not that fast. Shopify would rather have it very fast. We found people who got PhDs in dynamic language, garbage collection dynamic language, just in time compilation to make Ruby fast. And they're having delightful time at Shopify and Shopify is now very fast and it's very good. And everyone gets to profit from us needing this because we merged it back into Ruby Core and now everyone can have a JIT compiler. So it's great.

**Lenny Rachitsky** (01:36:40):
I feel like every story is a fractal into other hours of other podcast conversation we can have. I know you have to go run this nine, 10, 10,000 person company and do real work. So I'm going to end it there. We got through almost everything I was hoping to get through, but there's infinitely more questions I still have. Just to maybe end it, is there anything you want to leave listeners with? Any last nugget that maybe you didn't mention?

**Tobi Lütke** (01:37:03):
If I get to talk to again people who look at product and product management, here's my thing that I think maybe only implicitly came through in all this. Every product in the world, the quality at the end of the day is simply a reflection of how much the people who created it gave a shit about the product. And it is not possible to make great products if the people work on it do not give a shit about the product. And I actually think this is a very important role for product leaders to make sure that the team gives a shit. And I think this is something that can be done with building empathy for people using it, but also it can be done infectiously by the product leader. The product leader has to give a shit. Do not engage in product work on product that you don't care about because you cannot produce the thing that the person will give you a task is looking for.

**Lenny Rachitsky** (01:38:04):
And I think again, so much goes into this unquantifiable conversation which we already had, but boy is that important because it just isn't about the brief document, it isn't about the aligning stakeholders. Sometimes in some places some of those things are aspects of it and I think they correlate sometimes with people. But I think the higher order bit is if you are in product, you have basically free roads. You have to have the team sits, look around corners that they don't see. You just have to understand this thing that's being done better than everyone else because that's a role. Now, you don't have to do this by yourself. Everyone on the team is of a resource to you and you can query them and therefore determine what is around the corners and so on. But in engineering and UX and so on, but that's very important. Second is you've got to be exothermically infectious with actually caring about this thing because just that one thing alone will make a 10 times better product. It's crazy how much of a change this makes.

**Lenny Rachitsky** (01:39:25):
I feel like honestly what you're describing here is like founder mode, what the outputs are. Giving a shit, being exothermic excited and looking around corners. I feel like this is its own conversation here.

**Tobi Lütke** (01:39:37):
Yeah, I agree. I don't think founder mode, so I mean I think, yeah, founder mode is a super valuable term, but I actually do think that I answered the question about what I think about founder mode in the discussion about pumping heat into systems. That's what I do. It's just that founder mode in a founder run company, I think this has an easier time existing because people dislike the people who are pumping heat into a thing. People fight them, people probably take them out, and if the boss is like this, boss can protect the other people like this. And therefore I think founder run companies can be innovative longer because there's a substrate by which this can stay a thing for a long time. So anyway, that's my tear down on the founder mode discussion, which I think is a really fascinating recent discussion. I find it's very valuable. I'm sure it's misunderstood in all sorts of ways, but I think it's more actionable than people think.

**Lenny Rachitsky** (01:40:40):
Tobi, I appreciate you continuing to drop nuggets, even though I know you have to go, by the way, your video looks incredible right now. This is amazing editing, we're doing like a rainbow streaming across your body.

**Tobi Lütke** (01:40:53):
Yeah, we really hit the rainbows here like, "Let's go and invent." I hope you've got to gold that you were hoping and-

**Lenny Rachitsky** (01:41:00):
So much gold.

**Tobi Lütke** (01:41:00):
... I was sufficient leprechaun here to deliver.

**Lenny Rachitsky** (01:41:05):
What a metaphor. Okay, Tobi, thank you so much for doing this. This was incredible. I feel like this is going to help so many people build great companies and thank you from first principles.

**Tobi Lütke** (01:41:13):
Okay, I got to run. Bye-bye.

**Lenny Rachitsky** (01:41:15):
Okay, bye.

**Lenny Rachitsky** (01:41:18):
Thank you so much for listening. If you found this valuable, you can subscribe to the show on Apple Podcasts, Spotify, or your favorite podcast app. Also, please consider giving us a rating or leaving a review as that really helps other listeners find the podcast. You can find all past episodes or learn more about the show at lennyspodcast.com. See you in the next episode.

---

## [7/15] OpenAI researcher on why soft skills are the future of work | Karina Nguyen (Research at OpenAI, ex-Anthropic)

**Lenny Rachitsky** (00:00:00):
Not only are you working at the cutting edge of AI and LLMs, you're actually building the cutting edge.

**Karina Nguyen** (00:00:06):
When I first came to Anthropic and I was like, "Oh my God, I really love front-end engineering." And then the reason why I switched to research is because I realized, "Oh my God, Claude is getting better at front-end. Claude is getting better at coding. I think Claude can develop new apps."

**Lenny Rachitsky** (00:00:20):
What skills do you think will be most valuable going forward for product teams, in particular?

**Karina Nguyen** (00:00:26):
Creative thinking and you kind of want to generate a bunch of ideas and filter through them and not just build the best product experience. I think it's actually really, really hard to teach the model how to be aesthetic or really good visual design or how to be extremely creative in the way they write.

**Lenny Rachitsky** (00:00:42):
What do you think people most misunderstand about how models are created?

**Karina Nguyen** (00:00:46):
When you taught the model, some of the self-knowledge of you actually don't have a physical body to operate in the physical world, the model would get extremely confused.

**Lenny Rachitsky** (00:00:58):
Today my guest is Karina Nguyen. Karina is an AI researcher at OpenAI where she helped build Canvas, tasks, the o1 chain-of-thought model and more. Prior to OpenAI, she was at Anthropic where she led work on post-training and evaluation for the Claude 3 models, built a document upload feature with 100K context windows and so much more. She was also an engineer at New York Times, was a designer at Dropbox and at Square. It's very rare to get a glimpse into how someone working on the bleeding edge of AI and LLMs operates and how they think about where things are heading. Canvas

**Christina Cacioppo** (00:03:22):
Great to be here. Big fan of the podcast and the newsletter.

**Lenny Rachitsky** (00:03:25):
Vanta is a longtime sponsor of the show, but for some of our newer listeners, what does Vanta do and who is it for?

**Christina Cacioppo** (00:03:32):
Sure. So we started Vanta in 2018. Focused on founders, helping them start to build out their security programs and get credit for all of that hard security work with compliance certifications like SOC 2 or ISO 27001 today, we currently help over 9,000 companies, including some startup household names like Atlassian, Ramp and LangChain start and scale their security programs and ultimately build trust by automating compliance, centralizing GRC, and accelerating security reviews.

**Lenny Rachitsky** (00:04:04):
That is awesome. I know from experience that these things take a lot of time and a lot of resources and nobody wants to spend time doing this.

**Christina Cacioppo** (00:04:10):
That is very much our experience, but before the company and to some extent during it. But the idea is with automation, with AI, with software, we are helping customers build trust with prospects and customers in an efficient way. And our joke, we started this compliance company so you don't have to.

**Lenny Rachitsky** (00:04:26):
We appreciate you for doing that. And you have a special discount for listeners, they can get $1,000 off Vanta at Vanta.com/Lenny, that's V-A-N-T-A.com/Lenny for $1,000 off Vanta. Thanks for that, Christina.

**Christina Cacioppo** (00:04:41):
Thank you.

**Lenny Rachitsky** (00:04:45):
Karina, thank you so much for being here. Welcome to the podcast.

**Karina Nguyen** (00:04:48):
Thank you so much, Lenny, for inviting me.

**Lenny Rachitsky** (00:04:50):
I'm very excited to have you here because not only are you working at the cutting edge of AI and LLMs, you're actually building the cutting edge of AI and LLMs. You recently launched this feature, which basically... the first agent feature of OpenAI. I also just did this survey, I don't know if you know about this. I did a survey of my readers and asked them what tools do you use every day in your work and most use? And ChatGPT was number one, above Gmail, above Slack, above anything else. 90% of people said they use ChatGPT regularly. 

**Karina Nguyen** (00:05:23):
That's quite good. 

**Lenny Rachitsky** (00:05:23):
It's absurd. It wasn't around two years ago.

**Karina Nguyen** (00:05:25):
Yeah.

**Lenny Rachitsky** (00:05:26):
Also, we're recording this the week that OpenAI announced Stargate, which is this half trillion dollar investment in AI infrastructure. So there's just a lot happening constantly in AI and you have a really unique glimpse into how things are working, where things are going, how work gets done. So I have a lot of questions for you. I want to talk about how you operate and how you work at OpenAI, where you think things are going, what skills are going to matter more and less in the future, and also just where things are going broadly. So how does that sound?

**Karina Nguyen** (00:05:55):
Sounds great. Thank you so much. Yeah, I was extremely lucky to join early days Anthropic and learned a lot of things there. And I joined OpenAI around eight months ago. So, yeah, I'm excited to dive more in into-

**Lenny Rachitsky** (00:06:11):
Okay, I'm going to definitely ask you about the differences between those, but I want to start more technical and just dive right in. I want to talk about model training. People always hear about models being trained, these big models, how much data takes, how long it takes, how much money toss it takes, how we're running out of data, which I want to talk about. Let me just ask you this question. What do you think people most misunderstand about how models are created?

**Karina Nguyen** (00:06:36):
Model training is more an art than a science. And in a lot of ways we, as model trainers, think a lot about data quality. It's one of the most important things in model training is like how do you ensure the highest quality data for certain interaction model behavior that you want to create? But the way you debug models is actually very similar the way you debug software. So one of the things that I've learned early days at Anthropic was we've discovered especially this Claude 3 training, when you taught the model some of the soft-knowledge of, "Hey, you actually don't have a physical body to operate in the physical world." But then at the same time you had data that taught the model some of the function calls, which is like, "This is how you set the alarm."

**Karina Nguyen** (00:07:30):
And so the model would get extremely confused about whether it can set an alarm, but it doesn't have a body in the physical world. So it's like the model gets confused and sometimes it'll over accuse. So sometimes it says, "Look, I don't know. Sorry, I cannot help you." And so there is always a balance trade off between how do you make the model to be more helpful for users, but also not being harmful in other scenarios. And so it's always about how do you make the model more robust and operate across a variety of diverse scenarios.

**Lenny Rachitsky** (00:08:09):
That is so funny. I never thought about that. Most of the data that it's trained on is kind of assuming it's like a human describing the world and how they operate. It assumes there's a body and you could do things, and the model is told you don't have a body.

**Karina Nguyen** (00:08:20):
Yeah.

**Lenny Rachitsky** (00:08:21):
Okay. I want to talk a little bit about data while we're on this topic. I know you have strong opinions here. There's this meme that models are going to stop getting smarter because they're running out of data. They're trained in a large part on the internet and there's only one internet and they've already been trained on it, what more can you show them about the world? And there's this trend of synthetic data, this term synthetic data. What is synthetic data? Why do you think it's important? Do you think it's going to work?

**Karina Nguyen** (00:08:47):
I think there are two questions here. We can unpack one at a time. But people say we are hitting the data wall. I think people think more in the terms of pre-trained large models that are trained on the entire internet to predict the next token. But what actually the model is learning during that process is actually how do you compress the compression algorithm here? The model learns to compress a lot of knowledge and it learns how to model the world. So the next prediction of the word, like, "Teach me how to drive," basically. And you only have a few words that will match that, a car. So the model actually learns about the world in itself. So it's like it's modeling human behavior, sometimes it's modeling... And when you talk to pre-trained models which are very, very large, they're actually extremely diverse and extremely creative because you can talk to almost any Reddit user through a pre-trained model.

**Karina Nguyen** (00:09:56):
But I think what's happening right now with new paradigm of o1 series is that the scaling in post-training itself is not hitting the wall. And that's because basically we went from raw data sets from pre-trained models to infinite amount of tasks that you can teach the model in the post-training world via reinforcement learning. So any task, for example, how to search the web, how to use the computer, how to write, wow, all sorts of tasks that you trying to teach the model all the different skills. And that's why we're saying there's no data wall or whatever, because there will be infinite amount of tasks and that's how the model becomes extremely super intelligent. And we are actually getting saturated in all benchmarks. 

**Karina Nguyen** (00:10:52):
So I think the bottleneck is actually in evaluations that we don't have all the frontier, like evals like, I don't know, GPQA, which is a Google-proof question answering, PhD level intelligence. The benchmark is getting to, I don't know, more than 60, 70%, which is what PhD gets. So it's literally hitting the wall in like evals.

**Lenny Rachitsky** (00:11:19):
I want to follow both those threads. So the first is on this idea of synthetic data. Is a simple way to understand it, that the models are generating the data that future models are trained on and you ask it to generate all these ways of doing stuff, all these tasks as you described, and then the newer models trained on this data that the previous model generated?

**Karina Nguyen** (00:11:39):
Some tasks are synthetically curated. So this is an active research area is how can you synthetically construct new tasks with models to learn. Sometimes when you develop products, you get a lot of data from the product and user feedback and you can use that data too in this cross-training world. Sometimes you still want to use human data because actually some of the tasks can be really, really hard to teach. Experts only know certain knowledge about some chemicals or biological knowledge, so you actually need to tap into the experts' knowledge a lot. So yeah, I think to me synthetic data training is more for product... It's a rapid model iteration for similar product outcomes. And we can dive more into it, but the way we made Canvas and tasks and new product features for ChatGPT was mostly done by synthetic training.

**Lenny Rachitsky** (00:12:52):
Let's actually get into that. That's really interesting. I want to talk about evals, but let's follow that thread. So talk about how this helped you create Canvas.

**Karina Nguyen** (00:12:56):
So when I first came to OpenAI, I really had this idea of, "Okay, it would be really cool for ChatGPT to actually change the visual interface but also change the way it is with people." So going from being a chatbot to more of a collaborative agent, and the collaborator is a step towards more genetic systems that become innovators ultimately. And so the entire team of applied engineers, designers, products, research got formed in the air almost out of nothing. It's just like a collection of people who just got together and we rapidly started iterating with each other. 

**Karina Nguyen** (00:13:46):
Actually Canvas is one of the... I would say the first project at OpenAI, where researchers and applying engineers started working together from the very beginning of the product development cycle. And I think there's a lot of things that we have learned on the way, but I definitely came with the mindset of, "We need to do a really rapid model situation such that it would be much easier for engineers to work with the latest model possible, but also learn from user feedback or early internal dog food. How do we improve the model very rapidly?"

**Karina Nguyen** (00:14:28):
And it's really hard to kind of like figure out how people... when you deploy a product, how people would be able to use it. And so the way you synthetically train the model is physically figuring out what are the most core behaviors that you wanted the product feature to do. And for Canvas, for example, it came down to three main behaviors. It was how do you trigger Canvas for prompts like, "Write me a long essay," when the user intention is mostly iterating over long documents? Or, "Write me a piece of code," or when to not trigger Canvas for prompts like, "Can you tell me more about President..." I don't know, some of the general questions. So you don't want to trigger Canvas because the user intention is mostly getting answer, not necessarily iterate over the long document. 

**Karina Nguyen** (00:15:28):
The second behavior is how do we teach the model to update the document when the user asks? So one of the behaviors that we taught the model is actually have some agency and autonomy to literally go to the document and select specific sections and either delete it or edit, so highlight it and rewrite certain sections. Sometimes the user would just say, "Change the second paragraph to be something friendlier," and we would have to teach the model to literally find the second paragraph in the document and change it to a friendly tone. So basically you teach both how to trigger edit itself, but also how do you teach the model to get higher quality edit for the document?

**Karina Nguyen** (00:16:21):
In case of coding, for example, there's also the question of how good the model is of completely rewriting the document, versus having a very specific target edits. So that's another layer of decision boundary within edit itself is, "Let's select the entire document and rewrite completely, or do you want to have a very targeted custom behavior." And when we first launched the model, we would bias the model towards more rewrites because we saw the quality of the rewrites were much higher. But over time you are shifting based on user feedback and what you're learning from iterative deployment.

Lastly, the third behavior that we taught synthetically the model is how to make comments on any document. So the way we used that is we would use o1 model to seem a way of user conversation, let's say like, "Write me a document about XYZ." But then we used o1 to produce the document and then we injected user prompt to be like, "Oh, make some comments, critique my piece of writing or critique this piece of writing that you just made." And then we taught the model to make comments on the document on very specific [inaudible 00:17:45] So it's also what kind of comments you want the model to make. Do they make sense or not? How do you teach the quality of that? And it all came down to measuring progress via very robust evals. But, yeah, this is how you used o1 and a synthetic data generation for the training.

**Lenny Rachitsky** (00:18:07):
Okay, that's so interesting. So you talk about this idea of teaching the model and you mentioned how it's using synthetic data to teach the model different behaviors is a simple way to think about it. Basically that's where you do that by showing it what success looks like using basically evals. Is that the simple way to think about it? Like, "Here's what you doing this successfully would look like," and that teaches it, "Okay, I see this is what I should be doing [inaudible 00:18:31]"

**Karina Nguyen** (00:18:30):
Yeah, great. Yeah, amazing. Yeah, you got it.

**Lenny Rachitsky** (00:18:33):
Okay, got it. I want to start unpacking what your day-to-day looks like as you're building these sort of things. Is it like you sitting there talking to some version of ChatGPT, crafting these evals?

**Karina Nguyen** (00:18:44):
Sometimes I do that. Sometimes I do sit with ChatGPT. Actually, I think I learned this so much from Anthropic, is people spend so much time prompting models and where quality's a really bad batch all the time, and you actually get a lot of new ideas of how do you make the model better? It's like, "This response is kind of weird. Why's it doing this?" And you start debugging or something, or you start figuring out new methods of how do you teach the model to respond in the different way, have better personality, let's say. 

So it's the same thing of how personality is made in the models with those. It's very similar methods. But, yes, I think my time at OpenAI have changed. I think when I first came, I was mostly research IC work so I was like building a lot of... I was running code, training models, write evals, working with PMs and designers to learn, teach them how to even think about evaluation. I think that was really cool experience and I think it was just like an adoption of, "How do we do this product management of AI feature for our AI models?" Yeah, but now it's mostly management and mentorship. I'm still doing IC research code up to 4:00 PM, although. But I just kind of changed. 

**Lenny Rachitsky** (00:20:21):
All right, don't talk too much about being a manager.

**Karina Nguyen** (00:20:23):
Okay. 

**Lenny Rachitsky** (00:20:23):
Because everyone's in firing their managers. "Who needs managers anymore?" That's what I hear now. Just kidding. It's interesting that so much of your time was spent on teaching product teams how evals integrate and how important it is. And I've heard this a few times and I haven't personally experienced it yet, so I think it's an important thread to follow is just how writing these evaluations is going to become increasingly an important part of the job of product teams, especially when they're building AI features and working with LLMs. So can you just talk a bit more about what that looks like? Is it sitting there with an Excel spreadsheet basically showing, "Here's the input, here's the output, here's how good the result was"? Talk about what that actually looks like very practically.

**Karina Nguyen** (00:21:02):
It certainly depends on what you're developing, but there are various types of evaluations. Sometimes I do ask product managers, or there's also new roles that we have, model designers, to go through some of the user feedback maybe or think of various user conversations that should have triggered... Under these circumstances, it should trigger Canvas. And then you have this ground truth label of, "Okay with this conversation it should look trigger Canvas, under this conversation it should not trigger Canvas." And you have this very deterministic kind of eval that for decision-making behaviors is like this.

When we were launching tasks, for example, how do you make correct schedules is actually really hard for the model. But we built out some of the deterministic evaluations that is like, "Okay, if the user says 7:00 PM, the model should say 7:00 PM." So if you can have deterministic evals whether it's pass or fail. And the way it works is all the... Sometimes I ask product managers to just go create a double sheet, have different tabs and what's the current behavior, what's the ideal behavior and why, and some notes. 

**Karina Nguyen** (00:22:27):
And sometimes they usually use it with evals, sometimes we use it for training. Because if you give the spreadsheet to o1 model, it can probably figure out how to teach itself a good behavior. And I think there are second type of evals that is more prevalent is human evaluations. And you can have specific trainers or you can have internal people to when you have a conversation of the prompt and then you have various completion of models, you choose the win rate. Which model is the best? Which model produce the highest quality comment or edit? And then you can have continuous win rates. And as you develop new models it should always win over the previous models. So it depends on what you want to measure.

**Lenny Rachitsky** (00:23:22):
So interesting. Basically what I'm hearing, and there's something I'm learning about as I talk to people, is product development might move from this, "Here's a spec PRD, let's build it together and then cool, let's review it. Are we happy with this?" From that to, "Hey, AI, build this thing for me and here's what correct looks like," and I'm spending all my time on what does correct look like on evals essentially. 

**Karina Nguyen** (00:23:47):
You definitely want to measure progress of your model and this is where evals is, is because you can have prompted model as a baseline already. And the most robust evals is the one where prompted baselines get the lowest score or something. And then because then you know if you're trained a good model, then it should just hill climb on that eval all the time, while not also regressing on other intelligence evals. That's what I'm saying, it's more of an art than science. It's like, "Okay, if you optimize the model for this behavior, you don't want to brain damage in other areas of intelligence or..." This is happening all the time in every lab, in every research team. 

**Karina Nguyen** (00:24:35):
I would say prompting is also a way to prototype new product ideas. Early days at Anthropic when I was working file uploads feature, I remember I was just prompting the model to just... I remember we were launching a hundred key contexts. I was just prototyping this in their local browser. I did the demo. People really, really loved it. And they just wanted API for file uploads or something. And then that's when it clicked to me, and also one of the blog posts a long time ago, it clicked on me prompting is a new way of product development or prototyping for designers and for product managers. 

**Karina Nguyen** (00:25:20):
For example, one of the features that I want to do is have a personalized starter prompts. So whenever you come to Claude, it should recommend you starter prompts based on what your interests are. And so you can literally do it prompting for that. 

**Lenny Rachitsky** (00:25:42):
Mm-hmm. To experiment with that. 

**Karina Nguyen** (00:25:44):
Another feature was generating titles for the conversations. It's a very small micro experience but I'm really proud of. The way we did that was we took five latest conversation from the model, asked the model, "What's the style of the user?" And then for the next new conversation, the generated title will be of the same style. It's just like really little micro experiences like this. 

**Lenny Rachitsky** (00:26:12):
That's so cool. Did you do that at Anthropic or at OpenAI?

**Karina Nguyen** (00:26:14):
At Anthropic.

**Lenny Rachitsky** (00:26:16):
Okay, cool. I love the file upload feature that Claude has by the way. ChatGPT doesn't have that yet, is that right?

**Karina Nguyen** (00:26:16):
I think has the way. 

**Lenny Rachitsky** (00:26:23):
[inaudible 00:26:23]

**Karina Nguyen** (00:26:22):
I think the way it's implement is very different though.

**Lenny Rachitsky** (00:26:25):
Okay. Maybe it's the PDF feature, because I use it all the time with Claude.

**Karina Nguyen** (00:26:28):
Yeah. 

**Lenny Rachitsky** (00:26:28):
Okay.

**Karina Nguyen** (00:26:28):
That's cool.

**Lenny Rachitsky** (00:26:29):
Somebody needs to get on that. Main, it's wild how many features you built that I use every day and that many people use every day. This prototyping point you made is really important. It's something that comes up a ton on this podcast also of how that... is maybe the way that AI has most impacted the job of product builders recently is just prototyping instead of going from showing just like, "Here's a PRD, here's a design." PMs are more and more just, "Here's the prototype with the idea that I have," and it's working. You can play with it.

**Karina Nguyen** (00:26:54):
Yeah.

**Lenny Rachitsky** (00:26:55):
Yeah. Okay, I want to spend a little more time on how you operate. So you talked about you built this in launch of this tasks feature, is that the way to describe your tasks?

**Karina Nguyen** (00:27:04):
Yeah. 

**Lenny Rachitsky** (00:27:06):
So talk about how that emerged and let's better understand just how you collaborate with product teams and how OpenAI works in that way, whatever you can share there.

**Karina Nguyen** (00:27:14):
I think Canvas and tasks are going into the bucket of projects where it's more short or medium terms. And actually the way Canvas and tasks came about to be was it started with one person prototyping and creating a spec. It's kind of like PRD. It's like creating a spec of the behavior of the model. I don't think tasks is extremely groundbreaking feature necessarily. What makes it really cool is because the models are so general... Model can now search, they can write sci-fi stories, they can search for stocks, they can summarize the news every day. Because the models are so general giving something familiar to people that notifications is very familiar, having reminders is very familiar. So feeling like a form factor for the people who are very familiar, same as Canvas, Google Docs is very familiar, but then you add magical AI moment and it becomes very powerful.

But the way it comes usually operationally... Yeah, size is like a prototype, literally prompted prototype of how you would want the model to behave. For tasks, for example, you need to design... Literally design thinking is like okay, well, if the user says, "Remind me to go to lunch at 8:00 AM tomorrow," what information does the model need to extract from that prompt in order to create a reminder? And so this is how you design a spec for a new feature, like a tool. Canvas and tasks are all tools. So it's like how do you create the tool stack?

**Karina Nguyen** (00:29:09):
And then it's mostly like developing JSON schema. It was like, "Okay, from this problem maybe the model should extract the time that the user requested." And then you think about which format do you want the time to be? And then how do you want the model to notify you is basically the user should give instruction to the model. And then this instruction would fire off every day or something at that particular time. So, for example, if you say, "Every day I want to learn know about the latest AI news," the model should rewrite into, "Okay search for the latest AI news and this task will get fired at that particular type that the user requested."

And then your design is like tool spec. Actually, I don't know. I feel like sometimes it's through conversations I... Either people ask me to join the [inaudible 00:30:15] team and they're like, "Oh my god, we need researchers." Or like, "We need some support. We need to train the models," or sometimes. Canvas was mostly like I just pitched the idea of... It got staffed quite immediately during the break, so it's dependent on the project. And then usually with staffing is mostly a product manager, model designer, actual product designer, a couple of researchers and a bunch of applied engineers. Depends on the complexity of a project. And then for tasks it took, I don't know, like two months or so to go from zero to one basically.

**Lenny Rachitsky** (00:30:54):
Oh wow.

**Karina Nguyen** (00:30:54):
For Canvas this was like four, five months, I guess, to go from zero to one. And then you teach product managers how to build evals and maybe how do we not only ship the better feature, but how do we think longer term? What kind of cool features did you want tasks to have? I think it would be nice for tasks to be a little bit more personalized. It'd be nice to have to create tasks via voice on a mobile, right? This is how you get research roadmap right here is thinking how the feature will be developed in the future. 

**Karina Nguyen** (00:31:39):
And then from there it's like you start getting data sets. With evals, you want to make sure that goes well. And then you need to have a trade-off between what methods you want to use. And the reason why I really love relying purely on synthetic data instead of collecting data from humans is because it's much more scalable, it's cheap, less than half. You literally sample from the model and you teach the core behaviors of the models and that will generalize to all sorts of diverse coverage.

**Karina Nguyen** (00:32:15):
And when you launch the beta feature, you learn so much from the users that you can... All your synthetic sets can be shifted in the distribution and how the users behave on the product behavior. And this is how we improve. And this is what happened with Canvass too when we launched from beta to GA. 

**Lenny Rachitsky** (00:32:34):
Okay. 

**Karina Nguyen** (00:33:48):
So the project that I described are mostly product-oriented. Research is mostly product research. Another component of my team is actually more longer term exploratory projects. And it's more about developing new methods, understanding those methods under a variety of circumstances. So basically developing methods, you need to follow very similar recipe of building evals but it's much more sophisticated evals. You want to have outer distribution or if you want to measure generalization, you need to capture that. 

**Karina Nguyen** (00:34:26):
But it is basically more sciencey in a way where... If we talk about synthetic data, one of the hardest things about synthetic data is how do you make it more diverse? Diversity in synthetic data is one of the most important questions right now. And so it's like exploring ways to inject diversity as a general method that will work for all is one of the research explorations. Other ones is more developing new capabilities. I feel like it's always about you work on this new method and you have signs of life that it's working, either you think of how do you make it more general or you think of how do you make it very useful? And this is how the longer-term projects become more medium, short-term project.

**Lenny Rachitsky** (00:35:15):
That makes sense. Essentially working on developing ways to make the model smarter, o4, o5, o6. New ways to... o1 was a big breakthrough, right? 

**Karina Nguyen** (00:35:25):
Yeah. 

**Lenny Rachitsky** (00:35:25):
The way it operates where it's not just, "Here's your answer," it actually thinks and takes time to think through the process of coming up with an answer. Okay. 

**Karina Nguyen** (00:35:33):
Yeah. 

**Lenny Rachitsky** (00:35:34):
Very helpful. Speaking of that, of thinking about the future, where things are going, I want to spend some time on just this insight that basically you are building the cutting edge of AI, at the very bleeding edge of where AI is going and where it is. And so I'm very curious to hear just your take on how you think things are going to change in the world and how people work based on where you see things are going. And I know it's a broad question, but let's say in the next three years, how do you see the world changing? How do you see people's way of working changing?

**Karina Nguyen** (00:36:08):
It's a very humbling experience to be in both labs, I guess. To me when I first came to Anthropic and I was like, "Oh no, I really love front-end engineering." And then the reason why I switched to research is because I realized at that time it's like, "Oh my god, Claude is getting better at front-end. Claude is getting better at coding. I think Claude can develop new apps or something and so it can develop new features for the thing that I'm working." So it was kind of like this meta realization where it's like, "Oh my god, the world is actually changing." And when we first launched 100K context at that time, obviously I'm thinking about form factors that's like file uploads were very natural, very familiar to people. But you can imagine we could just make infinite chats in the Claude.ai app, as if it's 100K context. 

**Karina Nguyen** (00:37:04):
But because file uploads... It's like form follows function. It's like the form factor, the file uploads can enable people to just literally upload anything, the books, any reports, financial and ask any task to the model. And then I remember it was either enterprise customers, financial customers were really interested in that. It's like, "Oh wow." It's actually one of the very common tasks that people do in that setting. It's kind of crazy to see how some of the redundant tasks are getting automated basically by these smart models.

**Karina Nguyen** (00:37:48):
And they're entering the era where, I actually don't know for example sometimes if o1 gives me the correct answer or not because I'm not an expert in that field. And it's like, "I don't even know how to verify the outputs of the models." It's because all my experts know they can verify this. So, yes, so basically there are trends that are going on. The first trend is the cost of reasoning and intelligence is drastically going down.

**Karina Nguyen** (00:38:22):
I had a blog post about this. Maybe I should update on latest benchmarks, because at that time everybody was doing one benchmark and they'd be... quickly saturated the benchmarks. So I'm like, "Now we need to do the same plot but with another frontier eval." But the cost of intelligence is going down because it becomes that much cheaper. Small models are becoming even smarter than large models and that's because of the distillation research. 

This happened with Claude 3 Haiku. I was working with the training on the Claude 3 Haiku and I realized it was much smarter than Claude 2, which was way bigger, lots [inaudible 00:39:08]. But the power of small models become very intelligent and fast and cheap. We are moving towards that world. That has multiple implications, but the news is that people will have more access AI and that's really good. Builders and developers will have much better access to AI, but also it means all the work that has been bottlenecked by intelligence will be unblocked. 

**Karina Nguyen** (00:39:40):
I'm thinking about healthcare, right? Instead of going to a doctor, I can ask ChatGPT or give ChatGPT a list of symptoms and ask me, "Would I have a cold, flu, something else?" I can literally get the access to doctor almost. And there's been some research studies around that.

**Lenny Rachitsky** (00:40:05):
There was a New York Times story about that where they compared doctors to doctors using ChatGPT to just ChatGPT and just ChatGPT was the best of them. All doctors made it worse.

**Karina Nguyen** (00:40:18):
Yeah, that's crazy. Yeah. Yeah, that's crazy, right? Education I think I would have dreamt if I had the tool like ChatGPT when I was young and would learn so much. But it's like people can now learn almost anything from these models. So they can learn new language, they can learn how to build new look apps and write anything they do want. It's humbling to have... launch Canvas and bring that thing to the people, enable them to do something else that they couldn't have ever before. There's something magical around this experience. 

**Karina Nguyen** (00:40:57):
Education will have massive implications. I guess like scientific research, I think it's the dream of any AI research is to automate AI research. It's kind of scary, I'd say, which makes me think that people management will stay. It's one of the hardest thing to... Emotional intelligence with the models, creativity in itself is one of the hardest things. So writers, I don't think people should be worried as much. I think will alleviate a lot of redundant tasks for people.

**Lenny Rachitsky** (00:41:34):
This is awesome. Okay, I want to follow this thread for sure. And it's funny that what you described as you were an engineer at Anthropic and you're like, "Okay, Claude is going to be very good at engineering. This isn't going to be a potentially career long term, so I'm going to move into research and AI is going to need me for a long time to build it, to make it smarter."

**Karina Nguyen** (00:41:53):
I would say we still have... I think Canvas team has still have really cool front engineers that are really people who really care about interaction, design, interacting experience. I don't think models are there yet I think if... But we can get the models to this top 1% of front-ends and things for sure.

**Lenny Rachitsky** (00:42:16):
So what I want to move on to next along these lines is just, and this is just speculation, but what skills do you think will be most valuable going forward for product teams in particular? So folks are listening and they're like, "Okay, this is scary. What should I be building now to help me stay ahead and not be in trouble down the road?" What skills do you think are going to be more and more important to build?

**Karina Nguyen** (00:42:42):
Yeah, I think creative thinking. You want to generate a bunch of ideas and filter through them and not just build the best product experience. Listening. You want to build something that the most general model will not replace you. And oftentimes you build something and you make it really, really good for specific set of users and actually the mode is now in your user feedback. The mode is more in whether you listen to them, whether you can rapidly iterate. The mode is in here. I don't think we are yet to... There are so many ideas, I think there's an abundance of ideas that you can work on. I wouldn't be worried. I feel like in fact I just think people in AI field are like... I wish they were a little bit more creative and connecting the dots across the print fields or something like that to develop really cool new generation and new paradigms of interactions with this AI.

I don't think we've cracked this problem at all. A couple of years ago I was telling some people, I was like, "You want to build for the future." So it's like it doesn't necessarily matter whether the model is good or not, good right now, but you can build product ideas such that by the time the models will be really good, it'll work really well. I think it just happened naturally. For example, at Anthropic the Claude artifacts... And I feel early days of Canvas was, back in 2022 before ChatGPT, writing ideas was our knowledge [inaudible 00:44:36]. But I feel like Claude 1.3 model itself was not there to have made really extreme good high quality edits. For example, like coding. 

**Karina Nguyen** (00:44:47):
And I feel like I see startups like Kaeser was doing super well. And that's because they iterate so fast. They invent new ways of training models. They move really fast. They listen to what users like, massive distributions. Yeah, it's kind of cool. 

**Lenny Rachitsky** (00:45:08):
That's really helpful actually. So what I'm hearing is that soft skills essentially are going to be more and more important, powerful. You just talked about management, leading people, being creative and coming up with innovative insights, listening. There's a post I wrote that I'll link to where I try to analyze how AI will impact product management. And we're actually very aligned, and my sense was the same thing, that soft skills are going to become more and more important. And the things that are going to be replaced is the hard skills, which is interesting because usually people value the hard skills like coding, design, writing really well. And it's interesting that AI is actually really good at that because it's taking a bunch of data, synthesizing it and writing, creating a thing, versus all these fuzzy things around of what influences, convinces people to do things and aligning and listening, like you said, creativity, anything along those lines come up as I say that.

**Karina Nguyen** (00:46:01):
I think it's actually a really, really hard to teach the model how to be aesthetic or do really good visual design or how to be extremely creative in the way they write. I still think ChatGPT kind of sucks at writing and that's because it's bottlenecked by this creative reasoning. I think characterization is one of the most important... I think for a manager, I feel like... 

**Karina Nguyen** (00:46:28):
Actually, AI research progress is bottlenecked by management, research management. It's because you have constrained set of compute and you need to allocate the compute to the research paths that you feel the most convinced about. It was like you need to have a really high conviction in the research paths to put the compute, and it's more return on investment kind of situation. It's like, "Okay, I'm thinking a lot about across all my projects, which projects are higher priority?" Prioritization and also on the lower level, "Which experiments are really important to run right now and which are not?" and cut through the line. So I was thinking prioritization, communication, management. People skills like empathy, understanding people, collaboration.

**Karina Nguyen** (00:47:23):
I think Canvas wouldn't be an amazing launch if it wasn't about people and I think it's a wonderful group of people. And I get a chance to work with people like Lee Byron who's a co-creator at GraphQL and some of the best Apple designers. It's so cool to see... and how do you create this collaboration between people. It's just something that's still humane, I think.

**Lenny Rachitsky** (00:47:52):
Let me just follow through a little bit. I imagine people listening are like, "Okay, but once we have AGI or SGI it's like it'll do all this." There's a world where like, "Why isn't all this done?" I think it's easy to just assume all that. I'm curious this idea of creativity and listening, why you think AI isn't good at it, other than it's just very hard to train it to do this well. Is there anything there of just why this is especially difficult for AI and LLMs to get good at?

**Karina Nguyen** (00:48:20):
I think currently it's difficult for many reasons. I think it's still an active research area and it's something that I think my team is working on. It's like, "Okay, how do we teach the models to be more creative in the writing?" And so I'm thinking this new paradigm of wise that the models think more should actually lead to better writing in itself. But when it comes down to idea generation or discriminating of what is a good visual design or not, I feel like it hasn't had learned examples from people to discriminate it very well. I do think it's because there are not that many people who are actually really... It's not accessible to models to learn from these people I guess. So I definitely think that's why it sucks.

**Lenny Rachitsky** (00:49:19):
Yeah, that makes sense. Basically there's not enough of you yet, researchers teaching it to do these things, slash people that have incredible taste and creativity that can teach these things. You could argue this will come. 

**Karina Nguyen** (00:49:31):
Right. 

**Lenny Rachitsky** (00:49:31):
But we don't need to keep going down that thread. Let me ask you a specific question. In this post I wrote, I made this argument that a lot of people disagreed with that strategy is something that AI tooling will become increasingly great at and take over. There's the sense that that's the thing that people will continue to be much better at and you can't offload to AI basically developing your strategy, telling you what to do to win. My case is, "Isn't strategy, just take all the inputs, all the data you have available, understand the world around you and come up with a plan to win?" It feels like AI and LLM would be incredibly smart at this. What's your take?

**Karina Nguyen** (00:50:10):
I think so too. I think again, you teach the model all sorts of tools and capabilities and reasoning and it's like when it comes down to... For Canvas right now, it would be very cool for the model just aggregate all the feedback from users, summarize me the top five most painful flows on user experiences. And then the model itself is very capable of thinking of knowing how it's been made, figure out how to create a dataset for itself to train on it. And I don't think that we are far away from that self-improvement, models becoming self-improved by...

**Karina Nguyen** (00:50:54):
That, and the part of development, is basically self-improving. It's kind of like its own organism or something. Again, like strategies, it's more like data analysis and coming up with... I think what models are really good at is connecting the dots, I think. It's like if you have user feedback from this source, but you also have an internal dashboard with metrics and then you have other feedback or input and then it can create a plan for you, recommendations even.  And I think this is one of the most common use cases for ChatGPT too, is coming up with these sort of things.

**Lenny Rachitsky** (00:51:47):
That makes sense essentially a human can only comprehend so much information at once and look at so much data at once to synthesize takeaways. And as you said, these context windows are huge now. Here's all the information, what's the most important thing I should do?

**Karina Nguyen** (00:51:59):
Yeah, same as scientific research. Ideally the model would be able to suggest ideas, new ideas, or iterate on the experimental given the empirical results of the previous experiments like how do you come up with new ideas or the methods?

**Lenny Rachitsky** (00:52:18):
Yeah. Oh, man. Okay, so just to close the loop on this conversation, this part of the thread is the skills you're suggesting people focus on building and leaning into is soft skills like creativity, managing influence, collaboration, looking for patterns. Is that generally where your mind is at?

**Karina Nguyen** (00:52:40):
Yeah, I'm thinking a lot about how do we make organizations more effectively and I think this is mostly management, I guess. It's like how do you organize research teams or generally teams combined... Compose teams such that they will be at their maximally succeed or at the maximal performance of what can possibly... We can literally create the next generation of computers. It's just the matter of conviction and the way you manage through that. It's scaling organizations or scaling product research, I guess.

**Lenny Rachitsky** (00:53:15):
Yeah, I think you're basically building this thing and not efficiently doing it is limiting the potential of the human species right now.

**Karina Nguyen** (00:53:16):
Right. 

**Lenny Rachitsky** (00:53:26):
It's mismanagement within the research team in OpenAI and Anthropic and some of these other models.

**Karina Nguyen** (00:53:32):
Yeah, it's kind of crazy to think about it.

**Lenny Rachitsky** (00:53:33):
Holy moly. Okay, so speaking of Anthropic and OpenAI, you've worked at both. Very few people have worked at both companies and have seen how they operate. I'm curious just what you've noticed about the differences between these two, how they operate, how they think, how they approach stuff. What can you share along those lines?

**Karina Nguyen** (00:53:48):
It's more similar than different. Obviously there was a lot of... There are some differences always comes to nuances. I would say culture. I really love Anthropic and I have a lot of friends there. And I also love OpenAI and they still have a lot of friends though. So it's not about enemies. I feel like there's in AI, it's all like, "Yeah, they're competitors. There's enemies." It's actually like one big community of people doing the same thing. I would say what I've learned from Anthropic is this real care and craft towards model behavior, model craft, model training. 

**Karina Nguyen** (00:54:32):
And I've been thinking a lot about, "Okay, what makes Claude Claude and what makes ChatGPT ChatGPT?" And it's like I still have some sense of operational processes that leads to the outputs, to the model. It's the outputed model. And it's like the reason why Claude has so much more personality and is more like a librarian... I don't know. I don't know. I am visualizing Claude being like a librarian at some point, very nerdy or something. ... is because I feel like it's the reflection of the creators who are making this model. And a lot of details around the character and the personality and whether the model should follow up on this question or not. 

**Karina Nguyen** (00:55:19):
What's the correct ethical behavior for the model in these scenarios? A lot of crafts and curated datasets. This is where I learned that part of art, I guess, at Anthropic. I would say Anthropic is much smaller. When I joined it was, what, like 70 people? When I left it was tons of people. And obviously the culture changed so much. I really enjoyed being early days startup lives, and people knew each other as a family. But the culture shifted. 

**Karina Nguyen** (00:55:53):
I would say that I learned from Anthropic that they're much better at focusing and prioritization of... Very hardcore prioritization, I guess. And they need to do it. But I think OpenAI's much more innovative and much more risk-takers in terms of product or research. Actually, in way your full-time job can be just teaching the model how to be creative writers. And it's like there's some luxury in this research freedom that comes with scale, maybe. I don't know. I'd say I have much more creative product freedom to do almost anything, I guess, within OpenAI, evolve ChatGPT into the vision that we want. It's more probably bottoms-up, I guess.

**Lenny Rachitsky** (00:56:51):
Yeah, that's how I was thinking about it. It feels like OpenAI is more bottoms-up, distributed, people bubble up ideas, try stuff. And that leads to more products launching, I imagine more things just kind of being tried versus more of a, "Let's just make sure everything we do is awesome and great and craft and thinking deeply about every investment."

**Karina Nguyen** (00:57:08):
Right. 

**Lenny Rachitsky** (00:57:08):
That's really interesting. I've never heard it described this way. Karina, we've covered so much ground. This is going to help a lot of people with so many ways of thinking about where the future's going. Before we get to our very exciting lightning round, I'm curious if there's anything else that you think might be helpful to share or get into?

**Karina Nguyen** (00:57:23):
One of my regrets, I guess, when I was early days at Anthropic was that... I think there was some luxury of the time, because pre-ChatGPT, to actually come in with a bunch of ideas and prototype almost every day. And I think that we did a lot of cool ideas like Claude, and Slack was actually one of the first tool-usey products. It's like Claude could operate in your workplace now. It's kind of cool because you can add Claude to summarize the thread. So maybe you have an entire conversation with someone and then you want a summary of what happened you can ask Claude, "Summarize this." 

Also, it was really fun to iterate on the model itself. It's like when you just talk to the model in Slack forever. It created some social element, it was kind like [inaudible 00:58:19] and this Discord, people learned so much about prompting and how to work with Claude. Actually, one of the features that was early tasks prototype is every Monday Claude would just summarize the entire channel. Or every Friday we'd just summarize a bunch of channels and give the news about the organization, or something. 

**Karina Nguyen** (00:58:48):
And it's kind of like really cool form factor. I think thinking about form factor's a really important question in AI, especially we haven't even figured out how do we create an awesome product experience with o-series models. It's like the paradigm between synchronous real time give an answer paradigm into more asynchronous paradigm of agents working on the background. But then now the question is the agents should build trust with you, right? And trust builds over time, which is like with humans. And you start this collaboration which is why this collaboration model with you and the model is so important because you build trust and the model learns from your preferences so that it can become more personalized and it will start predicting the next action that you want to take on the computer or something. And it's more predictive, much more... We went from personal computers to personal model basically here.

**Lenny Rachitsky** (00:59:54):
Why is it not a thing? That seems like such an obvious feature that every LLM should have as a Slack bot version of them. Is that a thing I can help you install? Or is that not a thing right now?

**Karina Nguyen** (01:00:03):
I know that Claude and Slack was sunset in 2023 or something. I think it was after ChatGPT was mostly the focus on customer use cases or enterprise use cases.

**Lenny Rachitsky** (01:00:17):
Mm-hmm. Bummer. 

**Karina Nguyen** (01:00:19):
I think the form factor of Claude and Slack was kind of constrained a little bit when you want to talk about new features.

**Lenny Rachitsky** (01:00:28):
Bummer. I want that.

**Karina Nguyen** (01:00:30):
I know that ChatGPT had Slackbar tools. I don't know, maybe it will come back sometime.

**Lenny Rachitsky** (01:00:35):
All right, I would pay for that. Any other memories from that time of early days? Because that's a really special place to have been is early days Anthropic. Any other memories or stories from that time that might be interesting to share?

**Karina Nguyen** (01:00:48):
I think the very first launch when we felt... When click from use, again, was 100K context launch is when the models could input the entire book and give you a summary of the book or something. Or the financial... or catalog multi files financial reports and then give you an answer to the question, to very specific questions. I think there was something in there that was kind like, "Oh my god, this is a really cool new capability." Not model capability, but more like the capabilities that came from the product form factor itself rather than the model capability as much. 

I think other prototypes that we were thinking about... There's one part having a Claude workspaces and it's kind of the same idea of Claude and I would have this shared workspace and that share workspace is like a document and we can iterate on the document. And I feel like sometimes the ideas, [inaudible 01:01:55] and they're locked for two years, just like in this case. 

**Lenny Rachitsky** (01:02:00):
It's interesting, there's these milestones that kind of open up our view of what is happening and where things are going. ChatGPT think was the first of just like, "Wow, this is much better than I would've thought." You talked about 100K context windows where you could upload a book and ask it questions and have it summarize. I actually use that all the time. When I have interview guests and they wrote a book, I sometimes don't have time to read the whole book. So I use it to help me understand what the most interesting parts are. And then I actually dive into the book, just to be clear. And then, I don't know, maybe voice was another one where you could talk to say ChatGPT. Is there any other moments there that you're like, "Wow, this is much better than I thought it was going to be?"

**Karina Nguyen** (01:02:39):
Yeah, I think the computer use agents, like the model operating the desktop. And you can essentially think of new kind of experience where the model can learn the way you browse. And from that preference it can just browse as just like you. It's kind of simulated persona. And it's actually very similar to the idea of like, "Okay, maybe Sam Altman doesn't have a lot of time. Maybe I want to talk to his simulation and ask..." Or, for example, I really appreciate some of the technical mentorship. Yeah, cool. But he doesn't have a lot of time so it's like I really want to ask him this questions. How do you respond with simulated environments like this would be really cool.

**Lenny Rachitsky** (01:03:37):
That's a great place to plug Lennybot, have one of those. It's trained on all of my podcasts and newsletters. 

**Karina Nguyen** (01:03:42):
Oh, cool. 

**Lenny Rachitsky** (01:03:43):
It sits on many models. I don't know which exactly they use, but it's exactly that. And it's not even me, it's all the guests that have been on the podcast and on newsletter as I wrote. And you could just ask it, "How do I grow my product? How do I develop a strategy?" And it's actually shockingly good.

**Karina Nguyen** (01:03:58):
Do you feel like it reflects who you are?

**Lenny Rachitsky** (01:03:58):
Yeah. 

**Karina Nguyen** (01:03:58):
Or would it be... Okay. 

**Lenny Rachitsky** (01:04:01):
The best part of it is you can talk to it. There's an ElevenLabs voice version that's trained on my voice from this podcast, and it's actually very good and people have told me they sit there for hours talking to it. 

**Karina Nguyen** (01:04:15):
Wow.

**Lenny Rachitsky** (01:04:15):
And somebody told it, "Interview me like I am on Lenny's podcast, ask me questions about my career." And he did a half hour podcast episode with Lennybot.

**Karina Nguyen** (01:04:24):
Oh my god, that's so fun.

**Lenny Rachitsky** (01:04:27):
It's incredible. Future is wild.

**Karina Nguyen** (01:04:29):
Yeah. I think content transformation is... I would imagine sometime when you generate a sci-fi story in Canvas, you can transform this into audiobook where you have very natural content transformation of one media to another media. I think one of my earliest inspiration is one of the last episodes of Westworld where, I don't want to spoil, but where Dolores comes to her work at that time and she comes to this new workspace and she starts writing a story. And then as she writes a story, a 3D, virtual reality, starts creating on the fly. So I kind of want to create that. Kind of cool. 

**Lenny Rachitsky** (01:05:24):
Wow. Speaking of medium, I guess I was wondering if I should go in this direct or not, but real quick. Kevin Weill/Kevin Weill, I don't know exactly how to pronounce his last name, the CPO of OpenAI.

**Karina Nguyen** (01:05:35):
Kevin Weill, uh-huh.

**Lenny Rachitsky** (01:05:37):
Is it Weill or Weill? 

**Karina Nguyen** (01:05:37):
I think Weill. 

**Lenny Rachitsky** (01:05:39):
Weill. Okay. Okay. Let's just say that. We'll go with that. 

**Karina Nguyen** (01:05:40):
I hope, yeah. 

**Lenny Rachitsky** (01:05:43):
He did a panel at the Lenny and Friends Summit last year and he made this really fascinating point that chat is a really interesting interface for these tools because they're just getting smarter and smarter and smarter and smarter and smarter. And chat continues to work as a paradigm to just interact with them, similar to a human. You could talk to Albert Einstein. You could talk to someone not very smart and it's all conversation still. And so it's a really flexible way to interact with increasingly good intelligence. At some point it'll not be so great, and you were talking about all these ways that you're adding additional ways to interact. But it's interesting chat proved to be a really powerful layer on top of all this stuff. 

**Karina Nguyen** (01:06:22):
Yeah, that's real cool. I feel like chat also has social element which is very humane. It's like, yeah, you sometimes want to get into group chat. And having conversations with AI is kind of like a group chat in itself, as messaging. Actually, this idea of how do you build features like this? I see tasks as this general feature that will scale very nicely as the models would develop new capabilities themselves. The models will be able to do better searches and create new... come up with more creative writing on render, react apps and like HTML apps. And you can have everyday new puzzle for you, every day continue the story from the previous days. It scales very nicely.

**Lenny Rachitsky** (01:07:14):
You mentioned something as we were getting into this extra section that we ended up going down is this idea of the agents using a computer. I know this is actually something you are going to launch today, the day we're recording it, which will be out by the time this comes out, called Operator, can you talk about this very cool feature that people will have access to?

**Karina Nguyen** (01:07:33):
Yeah, so I unfortunately did not work on that, but I'm really, really excited about this launch. It's basically an agent that can complete the task in its own virtual computer, in its own virtual environment. You can do any literally task like order me a book on Amazon. And then ideally the model will either follow up with you which book do you want, or know you so well that it start recommending, "Oh, here is the five books that I might recommend you to buy." And then you hit, "Yeah, help me buy." And then the model goes off into its own virtual little browser and complete the task and buy the book on the Amazon. And then if you give the model credentials, credit cards, obviously it comes with a lot of trust and safety, then it will just complete the thing for you. It's a virtual assistant.

**Lenny Rachitsky** (01:08:37):
It's interesting how this just sounds like obviously this should happen. Why is this not yet a thing? Which is also mind-blowing that we're just assuming this should exist. Just some AI doing things for you on a computer we just ask it to do. 

**Karina Nguyen** (01:08:37):
Yeah. 

**Lenny Rachitsky** (01:08:50):
It's absurd.

**Karina Nguyen** (01:08:51):
It's actually really hard. And I think you're still cracking this, but feel like... I don't know if you use Tuple like a pair programming product.

**Lenny Rachitsky** (01:09:03):
No.

**Karina Nguyen** (01:09:04):
But at Anthropic we loved pair programming, so if you used-

**Lenny Rachitsky** (01:09:09):
Oh yeah, Shopify uses this. I remember it came up on a podcast episode.

**Karina Nguyen** (01:09:10):
Oh, nice. Yeah, so it is a very cool product where you can just call anyone at any time and then share screen and the other person can have access to the screen or start literally operating your computer. And it's very realtime... The allegiance is very... it's very high quality. And it's just like I kind of want the same. I want to pair program with my model and the model should even talk to me. Draw very specific section in my code and just go to tell me... Obviously teach me and we can have different modes. It's like right, this is a product right here for you. I don't know. Some people should build that. 

**Lenny Rachitsky** (01:09:58):
It sounds like a startup just got birthed-

**Karina Nguyen** (01:09:59):
Yes.

**Lenny Rachitsky** (01:10:00):
... from someone listening to this. You mentioned that it's very hard to do this agent controlling a computer as you and helping out. What makes it so hard for whatever, however much you can explain briefly?

**Karina Nguyen** (01:10:11):
Much of it is because right now the model's operating on pixels instead of language or whatnot. Pixels is actually really, really hard. The models [inaudible 01:10:25] perception, or visual perception. I think there's still a lot of multimodal research that's going on, but I think language scaled so much easier compared to multimodal because of that.

**Karina Nguyen** (01:10:38):
Another thing that I guess my team is working that is how do you derive human intent very correctly? It's like sometimes does the model know enough information to ask a follow-up question or to complete the task? You don't want an agent to go off for 10 minutes and then come back with an answer that you didn't even want. That actually creates much more worse user experience. And this comes with teaching the model people skills. It's like, "What do people like? Kind of like creating the mental model of the user and care about the user in order to ask certain questions. Actually, that part is hard to do for the models.

**Lenny Rachitsky** (01:11:28):
That relates to what we talked about earlier where this kind of the soft skill, people skills piece is not where these models are strong yet.

**Karina Nguyen** (01:11:34):
Yeah. 

**Lenny Rachitsky** (01:11:35):
Okay. I'm going to skip the lightning round. I want to ask just one question from the lightning round, something fun. 

**Karina Nguyen** (01:11:41):
Yeah. 

**Lenny Rachitsky** (01:11:44):
Okay, so when AI replaces your job, Karina, I'm curious what you're... And it gives you a stipend, gives you a monthly stipend. Here's your salary for the month. What would you want to do? What do you want to spend your time on? What will you be doing in this future world?

**Karina Nguyen** (01:11:57):
I've been thinking about this a lot times. I feel like I have a lot of jobs options. I would love to be a writer, I think. I think that would be super cool. You should write short stories, sci-fi stories, novels. I really like art history, so you know those conservationists in the museums who just try to preserve art paintings, but just painting through a long day? 

**Lenny Rachitsky** (01:12:28):
Mm-hmm. 

**Karina Nguyen** (01:12:29):
I think that would be really cool to do. Yeah. 

**Lenny Rachitsky** (01:12:36):
That sounds beautiful.

**Karina Nguyen** (01:12:36):
I don't know.

**Lenny Rachitsky** (01:12:39):
What I'm hearing is you need to Nerf these models to not get very good at writing so that you can continue... Although at that point you don't need to do it from... You don't need people to buy it, you're just doing it for fun, so it doesn't even matter if they're incredibly good at writing or art conservation. Oh man, what an episode of our conversation. What a wild time we're living in. Karina, thank you so much for being here. Two final questions. Where can folks find you online if they want to reach out and follow up on anything? And how can listeners be useful to you?

**Karina Nguyen** (01:13:06):
You can find me, I'm on Twitter it's KarinaNguyen. You can also shoot me an email on my website. And my team is hiring and so I'm looking for research engineers, research scientists, as well as machine learning engineers, people who come from product engineers who want to learn model training. I'm actually hiring for my team. My team is called Frontier Product Research, and we train models, we develop new methods but for product oriented outcomes.

**Lenny Rachitsky** (01:13:38):
What a place to work. Holy moly. What's the best way for people to apply for these very lucrative roles?

**Karina Nguyen** (01:13:46):
I think you can shoot me a DM on Twitter.

**Lenny Rachitsky** (01:13:49):
Okay. 

**Karina Nguyen** (01:13:51):
Or I'm yet to create a job description for them.

**Lenny Rachitsky** (01:13:51):
Okay. This is the job description.

**Karina Nguyen** (01:13:58):
Or you can apply into post training team. Yeah.

**Lenny Rachitsky** (01:13:58):
Okay. You're going to get a flood of DMs. I hope you're prepared. Karina, thank you so much for being here. This was incredible.

**Karina Nguyen** (01:14:03):
Thank you so much, Lenny.

**Lenny Rachitsky** (01:14:05):
Bye, everyone. 

**Karina Nguyen** (01:14:05):
It was fun. 

**Lenny Rachitsky** (01:14:09):
Thank you so much for listening. If you found this valuable, you can subscribe to the show on Apple Podcasts, Spotify, or your favorite podcast app. Also, please consider giving us a rating or leaving a review as that really helps other listeners find the podcast. You can find all past episodes or learn more about the show at LennysPodcasts.com. See you in the next episode.

---

## [8/15] An inside look at X’s Community Notes | Keith Coleman (VP of Product) and Jay Baxter (ML Lead)

**Lenny Rachitsky** (00:00:00):
The work that you guys do has had such a tremendous impact on the way the world works. I want to start with just giving people a brief understanding of what is Community Notes.

**Keith Coleman** (00:00:09):
Someone on X can see a post. If they think it's misleading, they can propose a note that they think other people might find informative. Other people can then rate that note.

**Jay Baxter** (00:00:18):
We actually look for agreement from people who have disagreed in the past. And what we see is when people actually have that sort of surprising agreement, that's what makes the notes so neutral and accurate and well- written, really, overall.

**Lenny Rachitsky** (00:00:31):
There's many people that are very polarized. How do you deal with people that are super anti-vax, super Jan 6?

**Keith Coleman** (00:00:36):
One philosophical thing that's important is that we want all of humanity to participate and sometimes people are surprised by that. We have all of humanity. We then have the data to understand what notes will be helpful to actual humanity. Every post is eligible for notes. We shouldn't exempt Elon. We shouldn't exempt government figures. We should be like everyone... Even advertisers can get notes.

**Jay Baxter** (00:00:58):
There have been external studies run by people totally independent of us who have found that if you take a post with or without a Community Note, that actually people's agreement with the core claims in the post does change if they see it with a note versus without.

**Lenny Rachitsky** (00:01:13):
Is there anything else along the lines of just working for Elon within an org Elon runs that might surprise people?

**Keith Coleman** (00:01:18):
If I were to start a company in that company, it would be even leaner than I would've made it before. I've been amazed with just how much the team is able to accomplish with a small group and I think because of a small group-

**Lenny Rachitsky** (00:01:33):
Today, my guests are Keith Coleman, Product Lead for Community Notes, and Jay Baxter, Founding ML Engineer and Researcher for Community Notes. This conversation may be my newest favorite podcast episode so far. Community Notes is one of the most impactful and clever and, also, underappreciated products in the world right now.

**Lenny Rachitsky** (00:01:52):
If you ever use X/Twitter and you see a note underneath a tweet correcting the misinformation in that tweet, that is Community Notes. I've never heard a deep dive into the story behind the product and the team that built it and I'm excited to bring you just that. We get into the surprising origin story of the product, how the algorithm actually works, how the algorithm emerged out of an internal contest within Twitter, the principles behind Community Notes, and why staying true to them has been so key to its success. Also, how it survived four different leaders, including Elon and Jack, and why it's now a big part of the solution to solving misinformation on the internet. Including recently being adopted by Meta as their main fact-checking tool. This is an incredibly special episode and I'm so excited to bring it to you.

**Lenny Rachitsky** (00:02:36):
If you enjoy this podcast, don't forget to subscribe and follow it in your favorite podcasting app or YouTube. Also, if you become a subscriber of my newsletter, you now get a year free of Notion and Superhuman and Granola and Linear and Perplexity Pro. Check that out at lennysnewsletter.com.

**Keith Coleman** (00:05:23):
It's great to be here. [inaudible 00:05:25].

**Jay Baxter** (00:05:25):
Thanks for having us on.

**Lenny Rachitsky** (00:05:26):
It's so my pleasure. I'm so thrilled to be having this conversation. The work that you guys do has had such a tremendous impact on the way the world works. So many product teams are always talking about driving impact and want to drive impact. You guys have actually built things that have changed the world in meaningful ways and continue to do that. And I've never really heard the backstory of how Community Notes came to be and how it works and all these things, so I'm really appreciative of you guys making time to chat.

**Keith Coleman** (00:05:52):
Yeah. First, thanks for saying that. That's why we built this thing is to help people and it's great to hear it. It's great to see people enjoying it and finding it useful.

**Lenny Rachitsky** (00:06:02):
I want to start with just giving people a brief understanding of what is Community Notes. I think a lot of people kind of heard about it, kind of maybe see it on X. As they scroll through, they see these notes but they're like, "I don't actually know what this is." So can you just briefly describe what is Community Notes?

**Keith Coleman** (00:06:18):
Community Notes is a way for the people, like the public, to add context to posts that might be misleading. The basic way it works is that someone on X can see a post. If they think it's misleading, they can propose a note that they think other people might find informative. Other people can then rate that note. And if the note is found helpful by people who normally disagree with each other, indicating that it's probably accurate, it's probably really neutrally-worded, it's probably informative, then it will show to everyone on X. The goal is just to get people more information about what they're seeing so they can make better decisions in their lives.

**Lenny Rachitsky** (00:06:57):
Amazing, and I think hearing this, it's absurd that this works. I think when people originally heard this idea like, "No way this is going to work." And so, just to dive a little bit deeper, can you give us a deeper understanding of how it actually works? Because I think it's the algorithm that you guys designed that is so clever that allowed this to work. So talk a little bit about that algorithm.

**Jay Baxter** (00:07:22):
Yeah. So I think a key misunderstanding a lot of people have if they haven't really dived into details, they just think that maybe someone can write a note and it appears immediately or we're just taking a majority rules vote of who thinks the note's good. I think both of those approaches would probably lead to biased or inaccurate notes. I think the key thing, really, that we do is we actually look for agreement from people who have disagreed in the past.

**Jay Baxter** (00:07:48):
And what we see is when people actually have that sort of surprising agreement, that's what makes the notes so neutral and accurate and well-written, really, overall. It's just that people who are very polarized, overall, often can't find agreement when things aren't accurate, right? I think it also provides some good anti-manipulation properties. I think people are often... If you said... I think back in 2020 before we started building anything here, whether this could work at all, I think a room of ML engineers would say, "Oh, you have to keep it closed source. People are going to be manipulating this all the time. You have to use ground truth labels from fact checkers. There's no way that you could bootstrap the system without external labels." But it turns out that you can do that with this kind of bridging-based agreement algorithm is what we call it.

**Lenny Rachitsky** (00:08:41):
Okay. So just to summarize and make it super clear. It's basically people... Someone writes a note. This information is fault... What's a good example, just as we talk about this, like a classic example?

**Keith Coleman** (00:08:50):
A really classic example is an AI generated image or an out of context image like, "Look what's happening here." But it's actually from five years ago in a different country and a different topic or something-

**Lenny Rachitsky** (00:09:00):
Oh, man. I've seen this so many times where it's like, "Look what's happening in San Francisco," and I'm like, "No, this is a whole different city and that's not-"

**Keith Coleman** (00:09:06):
Totally. Yeah.

**Lenny Rachitsky** (00:09:08):
Yeah. Okay. So someone posts this AI image. Someone writes a note, "This is actually five years ago in a different city," and this algorithm helps understand if this note is true and it's just regular people doing this.

**Jay Baxter** (00:09:23):
Yep. Regular people who have signed up to be Community Notes contributors. So there are a few checks, like you do have to have a verified phone number for instance. But yeah, at the end of the day, these are regular people. Not necessarily professional fact checkers or anything like that.

**Keith Coleman** (00:09:40):
And yeah, that was really important to us too. There was a question at the beginning, to the point Jay was making of like, "Did anyone think this was going to work?" Obviously, it was a crazy idea. We didn't know if regular people were going to be able to do this task and certainly people had concerns about whether they would do it effectively.

**Keith Coleman** (00:09:58):
Initially, some people inside the company were suggesting like, "Hey, why don't you have journalists or some select group be the first participants?" But very specifically we were like, "No. We're trying to move away from the idea of curated editorial decisions being made around this. This is supposed to be open to everyone." So we very intentionally try to allow all humans in. People are randomly selected and that's important to it feeling fair, feeling open, feeling trustable.

**Lenny Rachitsky** (00:10:27):
Yeah. And again, it's just like this sounds like the holy grail of understanding what is true and it actually works. And works so well that Meta recently, as you all know, decided to adopt this exact system for them instead of having tens of thousands of fact checkers reviewing things.

**Jay Baxter** (00:10:46):
One distinction that I would make, which maybe can come off as nitpicky but I think is important, is Community Notes adds additional context. It's not fact-checking necessarily, right? So there are cases where the post could be true. But maybe, it's just misleading because there there's no context or there's missing context. We cover those cases and I think that's an important distinction. We just have the philosophy that users should be able to make up their own minds, right? Like, "Here's extra context, take it or leave it," right?

**Lenny Rachitsky** (00:11:18):
Yeah. What I think about, you shared this with me, this example of a picture with a cat and somebody's Community Note was just, "That's a dog." Or is it the other way around or that's a-

**Jay Baxter** (00:11:31):
Yeah. "A Palestinian boy shares his bread with a dog," was the post and it's a picture of this cat. So obviously, this particular note is not super necessary because it just says, "That's a cat," and links to a Wikipedia for cat. It's a good example that the system is... This is not something a professional fact-checker or whatever or you think would need fact-checking. But it's proof that the system is really run by the users at the end of the day and adds some comic relief, I guess. And the note is correct.

**Lenny Rachitsky** (00:12:06):
Okay. It's important.

**Jay Baxter** (00:12:08):
Yeah.

**Lenny Rachitsky** (00:12:08):
When does a post get triggered to even be considered for Community Note? Is there a threshold or is it just you can write a Community Note on anything and people decide what they would vote on? How does that work?

**Keith Coleman** (00:12:19):
So every post is eligible for notes and that was, again, another really important principle. It's like, "We shouldn't exempt Elon. We shouldn't exempt government figures. We should..." Everyone, even advertisers, can get notes. So any posts on the platform can get a note. And if you look in practice, you'll see notes appearing on world leaders, on Elon, on ads, on media organizations, and on, obviously, just regular people using social media. But yeah, the idea is really that it's an even playing field. For a note to be proposed, the person proposing it has to have earned the ability to write notes. So there is that aspect where you have to earn in to be able to do this. And the way you earn that ability is through your ratings by demonstrating the ability to help identify notes that are found helpful to a broad range of people. So basically, if you have an ability to see and know, recognize what's helpful with a lot of people, then you have the ability to start proposing notes.

**Lenny Rachitsky** (00:13:20):
I actually signed up to be on... What do you call these people? Note take-

**Jay Baxter** (00:13:24):
Contributors.

**Lenny Rachitsky** (00:13:25):
Okay. Contributors. Yeah. So I've been rating. I haven't achieved-

**Keith Coleman** (00:13:29):
Nice.

**Lenny Rachitsky** (00:13:29):
I can't write notes yet.

**Keith Coleman** (00:13:30):
Yeah. It's not super easy. It takes some effort.

**Lenny Rachitsky** (00:13:33):
Are there stats you can share about the scale of Community Notes at this point, especially things that might surprise people?

**Keith Coleman** (00:13:39):
Yeah. I mean, the service is growing rapidly, so there are hundreds of notes per day. And to put that into context, I saw some stats recently from someone at UC Berkeley saying there was something like 10 traditional fact checks a day. So in contrast, there's hundreds of notes a day that are getting shown. They span a huge range of topics from, obviously, politics, news, out to entertainment, sports, gaming. Just whatever is going on that day.

**Keith Coleman** (00:14:07):
In addition to there being hundreds of these individual notes, they can also be matched to multiple posts. So if someone writes a note on an image or a video, like let's say it's AI generated or something like that, that note will automatically be matched to all posts that contain the same image. So you can have a single note matching to thousands of posts. And over let's say the last year, 2024, we had something like 95,000 notes that were seen about 30 billion times. That's more than double the prior year. Prior year was something like 37K notes seen 14 billion times. So that rate is increasing dramatically when you think about 30 billion views, that's a lot of information that is getting out there that might not have been out there otherwise, which is pretty cool. And part of the reason it is expanding like that is the contributor base is expanding. There's something like 950,000 contributors around the world. That's nearing a million people making this happen which is amazing.

**Lenny Rachitsky** (00:15:13):
Wow. And I'm one of those, right? I count as a contributor?

**Keith Coleman** (00:15:15):
Yeah. Yep. No. If you're signed up as a contributor, you count.

**Lenny Rachitsky** (00:15:16):
Okay. Cool.

**Jay Baxter** (00:15:18):
Then, there's more people on the waitlist too. So there's plenty of headroom for more growth. Regarding the matching on media and URLs, I think that's a huge way to get extra coverage. Also, I do think we've been very careful to make sure that those matches are precise. Because I think one thing that people love about Community Notes compared to other types of fact checking is that, actually, the notes are custom written for the particular claim you're seeing, right? So often, a fact check warning would just say something like, "Get the facts here." And then, there's a link to some generic page about voting information, which is so not helpful to have the information behind a click. So pulling the context up so that you have zero clicks that you need to make and keeping it specific is so important.

**Lenny Rachitsky** (00:16:11):
One feature I love that I imagine you guys thought deeply about is if I liked the post in the past, I get notified later if a community note shows up, so that I'm not remembering this false information.

**Keith Coleman** (00:16:22):
Yeah. I mean, we try to make notes as fast as we can, so we want them to appear instantly if possible. But inevitably, there's going to be a time gap between when a post goes live and when people figure out what's going on and when they get the note out there. And so, we send those notifications to try to close that gap. And yeah, we get a lot of love for that. We see people take screenshots and share them. They're excited about it. And it's also a pretty cool example of something you can do on the internet, in the social media world that was difficult in a print or standard news world where you would see maybe a correction the next day in a corner of a paper that was hard to read. Here, you're getting a ping about it if you've engaged with a post and note shows up.

**Lenny Rachitsky** (00:17:05):
One user feedback point is I'd love the push to just tell me, "Here's what you got wrong." Because I find that I actually have to go into it and read it and I feel like the push could just be like, "Here's more context to this thing." You're like-

**Jay Baxter** (00:17:20):
Agreed.

**Keith Coleman** (00:17:20):
We'll go take a look at that-

**Lenny Rachitsky** (00:17:21):
There we go. Live user feedback.

**Keith Coleman** (00:17:24):
Nice.

**Lenny Rachitsky** (00:17:25):
Okay. I want to get into the origin story of this whole thing. But two more questions, because we're on this thread. One is what's the the threshold for a note to show up on a note? Is that information you can share, just how does that work?

**Jay Baxter** (00:17:35):
So just because of the details of the way the algorithm works, it uses this machine learning algorithm called Matrix factorization where we fit it with Gradient Descent and whatnot. The threshold is it's 0.4 on this made up scale-

**Lenny Rachitsky** (00:17:52):
0.4. Great.

**Jay Baxter** (00:17:53):
Yeah. I mean, in practice, what it means is basically a majority of people... If there is a polarized divide relevant to the notes. Obviously, some notes are not about politics or something polarizing. But if there is, then a sizable majority of people on both sides would generally need to find the note helpful. And then, there are other rules that come into play beyond that main one. So even if it's above that threshold, it might get filtered out if... There's a separate algorithm that's looking at agreement between people's incorrect tags. So like maybe people found the note helpful but incorrect, right? It happens. And in those cases, it doesn't matter if it's above the helpfulness threshold.

**Lenny Rachitsky** (00:18:39):
This is probably the wrong way to think about it, but is it 40% of people that normally disagree, agree-

**Jay Baxter** (00:18:45):
No.

**Lenny Rachitsky** (00:18:45):
Okay. It's-

**Jay Baxter** (00:18:45):
It means nothing like that. It's just like on some arbitrary scale-

**Lenny Rachitsky** (00:18:48):
Okay.

**Jay Baxter** (00:18:48):
Yeah.

**Keith Coleman** (00:18:49):
Yeah. If we change random other things about the algorithm, that number would also have to change to an equally seemingly arbitrary number. We arrived at some numbers like that by gauging user feedback. So we could share a lot of notes with people, get feedback on which ones are helpful, and just a line emerged about indicating where things go from questionable to pretty clearly helpful.

**Jay Baxter** (00:19:13):
Yeah. And it is set right now, by the way, to be really conservative, I think. We just are pretty particular about quality and we really want note quality to be really high. I think Keith and I both believe that we live or die based on the quality of the notes at the end of the day. So we'd rather not show a note that maybe good, but we didn't have enough signal on than the other way around.

**Lenny Rachitsky** (00:19:41):
That makes so much sense. I've never seen a Community Note that is wrong and breaking that promise is a big deal. So I completely get why you guys are super conservative there. Okay. Two more questions [inaudible 00:19:53] because I'm just curious. These weren't on my list of questions to ask, but I feel like people wonder this. How many notes are written versus end up showing up and triggering on a-

**Keith Coleman** (00:20:02):
We probably show about 8% of notes that get proposed. It's been between, let's say, 7% and 10% or 11%, something like that over time. The number can vary a little bit. And as Jay said, there are undoubtedly... And you can see it, there's clearly more good notes than we show, but the goal is to hold a really high bar. We want to show a note when it's going to be helpful, when it's not going to appear biased and undermine trust in the system. We want these to be neutral, informative, helpful. And as Jay was saying, we view the worst possible mistake as showing a bad note because that's going to undermine trust and the trust is why people like the product.

**Keith Coleman** (00:20:47):
So yeah, the bar is there. And like I said, there's clearly some in that remaining, let's call it 90%, that are good. And then, there's a lot that are just not that great and there's some that are bad. And if you write one of these ones that are bad which bad being defined as people who normally disagree find the note not helpful, so it's like the inverse of the ones we show. If you write one that people normally disagree, find not helpful, you actually will ultimately lose your ability to write and have to earn it back. That other 90% is a mix. Sometimes people look at the number, they're like, "Oh, why don't you show more?" It's like, "Well, you probably actually don't really want us showing most of those." The gold here is that the system is able to filter out the good ones.

**Lenny Rachitsky** (00:21:31):
That makes sense. Okay. One other question is there's many people that are very polarized, like very disagreeable with a lot of things. How do they filter into this algorithm? How do you deal with people that are super anti-vax, super Jan 6, like all these very extreme potential views?

**Jay Baxter** (00:21:47):
If people really are so polarized that there isn't agreement among people that typically disagree, it's possible that this is one of those notes that might be correct, but it wouldn't be helpful to show as context. Maybe it's about a claim that people have really entrenched opinions about and they've read hundreds of things about it already.

**Jay Baxter** (00:22:15):
Probably this is just not going to improve people's understanding. It's just not going to be a helpful user experience. So it might not be the worst thing in those cases to not show the note. People, a few years ago, were pretty pessimistic that maybe fact-checking never changes people's understandings about what's true. Actually, there have been external studies run by people totally independent of us who have found that if you take a community note or posts with or without a community note... That actually, people's agreement with the core claims in the post does change if they see it with the note versus without. So we are having an impact on this thing that people previously thought was maybe not so easy to do.

**Jay Baxter** (00:22:59):
And so, it's nice to focus on the cases where there is the bridging agreement. I would also say there is this reputation component to the algorithm as well. So if you consistently rate notes in a way that is counter to the bridging-based consensus, then we will stop counting your ratings. So if you're the kind of person who constantly rates bad notes as helpful, we do filter you out. So there's a difference between those types of people versus the good but polarized ones.

**Keith Coleman** (00:23:30):
Yeah. I think one philosophical thing that's important is that we want all of humanity to participate. And sometimes, people are surprised by that. They'll be like, "Oh, aren't there people who shouldn't be doing this?", or like, "Their thinking is so extreme or something, maybe they shouldn't participate." But our view is it's actually we want to have all of humanity here. Because if we have all of humanity, we then have the data to understand what notes will be helpful to actual humanity. We can better model that better or better understand and better show those notes.

**Keith Coleman** (00:24:03):
So it's advantageous to have people who have all sorts of points of views and we don't expect that every note will be loved by every single person. That's an impossible bar. But we do intend to show the notes that 80% of people are going to read and say, "Wow. I'm glad I knew that." And so, in that sense, it doesn't matter how maybe extreme someone views a person's views as. It's still great to have them in the program. So no matter what your views are, please sign up and participate. It helps identify what's really helpful.

**Lenny Rachitsky** (00:24:39):
Cool. And we'll link to people if they want to actually sign up, so they know how to do this. Something we didn't actually specify, these are all volunteers. No one's getting paid to be doing these notes and voting, right?

**Keith Coleman** (00:24:49):
Yeah. It's totally based on intrinsic motivation and we think that's a great reason to be doing it. When you talk to the most active contributors, a lot of them, they want to have better information out in the world and that's a great motivation. So yeah, that's why they... If you think about, like for these people, the impact they can have is nuts. So when we first launched US-wide, this was like in 2022, a note appeared on a White House tweet and the White House deleted the tweet and reissued an updated statement.

**Keith Coleman** (00:25:25):
Imagine being the person who wrote that. You probably have 12 followers. Your posts probably get a couple likes. And here, you just put a note on the White House and they changed their public talking points based on what you did. That is an incredible amount of impact. So you could see why people are motivated to do it when they care about what's going on in the world. You don't have to be a big, well-known person to shape the discourse and information flow in a way that's helpful.

**Lenny Rachitsky** (00:25:59):
It's insane. There's so much to love about this. One is just the meritocracy of this whole operation of just anybody that is true and correct can participate and have impact. Also, it just shows you how much information we get that is just wrong. We had no idea how often we see things that are wrong and now we do.

**Keith Coleman** (00:26:18):
Working on this product has made me realize just how many things I used to trust by default, that now I look at more skeptically.

**Lenny Rachitsky** (00:26:26):
Definitely mean these days. Okay. Before we get to the origin story, is there anything else along those lines you guys think might be really important to share, that are really interesting?

**Jay Baxter** (00:26:36):
Sure. I guess one other thing is that although we don't actually use the fact that a post was noted in the core ranking algorithm, which we think is a nice property. There is a really big impact just organically, meaning not from the algorithm but just from user behavior, where people will like and re-share or quote posts way less when-

**Jay Baxter** (00:27:00):
Quote. Posts way less when notes are applied. I don't know, for people out there who typically run A-B tests on big platforms, you may already be familiar with this, but 1% is typically an awesome effect size for any algorithm change. We saw more like 30 to 40% engagement rate drops for likes and reposts in A-B tests we were ran when showing a post with or without a note, which is just crazy big. That's just an A-B test on the engagement rate, so that's not the network effect. If you capture the overall network effect of how post spread less by that person's repost, basically if you look top line with a difference in differences approach, multiple different external research groups have both found consistently that there's a 50 or 60% drop in total reposts, which is just nuts after a note is applied. It's having a really big impact on spread actually, too.

**Lenny Rachitsky** (00:28:05):
That's so great to hear. It's what I would want to see and it's incredible impact. Basically, an AI image of something false would just go crazy on Twitter, and did before Community Notes came out, and now what you're saying is just adding that context, not actually... Like you're saying, the algorithm doesn't demote it. If there's something incorrect, it's just people are like, "Okay, this is false, why would I want to retweet this?" That makes sense.

**Keith Coleman** (00:28:28):
Correct.

**Jay Baxter** (00:28:29):
Right.

**Keith Coleman** (00:28:29):
Yeah, the notes just totally take the wind out these stories. The thing will be going viral, note appears, resharing drops 50 to 60%, and that's it. At 50 to 60% per generation, the virality quickly goes to zero.

**Jay Baxter** (00:28:45):
By the way, I have very mixed feelings about this next one, but authors become 80% more likely to decrease, sorry, to delete their post after they get noted, which okay, that's great, because less misinfo out there, but I'm pan about, because those are usually the best notes. If the note was so just good that you had no other option but to delete your post, those notes don't get seen by other people, right? Because-

**Lenny Rachitsky** (00:29:13):
That's hard.

**Jay Baxter** (00:29:14):
There's an argument, by the way, that seeing... Just because you might see the same misleading claim elsewhere off X, or somewhere else on X, it might be good to actually show... Better to have seen the post with the note than not see it at all.

**Lenny Rachitsky** (00:29:28):
Yeah.

**Jay Baxter** (00:29:29):
Unsure about that claim.

**Lenny Rachitsky** (00:29:31):
That is so interesting.

**Jay Baxter** (00:29:32):
Yeah.

**Lenny Rachitsky** (00:29:33):
Yeah, I'd be so sad if I was that community note writer and just... Man, it's so good. They just can't even keep the post up. Okay. Coming back from today's world, where this small amount of code is changing the way people understand the world and what they believe, and making the White House rescind their announcements, zooming back to the beginning of how this whole project started, what I heard just briefly is, Keith, you were just tired of managing PMs, you wanted to just work on something yourself, you wanted to work on something impactful away from corporate BS, and you basically just started looking for something that was impactful, important, and you found this. Talk about just how it all came to be at the beginnings of the story.

**Keith Coleman** (00:30:19):
Yeah. I mean, for me, the beginnings actually go back to why I joined, it was then, Twitter in 2016. I had a startup and we'd had some acquisition offers, and one of them was from this company, Twitter. It was 2016, it was the middle of the election between Donald Trump and Hillary Clinton, and there were something like three televised debates, but every day, there was a debate happening on Twitter, and it was very clear, this is where people are talking about these things that matter, where information is being shared, where ideas are being formed. As a user, it was obvious that I could get good information there, but it was also obvious that there was questionable information floating around. I remember just looking, as an outsider, thinking like, "Wow, this is a really hard problem and it also seems really important," so we ended up going to Twitter and the company was in a turnaround at that point.

**Keith Coleman** (00:31:21):
My first three years was just helping to get the company growing again, working on everything that was the consumer product, getting user growth going back and people wanting to work there again, et cetera, but a few years in, I was reflecting on what we had done. I think we had done a lot of good work getting momentum going, and people in the us and in the industry had tried things to deal with misleading information, but nothing was really working. It was obvious nothing was working. Nothing could handle the scale of the problem, nothing could handle the speed, and a lot of people just didn't trust the existing approaches. The existing approaches were either fact-checkers or internal trust and safety teams making decisions about what was or was not misleading. A lot of people just didn't want or trust that to be the way this was decided, which is very reasonable.

**Keith Coleman** (00:32:19):
I'm looking at that, I was still managing a large PM team. That's a whole story in itself. That job required a lot of energy in, and I didn't feel like I always saw the output that I wanted to see from it. I didn't see the change in the product I wanted to see and I was contemplating, "Should I go start a company? Should I do something else?" And I kept coming back to this problem. I'm like, "Man, how is the world going to deal with this information quality issue of what we get on social media?" Wherever get it. I'm at this company where you can make a difference on this problem, why not go and try some crazy ideas and see if one of them might work? I had a kid, I came back from paternity leave, I went to my boss, Kayvon. I was like, "Hey, Kayvon. How about I just stop doing my job and I go work on this instead? 'This' being trying some crazy ideas to see if we can deal with misleading info."

**Keith Coleman** (00:33:24):
He was stoked, so I went off and started working on that. It started with just reading any research I could on the problem and existing solutions. What was or was not working, what were the issues, and then into prototyping. Then it ultimately led to us building and piloting this idea that became Community Notes.

**Lenny Rachitsky** (00:33:46):
Amazing. I have so many questions and we're going to keep going through the story, but when you joined Twitter, what was the... It was called Twitter. At this point, I'm going to try to call it X now, which I know is important to your boss. What era of Twitter was it at that point? It was Kayvon joined and who was the CEO? Because there's been many.

**Keith Coleman** (00:34:05):
Okay, yeah. I came in December 2016, so Jack had relatively recently come back as CEO to turn the company around, and just to give you a sense of the state of the company, something like a third of employees were leaving every year. Just imagine a third of your team gone every year. The stock was in the toilet, the product was not really growing, so Jack was working on a turnaround and Kayvon was there already. Kayvon was running Periscope with a bunch of video stuff, and that group continued to... Jack was there up through the start of the Community Notes, then Birdwatch Project, and... Yeah.

**Lenny Rachitsky** (00:34:50):
Okay, and it was called Birdwatch. I don't think we've used that term yet, but that's an important point. It was called Birdwatch initially.

**Keith Coleman** (00:34:55):
Yeah. It was originally called Birdwatch when we started the project, but obviously, somewhat famously the name changed along the way.

**Lenny Rachitsky** (00:35:05):
Yeah, maybe let's just tell that story real quick, and I know we're zooming it forward, but just... I have this Twitter thread that I saw between Jack and Elon when they're debating what to call it, and Elon's like, "Birdwatch sounds creepy, I want to change it". Is there anything there you can share?

**Keith Coleman** (00:35:19):
Yeah, the story there... The story, that's funny. Elon came in, acquired the company, and we had just launched the product relatively recently in the US. It had been in pilot for a year, but we had just made it available US-wide, and I guess he'd been seeing the notes. Soon after the exhibition, he DM'd me and he was like, "Hey, this Community Notes thing is awesome," and I was like, "I'm glad you like it, let's talk," so we talked the next day and he kept referring to it as "This Community Notes thing." I was like, "It's interesting you keep calling it that, because that's actually the very first thing that I called it." The very first figma mockup I made depicting this thing was called "Community Notes." I don't know why, it just felt really natural, so that's the first prototype we had tested.

**Keith Coleman** (00:36:14):
Later, the project changed the same to Birdwatch, but Elon was like, "Hey, let's just call it that." The next day, we just changed the name. It's always notable for the team when you change the name, but really, the team was excited about it. I think it is a much more understandable name. Jack has made fun of it, calling it "The ultimate Facebook name," or something like that.

**Jay Baxter** (00:36:41):
The most boring Facebook name [inaudible 00:36:44].

**Keith Coleman** (00:36:44):
Boring name, which is funny, because they're now launching Community Notes. I think it is a very understandable, intuitive name, and I think it has served the product really well. There's a reason it was the name in the very first mockup.

**Lenny Rachitsky** (00:36:57):
Yeah, I think descriptive names just makes sense. This connection with Elon, and I want to talk later about just how you've dealt with so many strong personalities over and kept this alive throughout so many changes, but before we get to that, you did something that I think a lot of product leaders, angel leaders, just people that have managed people dream of give up all this power, in air quotes, and career trajectory and influence and just, "Forget all that. I'm going to go back to just building something awesome, small team." Is there any advice there that you could share from that experience that you think might be helpful for other leaders to share or to hear to help them maybe do that same jump? Because that's really difficult in practice. Easy to talk about, hard to do.

**Keith Coleman** (00:37:42):
Yeah, I think it is a difficult jump. I've done it a bunch of times in my career and I've always been very happy with it, where I started with a small team, that it grew into something bigger, and then I was like, "We're dealing with a lot of big production stuff, team's really big. I want to go back to doing something like crazy and new with a small team again." I've done that sawtooth leap a bunch of times, but it can be hard, because certainly, the natural... The classic career path is, I don't know, rewards or running a large organization or being a manager, or things like that, but I think, at the end of the day, you got to work on stuff you love, you got to be having fun, and I think people want to be having impact.

**Keith Coleman** (00:38:29):
I think there's one myth that can get in people's ways. The idea that the more people you manage or the larger your scope is, the more impact you have. I definitely do not think that is true. I mean, look at Community Notes for example. If I had stayed running a large consumer PM team, what would I have produced? 16 more pages of OKRs? I don't know, a bunch of documents? I think building Community Notes has had way bigger impact on the world. It's become the industry standard for how to deal with this now, which is super cool. People love it, it's the first thing that is plausibly dealing with the internet-scale issue of information quality. I think it's unquestionably a bigger impact than I would've had if I were just doing whatever, doing some standard management track thing like I was doing before. I think that's true of so many other small companies and startups. Someone screenshotted I think it's Blake Scholl's LinkedIn the other day. He went from director of coupons or something to building the first supersonic-

**Lenny Rachitsky** (00:39:37):
Yeah, from Groupon.

**Keith Coleman** (00:39:41):
Those stories are everywhere when you look, so I definitely have found that, for me, I love building hands-on, I love trying crazy new ideas. I love the zero-to-one experience. It's fun to scale things up too, and it can be fun to operate at scale, but this team is a good example of one that operates at a very large scale, but that is still very small.

**Lenny Rachitsky** (00:40:03):
Yeah, I think the way you guys operate is what more and more companies are trying to do, remove middle management layers, create small teams that just execute and build impact, just like Ics. Whenever I say IC, I have a comment on YouTube, where like, "What is IC?" I'm just going to explain, individual contributor, non-manager is when I say the word IC. Let me follow this thread, and when I asked people about how you set up the team to operate effectively and protect it initially, there's this term, "Thermal," that came up a lot. It was like a thermal team, if that's how you describe it.

**Keith Coleman** (00:40:37):
Yeah.

**Lenny Rachitsky** (00:40:38):
What is thermal?

**Keith Coleman** (00:40:39):
Yeah, so anyone who's worked in a larger company probably knows that things can get bureaucratic or bogged-down, decision-making can be slow. There's these large planning cycles, people can try to take someone from one team, move them to another at random arbitrary times that can disrupt a project, all sorts of things like that. Our company, this is a number of years ago when we started this project, we had a lot of founders in the company. Kayvon is an example of founder who is helping to run the company, and he had this idea, "Hey, why don't we create this program, call it Thermal, where we could have teams that were somewhat isolated from that." They could run through their own process, they would have one clear owner. The team would be entirely dedicated to that project and we would just repeatedly make funding decisions as to whether to continue the effort.

**Lenny Rachitsky** (00:41:31):
Why was it called Thermal, by the way? What was the idea there?

**Keith Coleman** (00:41:35):
I think it was an old bird analogy of thermals lifting the bird on their wings. Twitter 1.0 obviously had a lot of bird analogies, bless its heart, so that was one of them. I loved the idea, as someone who liked the startup environment, so when we were starting this project, I was like, "Hey, Kayvon. Why don't we make this the first Thermal project?" And he was like, "Yeah, let's do it," so we started with that way of operating and it gave us, from day one, a lot of freedom and autonomy that I think was really important to make the product work.

**Lenny Rachitsky** (00:42:15):
Just be very specific about it. What makes it a Thermal project? How do you set that up? This is asking from perspective, if a company wants to build their own something like this, what does that look like?

**Keith Coleman** (00:42:24):
Yeah, I think there's a bunch of key attributes. One key attribute is there's one clear driver of the project, who's effectively a founder. I guess maybe you could have two or something, but really clear, there's driver of the project and also there's one clear decision-maker that they go to.

**Lenny Rachitsky** (00:42:43):
Outside of the team?

**Keith Coleman** (00:42:44):
Outside of the team. That was true back when we started and it is true now. If we need something or have a question about something, I talk to Elon. It was like that from the beginning, it's like that now, and I think that's a big reason we're able to make decisions effectively, quickly, in a simple way.

**Lenny Rachitsky** (00:43:02):
It probably has to be someone very senior, not [inaudible 00:43:05] manager.

**Keith Coleman** (00:43:06):
Someone senior who can make the decisions you need made, whatever they are. I think that's really important, that clear decision-making structure. Another was 100% focus, so everyone on the project is expected to be totally focused on it. A lot of companies, it can be easy to have people's attention spread across a bunch of things, and it makes it hard to get stuff done. You'll talk to whoever that person is, you'll ask them for help on something, and they'll be like, "Yeah, I'll help you. I got to finish this thing, and it'll take me a week or two and then I'll get to it." A week or two delay totally changes the momentum of a project. When we were 100% focused, we talk in the morning, it's like, "Hey, Jay. Why don't we try this thing in the algorithm?" He's like, "Yeah." Then that afternoon or the next day, we're looking at results.

**Keith Coleman** (00:43:59):
Because of that total focus, the rate of iteration goes way up. Then beyond that, there was also just the ability to use whatever our own decision-making process was. We didn't need to write OKRs or... For others standard practices. Obviously, we had to make sure we were responsibly building the product and everything, but we didn't need to use the standard practices. I think that's another great example, OKRs, I understand why they can be helpful, but they can also be not necessarily the right cadence at which to set goals. I think it's really unclear that quarterly or annual goals are actually the right pace. We would set the goal for the next milestone that mattered, and we would work on that. We reached that milestone, we would have an idea of what was coming after, and then when we hit that, we'd set the next milestone. Whether that was two weeks, a month, three months, whatever it was. We set our own pace and goals at that pace, and that just I think is a lot more natural for the development of something.

**Jay Baxter** (00:45:06):
The whole OKR determination and planning process took longer than it would take us to pick a goal and then execute on it and finish it.

**Lenny Rachitsky** (00:45:15):
How big was the team early on that you set up? How many engineers?

**Keith Coleman** (00:45:19):
It started with just me and then, when we decided to build the thing, we figured we needed about five. We wanted it to be as small as we possibly could. It was clear we needed someone on ML doing scoring, it was clear we needed someone to do some client engineering work, someone to do backend engineering work. There may have been one or two other. We needed a designer and a researcher to help us understand the customer base and make sure we were building the thing in a way that was actually going to resonate with people. I think it was backend, frontend, ML, design research. That was the original team, from what I remember.

**Lenny Rachitsky** (00:46:01):
Amazing. Basically, one of each function. A question I have for Jay, actually, is there's all this talk of small teams and moving fast, but sometimes you just need more engineers to build the thing. Is there anything you've learned about just how to keep a team small while moving as fast as you are, and not need or need to hire more engineers?

**Jay Baxter** (00:46:22):
I think, in the beginning when we were iterating on what should even the requirements be, it was definitely good to just have one ML engineer, but I think, at some point, we got clear on what the goals of the algorithm should really be and we tried... I think, at the very beginning, it wasn't clear that we needed to build this bridging-based algorithm. The actual first algorithm that I put into production was very focused on anti-manipulation. It was this page rank variant, but it didn't solve the problem of bias, basically. If there are more users on one side, a page rank type graph algorithm can actually amplify those biases. I think, after building that prototype and getting data from that, it was clear that the bridging-based algorithm was going to be the way that we needed to solve it, and at that point, basically I set up a bake-off. Kind of a Kaggle competition or something. That was the key time where it was really important to pull in other engineers.

**Lenny Rachitsky** (00:47:34):
That is such a cool story. I want to follow that thread. Before we do that, you just mentioned you guys yell "Thermal." What does that mean? Is that YOLO, like a version of... Okay.

**Keith Coleman** (00:47:43):
We're just going to ship, because we're thermal project.

**Jay Baxter** (00:47:46):
Ship it.

**Lenny Rachitsky** (00:47:47):
Okay. Marketers, I know that you love [inaudible 00:47:52], so let me get right to the point. Wix Studio gives you everything you need to cater to any client at any scale, all in one place. Here's how your workflow could look. Scale content with dynamic pages and reusable assets effortlessly, fast-track projects with built-in marketing integrations like Meta, CAPI, Zapier, Google Ads, and more. A-B test landing pages in days, not weeks, with intuitive design tools, connected tracking, and analytics tools, like Google Analytics and Semrush. Encapture key business events without the hassle of manual setup, manage all your client's social media and communications from a unified dashboard, then create, schedule, and post content across all their channels. If you're working on content-rich sites, Wix Studio's No-Code CMS lets you build and manage without touching the design. When you're ready for more, Wix Studio grows with you. Add your own code, create custom integrations with Wix-made APIs, or leverage robust native business solutions. Drive real client growth with Wix Studio. Go to Wixstudio.com.

**Lenny Rachitsky** (00:48:52):
Okay, so coming back to this algorithm, this is actually really interesting, because I've never heard any of this. I was going to ask just what inspired this actual algorithm, and you basically did an internal competition amongst ML engineers to see who had the most successful algorithm. Netflix-contest style, Kaggle style.

**Jay Baxter** (00:49:09):
Yeah, yeah. This particular idea of finding content that is liked by people on opposite sides of a polarized divider who typically disagree, this was not an idea out of thin air. I think Keith had found some of Chris Bale's work, he had made this list of accounts that were often liked by people who were on both sides politically. There is other projects, like polls out there that look for agreement among people who typically disagree, but I think that it wasn't obvious that our project definitely needed to use that from the very beginning. When you implement it and compare it against these other type... PageRank seems, obviously, it's designed to be manipulation-resistant. Naturally, if you just have a voting ring of people who all vote themselves up, then PageRank can filter that out very well, but that just wasn't the main attack vector, I guess.

**Jay Baxter** (00:50:15):
We had to get some real data from the pilot to realize that, "Okay, the real thing going on here is people are polarized," so it was only once we got that, the real data from the pilot, that I think it was clear that the bridging-based algorithm was the direction we really needed to go.

**Lenny Rachitsky** (00:50:34):
I want to come back to the way you operate the team. I hear that you run the whole team off a single Google Doc that's like a four-year-old doc that you just keep adding goals to, bullet points. Is that true?

**Keith Coleman** (00:50:47):
There is a very long-running doc that has had to be chopped and purged, because it was breaking Google Docs in Chrome at various points in time. It's like a note-taking doc. It's really where we coordinate what we're doing. The team meets on a daily basis, we spend whatever amount of time we need to get on the same page about what we're building. We might talk about anything from what's most important right now to, "What should we work on next?" To, "What are we trying to launch right now, and why is it not launched? What's in the way of launching it?" We might review new modeling or scoring algorithm update and try to understand what's working in it, what's not. We'll just cover whatever we want or whatever feels most important. As you said, we set our goals very dynamically, so whatever seems like the most important thing for us to work on now and next is what we spend our time on. I think that's served the project really well versus feeling attached to some quarterly goals, or something. We'll look at, "What is going to help people the most?" Or, "What's the biggest problem right now?" What are either one of those? And we will go tackle it. We might change our roadmap multiple times in two weeks based on what we see.

**Lenny Rachitsky** (00:52:08):
I'm hearing no Jira, no Asana, no Monday.com.

**Keith Coleman** (00:52:11):
No.

**Lenny Rachitsky** (00:52:12):
Okay.

**Keith Coleman** (00:52:12):
Yeah, I mean, we have to use Jira to coordinate with some other teams. Sometimes when we file a request, we have to make a Jira ticket. But no, I am not a fan of heavyweight task management. I love being on the same page, being able to keep most things in my head, and having a really light way to write down the things that the team can't keep in its head.

**Jay Baxter** (00:52:34):
We did use Asana briefly, but my memory of it is that you spent more time in the meeting grooming a backlog of irrelevant stuff than actually talking about the proper priorities. I think it's nice in the Google Doc that, if something becomes irrelevant, it can just fall off without needing explicit backlog grooming.

**Lenny Rachitsky** (00:52:58):
Just to maybe summarize a little bit of how you guys operate that might inspire other companies to set teams up like this, so I'm going to go through a few things you shared. One is one person in charge of the team, like the founder almost. They're basically the founder of the team. They have one very senior, essentially, sponsor/decision-maker that they interface with. In your case, Elon, no big deal. In other cases, it could be the CTO, CPO, someone like that. The team is focused 100% on its product and goal. You keep the team very small, so you start with one person of each function. One front-end engineer, back-end, ML person, designer, researcher, PM, and then Google Docs is almost basically for your project management. Yeah, it's basically run with Google Docs, stop, don't use big, complicated products.

**Keith Coleman** (00:53:50):
I think that's a pretty good recipe. On the Google Docs, people can do what they want. If they want to use thumbnails, go for it. I think those first ingredients, really, are key structurally. Then beyond that, it's a matter of having an ambitious goal that gets-

**Keith Coleman** (00:54:00):
And then beyond that, it's a matter of having an ambitious goal that gets people fired up to go do great work.

**Lenny Rachitsky** (00:54:06):
Yeah. Awesome. I think there's a lot there that a lot of people think they should do when they set these teams up, but they don't actually do, and it feels like each of these is just a really key ingredient to it to actually succeeding.

**Keith Coleman** (00:54:17):
It definitely really helped us succeed. I don't know that the project would be here if it was not for some of those elements.

**Lenny Rachitsky** (00:54:24):
That's a powerful statement. This thing that has changed the way the world understands what is true would not have existed if you didn't set it up in this specific way.

**Keith Coleman** (00:54:34):
Yeah. I don't know if I would've begun the project had I not known. We had that structure, that ability to make decisions, the autonomy, the speed, the ability to go fast. We started with that in 1.0 and it's been continued and if anything, furthered in X. X as a whole company operates with a lot of those attributes, and I think it's one of the reasons the product is successful. I think those are big reasons why at least, Jay can speak for himself, I have so much fun working on this. I love working on it. It's great to wake up every day and solve these problems. We get to do them efficiently, make decisions quickly, build stuff that helps a lot of people. It's awesome.

**Jay Baxter** (00:55:25):
This whether thermal or Elon way of operating is definitely more fun and the fact that... That combined with the awesome mission is super important for internal recruiting. I remember when I was first chatting to Keith about this back in early 2020, I had another project. I worked on a few, but one was like personalize the number of push notifications that we send, and it drove a lot of DAU without losing opt-outs significantly. So that was setting me on track, or if I had kept working on that, I could have probably gotten a promotion from that with low risk, or I could take this huge career... It's not as big at a career risk as joining or founding an actual external startup, but there is still career risk, I guess, in joining a team like this. I think all of the same aspects of recruiting that apply to external startups and apply internally, and if you can have an exciting vision, that is key.

**Keith Coleman** (00:56:30):
Related to that and your list, Lenny, one thing we missed that's super important is that on this project, and I think of successful projects like it in startups, is that people are self-selecting to join. We did not assign anyone to this project. People reached out to join or they applied to join the job. I and the team interviewed every single person that joined the team and we were like, "We want that person on the team. They want to be on the team." And so people are totally bought in to the goal, mission, the way the team works, the other people they're going to be working with. And that makes a huge difference.

**Keith Coleman** (00:57:10):
So a great time to do that is at the start of one of these things. If you're going to try something crazy, it's going to be tough if you're just assigning random people to it. But if you let people opt in and self-select much more likely to be successful. And one thing that I have observed at X, which really surprised me was that this is also possible at a large scale. One of the things that Elon did when he bought the company was he basically asked people to self-select to stay. You had to click the button. And he sent an email out that was like, "Hey, Twitter 2.0."

**Lenny Rachitsky** (00:57:44):
Fork in the road. Right? [inaudible 00:57:46]

**Keith Coleman** (00:57:45):
Fork in the road. Fork in the road. Exactly. He's like, "Twitter 2.0, now X, it's going to be hardcore. We're going to do ambitious things. You're going to work your butt off." And you had to click on the form and say, "Yes, I want to join." And I think that was really important for the company because you want people to opt into that. You want the people to be saying, "Yeah, that's what I want to do," and the company's going to be a lot more successful. If people aren't sure, it's better for them probably to go do something else and where they're naturally more aligned and happier. And I thought that was a great approach to taking a large company and getting it down to people who are really excited about working together on a mission. So for us, we did it from day one, which I think is an easy way to do it, but it's possible to do it later as well.

**Lenny Rachitsky** (00:58:33):
I love that you described it as fun and I think a lot of people when they see Elon laying off a bunch of people, being very hardcore himself, people don't imagine it as a fun place to work. And it's clear how much you guys love working on this, how fun it is and how interesting it is. And it's interesting to hear that 'cause I think a lot of people don't feel that externally. Is there anything else along the lines of just working for Elon within an org Elon runs that might surprise people about just the way of working that's interesting or surprising or you think other companies might want to think about adopting?

**Keith Coleman** (00:59:07):
I've always liked lean teams, but my experience at X has made me change the way I would think about running a future org-... If I were to start a company and had to change the way I think about starting that company, I would be even leaner than I would've made it before. I've been amazed with just how much the team is able to accomplish with a small group. And I think because of a small group, shortly after the acquisition, we had this product called Spaces. It had been in the product before, but it was pretty small scale, and Elon wanted to run these large spaces. I forget who the first people he was going to bring on were, but he was going to be there. Ultimately, these things have gone on to host politicians and things like that, and he's like, "Guys, we got to scale this up." I forget the numbers.

**Keith Coleman** (00:59:57):
He's like, "We need to be able to scale a million people," or something like that. I'm getting the numbers wrong. "You need to be able to scale way up." This is the kind of thing at 1.0 That would've taken a year if it had ever happened, and the team did it in two or three weeks. And it was really exciting and inspiring to see. I didn't work on that, but I watched it from the outside. I'm like, "Wow, with this tiny team motivated behind a big goal that was like, 'Hey, guys, it's not like, are we going to do this?' It's, 'We are going to do this.'" They got it done in two or three weeks. That must've felt amazing for them. It was certainly exciting to see. But I've definitely come to appreciate just how lean something can be and not just get by but actually thrive because it's that lean.

**Lenny Rachitsky** (01:00:42):
I think the point you made about people opting into that is important, 'cause I think a lot of people hearing that would be like, "I would never want to be asked to build something like that in two weeks." And I think a lot of people do, and we love that kind of experience, especially working with the Elon, especially shipping something at that scale. But I think there's an important element there of just like, "Okay, I don't want to do that. I have other things to do in my life other than ship spaces." So I think that's a key point you've raised of just there's an opt-in step.

**Keith Coleman** (01:01:10):
Totally. I think the opt in is important, and it may even be that you want to opt in at one point in your life, and maybe at another point in your life something else is better. I think whatever it is you're choosing to do, it's nice to be opting in to feel like it's aligned with how you want to spend your time.

**Lenny Rachitsky** (01:01:25):
Something on my mind, and I don't know if you guys want to go here, but it's something I think a lot of people think about is when Elon came in, he let go of 80% of folks. And everyone's just like, "Twitter is dead. It's all going to fall apart. There's no way they can run this thing with that small of a staff," and clearly they were wrong. Clearly, it's working great. It's becoming a massive deal in the world and continues to grow. Is there anything about that that you were surprised by or anything about just how it continues to operate so well in spite of that big shift?

**Keith Coleman** (01:01:57):
I think the leaner team, the reduced process in bureaucracy is a big reason it does move as fast as it does. It's easier to get stuff done faster here. Yeah. I think that shrinking is actually a big reason for the increased pace of launches, the increased pace of experimentation. One thing that I noticed a result of that is the people who are here, they seem to all really feel like owners. They take the sense of responsibility that an owner takes in the product. They'll try to track down what's wrong, fix whatever is needed, jump in to help build or fix, improve any system that needs help, even if it's outside of their space. And there's the flip side of that too. For people who've worked at big companies, they may have experienced this thing where there's like ano-... You want to change something in some other system or product, and so you reach out to that team. And maybe they're a little resistant, they'll maybe be like, "Oh, we'll get to that next quarter or so-

**Lenny Rachitsky** (01:03:07):
They have their own goals to hit. Yeah. [inaudible 01:03:08]

**Keith Coleman** (01:03:08):
Yeah. Exactly. They don't really necessarily want to help you or they're busy. Here, you're like, "Hey, guys, we need to do this thing with that other system you work on." And they're like, "Great! Here's the code. Here are the docs. Send us the fab if you have any questions, and we'll get it in." And it's just the thing, you can just jump in and get it done. And that kind of collaborative effort, like the sense of shared ownership, I think from my experience came from or was a result of the shrinking of the team down to people who wanted to be there and work together to build this thing. So I think that's been a really positive impact. It's not always easy. Certainly, a lot of people have a lot of responsibilities, but they're here because they're up for it.

**Jay Baxter** (01:03:53):
Yeah. I think one other thing that's key is when you are forced to have such a small team, well, this is important anyways, but deleting code is more important than writing it a lot of the time. So I think so often maybe due to promotion incentives or just regular human tendency, engineers have a tendency to add these little incremental wins that actually add more of a long-term maintenance cost than is clear, because you just run a little one month A-B test, you see this significant win and you don't realize the maintenance burden you just added to your team for the rest of eternity until you turn the thing off. So I think there's a lot to be gained and you get forced to do this, by the way, when you have such a small team. It's just auditing parts of your system and deleting the things where the maintenance cost is worse than the gains. So I think we did have to do this across the company after the big layoffs, and systems are leaner now and they can be worked on by fewer numbers of people.

**Lenny Rachitsky** (01:05:02):
That's an amazing point. I remember Elon's being like, "Here, we have to throw away the whole thing. We have to re-architect everything. It's stupid the way it's built." And it sounds like that actually worked.

**Jay Baxter** (01:05:10):
Yeah, so-

**Lenny Rachitsky** (01:05:10):
Well.

**Jay Baxter** (01:05:11):
You don't have to rewrite everything from scratch. Some things are good, I guess, to rewrite. But just even deleting the unnecessary cruft and keeping the rest of the core system, that's awesome.

**Lenny Rachitsky** (01:05:23):
I love that we're creating a formula to run these sorts of companies and teams. There's so much here. I want to go back to the building of the original product. I took us on a long tangent and an amazing tangent, but I heard a story of when you launched Birdwatch at that point. You specifically wanted to keep expectations very low and there was a GIF in the thing, and it just looked like clearly this is not ready for prime time. Talk about just how you did that, how you launched it in a way where people weren't like, "It's never going to work."

**Keith Coleman** (01:05:53):
We were very disciplined, I guess you could say, about having the product prove itself at every given point. When we built the first mockups, these were just pictures of depicting what community notes might look like. We showed those to people across the political spectrum. We saw, hey, people really like these. Whether they're on the right or left, they seem very open to reading these community notes even when they're critical to people of their own side. So we're like, "All right. That gives us confidence that if we can build this, if we can actually make this as a reality, it's going to work." Then there's a question of can we make it a reality? Will people in the real world be able to write notes that are of this quality?

**Keith Coleman** (01:06:35):
And so we had an internal pilot test version of this where you could write notes. And we first basically ran this through an Amazon MTurk type of participant test just to see if you just put some normal people in there, will they be able to write these notes? All those notes weren't good, but it was clear that there were people out there who could write good notes. So then like, "Okay, this is possible. What will happen if we actually do this out in the real world? And let's run a pilot and find out." And so we took that pilot that we'd run the MTurk of test on, and we released it to at first 1000 people, totally out in public, and we didn't know what was going to show up. You could imagine the notes could have been terrible.

**Keith Coleman** (01:07:27):
And so we were talking, "Well, what do we do? We're going to put this out there. Everyone's going to have all these questions. They're probably going to be really skeptical, and we know it might be a total dumpster fire. And so what do we do to set expectations appropriately?" We felt like we could probably get there in the end, but we just didn't know what was going to happen at first. We wanted to set expectations, and so we're like, "Well, why don't we just stick..." There's the page where you see a post in the notes below. We're like, "Why don't we just stick a dumpster fire GIF on that page?" And you go there, you're like, "Hey, anything you see below here might just be a total dumpster fire. At least it would show we were aware of that as a possible risk." In the end, we did not do that. It cracked me up, but we thought it was like-

**Lenny Rachitsky** (01:08:13):
Oh, you didn't actually launch. Okay. That was just a concept. Okay.

**Keith Coleman** (01:08:16):
We had mockups of it, and every time I looked at the mockup, I laughed, but ultimately we had so much to explain on that page, like, what is this thing and how does it work? Ultimately, we're like, "Okay, this is probably going to distract from the point." So we pulled it. I wish maybe it had seen the light of day at one point, but yeah, ultimately we kept it simple and we focused that page on explaining what was going on here. But again, as has happened many times with the project, we put the pilot out there and the notes were good.

**Keith Coleman** (01:08:48):
They weren't all good. It was a mixed bag, but there was gold in there. And from the very early days with just 1000 contributors, it was obvious that people could write notes that were informative, that were neutral, that spoke to controversial challenging topics, and that if we could just identify those from the rest, this was going to work. It was going to work as well as the very first mockups we had made. So that became the focus that is, how do we sift out the gold from the rest?

**Lenny Rachitsky** (01:09:19):
I think you may have shared with this with me, when someone noticed you guys were testing this and they took screenshots and tweeted it, and I think Elon replied, "This is cool."

**Keith Coleman** (01:09:27):
Yeah. Yeah. So in the very early days when it was just a Figma prototype, we were running these usertesting.com on moderated studies. I guess one of the participants sent one to an NBC reporter who wrote a bunch of stories on it. Anyway, that day, there was a lot of chatter about it on the service, and Elon... To put this back in time perspective, this is, I think, 2020, so two years before any acquisition stuff happened, Elon is just a Twitter user building rockets and electric cars and other cool stuff and stumbles on this thing that depicts the prototype that we've been testing. And he writes back, "Definitely worth trying, IMO." And I remember thinking that was cool back then and it's interesting to see, he's obviously had a very consistent point on it. I think the idea was appealing and he has obviously been a big fan of it in the product and had been a big supporter proponent. So yeah, it was cool that it came from... that support has been from the very early days before he was ever involved in the company.

**Lenny Rachitsky** (01:10:36):
I love that moment. That must have felt really wild for Elon to be commenting on this Figma prototype retesting.

**Keith Coleman** (01:10:42):
It was cool. It was cool.

**Lenny Rachitsky** (01:10:44):
Oh, man. So when we were preparing for this interview, I asked you guys what's the main thing you want to make sure people get and understand about why community notes has been so effective? And Keith, you specifically said that it was the principles behind how you wanted to approach this and how you continue to stick to this throughout. And we'll talk about how you kept it alive throughout all these different CO changes in leaders. But just talk about these principles, what the actual principles are and why that was so key to it working out.

**Keith Coleman** (01:11:16):
There are a number of principles that I think when we first shared them with people at the company seemed maybe a little bit crazy. But I think they are the reason the product works, and I think they've been very important, and we do. We come back to them regularly, today, all the time. Probably the craziest one is just that this thing is going to be the voice of the people. It's going to represent the voice of people. It's not going to represent the company's voice. So it is not a tech company deciding what shows. It is the people deciding what shows, and that had a lot of implications on the design. First of all, we don't have a button that will change the status of a note. So if a note is showing because the people have rated it and found it helpful, it is going to show. We can't change that.

**Keith Coleman** (01:12:08):
And that is the kind of thing that when we first propose this, that's unsettling to people. They're like, "Wait, so something can go up and the company can't take it down, or can't change its status, get it to stop showing." And we're like, "Yeah, and it has to work that well. If it doesn't work well enough to do that, then it doesn't work." This is one of our key principles was, if there's a problem with a note that's so bad, you want to do something about it's a problem with the system. We need to redesign the system to be showing good notes. And so yeah, we had to get everyone comfortable with the idea that there was no button to change the status of a note. Similarly, as we talked about earlier, we wanted this to represent all of humanity.

**Keith Coleman** (01:12:53):
And so we didn't want to be arbiters of who can come in and be a contributor and who can't. So we open it to everyone. You just have to meet a really basic objective criteria. You have to have a verified phone to help reduce the likelihood of having bots or things like that participating. But beyond that, it's random selection and it still is that way today. And again, that people took some time to get people comfortable with it. But I think that the fact that this is the voice of the people and reflects their output through an open and transparent process is so key to both why it is good, why it works, but also why it's trusted. So that's number one and I think will forever be the heart of the product. Another one that people thought was crazy was transparency.

**Keith Coleman** (01:13:49):
The previous approaches to dealing with misleading info, it felt to a lot of people, like black box tech companies or media companies or leads or whatever making decisions. We're like, "People need to get comfortable with this. They need to trust this. So the whole thing has to be out in the open." The code that decides what notes share has to be out in the open. All of the data and ratings that make it happen have to be out in the open. People should be able to take the code and data and replicate the whole service and that we have done exactly what we've said we've done. And they should be able to audit it. They should be able to go and look and say, "Hey, I think this part could be better."

**Keith Coleman** (01:14:28):
Or if they think we're biased, they should be able to work with the data and point it out. And if people have good observations, that should factor back into the code. And this is, again, something that's difficult to get people comfortable with, that everything is out there, you can't cover anything up. But I think that's so essential to people trusting it. Yeah, we set these out on day one. We go back to them constantly because we're always evolving the product, and we're always like got to make sure every new change is open. Whenever we update the scoring system, there's an update in GitHub when the data is published daily so you can download it. And so yeah, I think those have been really essential to the thing working.

**Jay Baxter** (01:15:13):
And by the way, these do not come without a cost. It's actually really hard from an end perspective to actually open source the actual algorithm that's running on the actual data. Because the way large-scale services like this are usually architected does not naturally lend itself to being run as a script by someone who's downloaded a TSV. So we actually have to take weird architectural decisions to make this possible in a way that probably wouldn't have been if we didn't start with this assumption from scratch. We would've had to maybe rewrite the system to make it like this.

**Lenny Rachitsky** (01:15:49):
What's an example of that?

**Jay Baxter** (01:15:50):
For instance, there's a matrix factorization that we train. Usually, you would train a matrix factor... train your ML model once and then serve it, I guess with a separate service. But we didn't want to have people externally spinning up services to be able to replicate the system that we had. So basically, I don't think it would've been actually very cool if we had open sourced the code in a way that wasn't actually runnable, I guess, by someone just... At this point, you can download Python code and run a script. You do need a lot of RAM right now, but you can do it on one machine.

**Lenny Rachitsky** (01:16:35):
Okay. How much RAM are we talking about?

**Jay Baxter** (01:16:37):
Oh, only like 500 gigs.

**Lenny Rachitsky** (01:16:41):
Okay. Okay. That's reasonable.

**Jay Baxter** (01:16:41):
It'll take a day if you don't do anything special to speed it up. Good to know, but yeah.

**Lenny Rachitsky** (01:16:45):
Cool.

**Jay Baxter** (01:16:46):
Possible is the key thing, and people have done it. Vitalik Buterin had a blog post where he talks about his explorations, making sure the algorithm really does what it says it does. And I think just the fact that a handful of people have done this, there's enough people who have done it that there's someone you'd probably trust who's verified it.

**Lenny Rachitsky** (01:17:11):
And that's rolling out to Meta. No big deal. I love just as you described these principles, just I could imagine a PM at a company being like, "Okay, guys. Here, I want to do this project." There's so much idealism to it that rarely works in real life; going to be open source. You're going to give it to everyone. We don't have actual control over what it's going to do, don't worry about it. It's going to just change the way people see this thing that we've been very careful about and then it works. And I think that's very rare and it's really impressive. And what I'm hearing partly is that sticking to those principles was actually really fundamental to it working and not bending over when someone's like, "No, no, no, we can't do this. What if we change this part?"

**Keith Coleman** (01:17:54):
I think if we had broken with any of those principles, if there was anything black box, if there was whatever, the product would be a lot harder to trust. And so I think it's because we've just stuck to them so cleanly simply that people can trust it.

**Lenny Rachitsky** (01:18:11):
You've talked about a few moments when it was like, wow, the White House changed their announcement because of the community note. We talked about the dog is a cat. Are there any other moments that after you launched of, "Holy shit, this is working? This is going to actually work."

**Keith Coleman** (01:18:26):
All along, we saw it working. We wanted to be confident whenever we expanded it to new audiences or new countries or whatever, we wanted to be confident it was going to work. So maybe held our breath a little bit just to see that it would do what we expected, but we always expected that. But that said, there were definitely stress cases. The one that comes to mind is the start of the Israel Hamas conflict in 2023 in October. That was probably the largest deluge of misleading information I've ever seen shared on the internet at one time. It was overwhelming. A number of photos and videos and whatever coming out related to that, it was insane. And just to give you an example, I think it was first three days or something of that conflict, we had 500 notes covering all sorts of different... out of context imagery.

**Keith Coleman** (01:19:32):
Someone would say, "Hey, this is happening here." It's actually from 2013 in Syria. There were people making fake battle footage in the video game simulator Arma 3. So there were notes explaining, this stuff looked realistic. And unless you saw the note, you wouldn't really know. There are all sorts of claims about what was going on in the ground, and that was definitely... The product was still pretty new at that point. We'd expanded in the U. S. less than a year before that. We had been rolling out throughout the world that year and then this large event happened. And I felt like we were just enough prepared at the right time for the system to be able to handle that.

**Keith Coleman** (01:20:16):
Probably one of the most important things we did right before that was launch the ability to write notes on images and videos and have those matched to other posts. I remember at that time thinking, "Wow, I'm glad we launched that feature a few months ago versus still had it on the shelf," because it was really important in that conflict. And I think even it was just a few weeks before we had launched a major speed up in notes too. When we first built the product, the number one focus was always quality. We knew that the product would live and die by the quality of the notes. That was the thing we could never give up on. We also knew it needed to deliver speed and scale, but we're like, "We will get the quality in the right place, and we can speed it up and scale-

**Keith Coleman** (01:21:00):
Get the quality in the right place and we can speed it up and scale it out over time. And we had actually just launched a speed-up that took three hours off the time it needed to go live, and it was I think a matter of weeks before that conflict happened, so again, super glad that was out there. In the first few days of the conflict the median time from a post going live to a note showing up was five hours, which is like crazy fast. Typical fact checking is like two to four, at least it's really common to see it take two to four days. These notes were showing up in five hours and we're like, we are so glad we got those things out before this happened, it made the service a lot more helpful.

**Jay Baxter** (01:21:40):
One other thing that was, I think, nice to see working then was, one criticism of Community Notes some people bring up is, well if you always need agreement from people who typically disagree, then in these super polarized settings, that conflict being probably number one, then you wouldn't see any notes. But actually the reality was there were tons of notes about that conflict. So I think there was this kind of nice property where actually, and maybe this is a surprising fact, that there's more agreement out there across polarized divides than maybe conventional wisdom says, and the places where people agreed were really objectively true and verifiable. I guess maybe this is more true the more polarized the setting is, but where the agreement actually lends you, and basically notes that are very neutrally written, very focused on the facts and easy to verify information.

**Lenny Rachitsky** (01:22:46):
There's this talk for a while of just there's no more facts, nobody believes there is a single true fact anymore, everything is subjective, and I think Community Notes proves the opposite. Facts matter, there are facts that we can all agree with even on the most controversial topics.

**Keith Coleman** (01:23:04):
Yeah, we saw this really from day one, when we would show those prototypes to people just depicting the idea, it was really obvious that people cared more about, or they cared a lot about understanding reality and what was going on and they were willing to disagree with their side, so to speak, to recognize that. And I think that's not always that obvious to people. The world does feel really polarized, but people definitely are willing to cross partisan boundaries to get to accurate information and that's why the product works.

**Lenny Rachitsky** (01:23:38):
It feels like as we rely more and more on what we know and understand about the world is becoming social media online and moving this quickly, it's like I'm so thankful this exists because otherwise it'd just be, what do we trust anymore? This being out aligns with we need this thing to exist at the same time. And it feels like at the same time there's also people I just don't trust. I think people have shifted from I trust what I read to, okay, I shouldn't just believe everything I'm reading. Is there anything there you're noticing about just how people think about news they see and their shift of just like, I'm not going to believe everything. Is there anything that you've noticed about just human behavior or just the way we've shifted understanding what is true?

**Keith Coleman** (01:24:30):
We haven't done any research to look broadly at how people's perceptions are changing there, but I certainly have found myself that particularly seeing notes, I am more skeptical about what I read at first, and I think that's been helpful. And we hear that from people, that they think about things a bit more, and I think that's a good secondary effect and benefit of something like this, which is the more you see the patterns of how what you're reading can be wrong, the more you can thoughtfully question it and try to get a better understanding of what's really going on. So historically I think this was called media literacy, but basic idea of can you understand the ways in which things can go wrong and try to cut them yourself.

**Jay Baxter** (01:25:21):
Another aspect I think we help with that is discovery of the Community Notes. I think often before Community Notes you could have just been living in a little news filter bubble, or maybe there were fact checks out there that you should have been reading but you weren't discovering them. So the fact that the note applies, it is directly attached to the post and visible by anyone who sees the post helps cross those filter bubbles and can kind of... I think for some people it's the first time they've actually seen counter arguments to claims made in their own little echo chamber.

**Lenny Rachitsky** (01:25:59):
That's incredible, yeah. I love the point you're making about how it actually teaches people to be a little more skeptical of the things they read. It's an education system more than just, here, this one thing is wrong. I love that.

**Lenny Rachitsky** (01:26:13):
Okay, just a few more questions. There was an audience question asked on Twitter, we all asked on Twitter, "What do people want to know about Community Notes?" one was actually why you guys switched to anonymous contributors, what was the decision behind that?

**Keith Coleman** (01:26:26):
Yeah, we had this pilot where we were testing with a small number of contributors, a few thousand contributors, and we learned a lot through that pilot. Probably the biggest thing we learned was related to anonymity or pseudonymity of contributors. We had originally assumed that it was important that people contribute under their real handle, or their real name, or whatever it was. The first prototypes depicted that, we kind of thought that would be important for people trusting the note, and actually it was totally wrong. The best option was actually opposite of what we first tried.

**Keith Coleman** (01:27:02):
We found a few things. One, people were hesitant to write a note on a controversial topic because they didn't want to get attacked or harassed online. And so some people were comfortable doing this but others were not, and so it meant there was more potential good notes to be written than were getting written, and this was very clear feedback from the pilot.

**Keith Coleman** (01:27:24):
Two, and this is super interesting, people are actually more willing to cross partisan boundaries when they are anonymous or pseudonymous than when they are under their real name, and it intuitively makes a lot of sense. If you publicly are using your name, you feel are affiliated with one side versus the other, you might hesitate to be perceived as breaking with that side. But you may actually, for example, find a note helpful that's critical of that side, and there's a bunch of studies that show when people are anonymous, they're much more willing to cross partisan boundaries and work with the other side, agree with the other side, and we saw that too. And so by allowing people to be pseudonymous, you actually get more honest answers about what they really think and it helps find disagreement that really-

**Lenny Rachitsky** (01:28:11):
That's so counterintuitive.

**Keith Coleman** (01:28:12):
Yes.

**Lenny Rachitsky** (01:28:12):
You never hear the opposite always, and it's so interesting it's the opposite.

**Keith Coleman** (01:28:15):
Yeah, yeah.

**Jay Baxter** (01:28:17):
I think the same principle applies to making the likes private.

**Lenny Rachitsky** (01:28:21):
I was just thinking that.

**Jay Baxter** (01:28:21):
Yeah.

**Lenny Rachitsky** (01:28:23):
Yeah, I like a lot more stuff that's a little, definitely, I wouldn't have liked, yeah.

**Keith Coleman** (01:28:28):
It allows freedom for honesty, which is pretty great. And one of the criticisms of pseudonymity is it can generate, maybe people have reached the quality threshold that they put out there, but we have so many quality mechanisms in the system that that wasn't an issue, so we could keep quality high while opening up for that honesty.

**Lenny Rachitsky** (01:28:48):
Another question, you touched on this a little bit, which is around navigating the existing trust and safety apparatus of Twitter, which as you described, basically, previously, it was like we make decisions on what is true and not, and every company works this way, you guys basically upended that like, here's a completely different way, you have no control over what we say is true or not. Talk about just that experience of overcoming that, I imagine, very difficult hurdle of like, okay, forget all that, we're going to do it totally different.

**Keith Coleman** (01:29:20):
Yeah, it was definitely, what we were proposing was very different. I will say that I think people were sort of open-minded to it, generally speaking, and I think everyone had a sense that what was being done at the time wasn't really working that well or solving the problem, and people were open to new ideas, so that's a good foundation.

**Keith Coleman** (01:29:39):
But I think one thing we did that was probably very helpful in that is we wanted the product to prove itself at any point. First it had to prove that people could possibly find notes helpful, then it had to prove that people could possibly write these notes that would be good quality. And so anytime that we were proposing doing something with the product, like running some research test, or running the pilot, or expanding the pilot, we always had the data that had convinced us that that was a good decision, like we were stepping into the next phase of expansion that made sense. And so I think we probably rarely proposed anything that seemed unwise, because we were holding such a high bar for quality ourselves, and I suspect that went a long way.

**Lenny Rachitsky** (01:30:33):
So it's partly, what I'm hearing is, take it step by step to prove this is actually working, and partly be confident it is working to yourself before you try to convince the trust and safety team this is the way to go.

**Keith Coleman** (01:30:46):
Exactly.

**Lenny Rachitsky** (01:30:49):
Was there a moment along that journey it shifted from no way this is a thing to okay, wow, let's actually consider this? Or was it this very gradual process?

**Keith Coleman** (01:30:59):
Whether other people were saying no way to wow, let's actually-

**Lenny Rachitsky** (01:31:03):
Yeah, just internally of just like, okay, we're going to actually stop this trust and safety way of operating and instead rely on Community Notes, was there a moment of like, okay, let's actually make that switch, or was that Elon actually, is that the big switch?

**Keith Coleman** (01:31:15):
The biggest change there happened in X, the biggest changes prior to that were just the decision to put this out there and have it be operating in public at first US wide scale. But yeah, then the bigger switches came in the X period.

**Jay Baxter** (01:31:36):
I think even though there was original research before Birdwatch had even started, or Community Notes had even started, from external researchers showing that crowdsourced fact-checkers, laypeople can do about as well as fact-checkers and actually the agreement rates were kind of similar between the groups. I think even though that research was out there, I think there were definitely a lot of people who didn't really believe it could work until it already worked.

**Lenny Rachitsky** (01:32:04):
Basically prove it, prove that it works. Yeah, that makes sense, versus just a bunch of docs and strategy and thinking, it's just like, look, it's actually working, you can see for yourself.

**Jay Baxter** (01:32:13):
Yeah.

**Lenny Rachitsky** (01:32:14):
Makes sense. Okay, possibly last question, we'll see which fractals of questions you guys bring up here. I referenced this a couple times, this incredible achievement of keeping a project alive through Jack and then, I have this note, and Kayvon running the show then, and then Parag running Twitter, and then Elon, and then Linda taking over as CEO, quite rare, especially something this visible, this impactful to everything that X is. Any lessons or keys to that actually working, of this project surviving throughout so many work changes and leaders?

**Keith Coleman** (01:32:54):
It definitely has been a crazy time to be building something. It's been fun. The craziness has been entertaining. I think one reason perhaps the product has done so well and survived is the nature of the product itself. It is designed to produce information that is found helpful by people who normally disagree. And so even if you have CEOs or leaders who might disagree, there's a good chance actually they'll find it helpful, they'll be like, wow, this thing does produce pretty useful output. So I think there's something in the nature of the product itself, that when people see it, whatever side they're on, left, up, down, they're likely to find it pretty helpful, so I do think that helps.

**Keith Coleman** (01:33:39):
I also think the team executed really well. We had ambitious goals that were exciting, they solved a real problem. This is a real problem that matters in the world. At every step, as we talked about, the product needed to prove itself, and we would make sure it proved itself and we would bring the results that convinced us and we'd share those with people. And so they would say, oh yeah, I agree, it kind of proved itself, let's take the next leap. And we've done that all along the way and we continue to operate that way, and I think that focus on the outcome and goal that matters, and executing against it, really helped.

**Keith Coleman** (01:34:24):
The team did not get distracted by much all through the period during which the acquisition happened. There was a lot of opportunity for distraction. This team was shipping every week, we were super focused on the goal, let's make this thing work, let's get these notes out there, and I think people saw that execution and were excited to support it.

**Lenny Rachitsky** (01:34:49):
Yeah, like it's working, why would we mess with that? And it's important, and it keeps us from having to hire tens of thousands of people to fact check.

**Keith Coleman** (01:34:57):
The interesting thing about that is no one ever asked us or brought up or seemed to care about anything related to cost savings in this process. And I think that's an assumption people have outside the company, that this must have been a reason there was interest in it. But that was never a goal, it was not at all why the project was started, it was not why people were excited about the project. And I think that's also, for people outside who maybe don't see the conversations, it's kind of a heartening thing to know, is that the focus was always on solving the problem. The other approach is even if you had 10,000 people doing it, the real issue is that they don't work that well because they're not trusted or they don't scale or they're too slow. And so the goal was really always just help people stay informed at scale. Let's build an internet scale solution to an internet scale problem that people like.

**Lenny Rachitsky** (01:35:51):
Something I heard about you, Keith, when I was asking people about how this worked and why this worked so well is that they describe you with having a very low ego, and that allowed you to give up this whole team and power and influence and just the name, forget it, whatever you want, we'll call it Community Notes, great. Is there anything in there you can share of just how you think about that and how important that is as a product leader to have a low ego?

**Keith Coleman** (01:36:19):
For me, this project, I feel like I get to do community service with this project. I see my work as in service of the people and the community, and that's what motivates me. The only thing that I care about is delivering the outcome that the world finds helpful. And so in some ways the project has not been about ego, it's about truth-seeking, let's find... Not truth in the sense of what information is true, but let's find out what's actually going to make this work. How does it need to be structured, what should it be called? Whatever is going to produce the best outcome is what we should do. So I think I feel more attached to the product being helpful than to anything else, and so to whatever degree it might seem like low ego is probably more a result of wanting to actually solve the problem.

**Lenny Rachitsky** (01:37:15):
And I think partly what I'm hearing is just if you win and succeed, good things will happen, so focus on that.

**Keith Coleman** (01:37:20):
Certainly satisfying things will happen, it's very satisfying to have people appreciate it. It's satisfying that people on the left and right love it. It's satisfying that even people who receive notes, love notes, and reach out to them and post them, that's amazing, it feels so good to have helped give people that, and yeah, it's very motivating. It's a great reason to wake up in the morning.

**Lenny Rachitsky** (01:37:44):
It's absurd this has worked, but it's also like of course this would work, of course something like this should work. It's like such interesting-

**Keith Coleman** (01:37:50):
It's the internet, it's of the internet, that's why it works.

**Lenny Rachitsky** (01:37:55):
Oh man. Where's Community Notes going from here? What's happening, where's it going, what's the future?

**Keith Coleman** (01:38:03):
We're always working on basically more better notes faster. So there's clearly an opportunity to get more notes out there, we want them to stay as good or better than they are, we want to get them there faster, so we're always working on core product changes to help deliver that. Recently, for example, we just released an update to what we call the Community Notes bat signal, or the ability to request a Community Note. So anyone on X can say, "Hey, I think this post needs a Community Note," and now they can even add a source explaining why so that when a prospective writer sees that it's much easier for them to write a note. So we're always working on core things like that, core algorithm improvements.

**Keith Coleman** (01:38:47):
I think there are also new frontiers that show a lot of potential, AI and LLMs are one. It's easy to imagine a lot of ways that AI could assist the people in this task they're doing of trying to get information out there quickly. And maybe Jay should talk about the Supernotes work that we've done with some folks outside the company.

**Jay Baxter** (01:39:13):
Yeah, so one cool thing about having public data and code is that external researchers can collaborate with you, and in this case Supernotes had this idea that we can basically take existing notes as input, existing proposed notes that maybe they have some problem, maybe they have part of the story, maybe they're worded in kind of a biased way. Basically take all these in, have an LLM generate a ton of different variants, and then basically make the simulated jury to basically get a representative group of contributors for community notes who would be rating the note and try to predict based on their past ratings how they would rate these LLM generated notes. And so this way you can actually, rather than just having an LLM write a note from scratch and hoping it's good, you can simulate the entire community notes rating process and explicitly create notes that are likely to be rated helpful by people.

**Jay Baxter** (01:40:18):
So I think ideas like that are very promising for the future, and it's a nice way that LLMs and humans can work together. Obviously agents can browse the web too, and that's one way that you could imagine agents assisting humans is maybe checking whether a note is actually supported by the source. Although then you get into things like, well, are people going to actually be as diligent? Right now I think raters are very diligent because they know just some Community Notes contributor wrote this like, I better check this before I rate it helpful. But hopefully we can design things in a way such that people don't trust the output and actually verify it themselves before issuing a helpful rating.

**Lenny Rachitsky** (01:41:11):
Yeah, that is such an interesting area to explore where you want to avoid AI hallucinating slop versus make it easier and scale it even further. What an interesting challenge.

**Keith Coleman** (01:41:23):
What's cool about this project, in addition to the AI element, is that it's being done outside the company. We talked earlier about the open source transparency. The key reason we made this all open source was so people could see how it worked, but the dream is actually that, it's not just that the contributions to the notes and ratings are from the people, but the dream is actually the product is built by the people. What if the scoring algorithm were significantly or entirely written by the public? That would be incredible. And Supernotes is probably the first very substantial potential change in the algorithm of the way it works, that was kind of coming from the outside and plausibly could be part of the core, so we'd love to see the product go in that direction as well.

**Lenny Rachitsky** (01:42:08):
Sweet, go Supernotes. Well guys, the work you're doing is tremendous. This is every product person's dream, I think to work, on something like this. Small team, lots of support, lots of impact, just innately interesting, and so I think this is going to inspire a lot of people.

**Lenny Rachitsky** (01:42:27):
So let me just ask you, is there anything else you wanted to share? Anything else you think might be helpful for folks to leave them with?

**Jay Baxter** (01:42:33):
Sure, I guess one thing that just I thought was interesting over the course of working on this product is just there's... I think in a similar way to how retweets originally were not something Jack came up with, I think users just started doing it and then it became a core part of the product. There's a huge way already in which there's just a lot of surprising things that people wanted to use Community Notes for that I don't think we really expected, and it's kind of cool to see those user desires emerge.

**Jay Baxter** (01:43:04):
I think one example, I guess we had always been imagining political type of misinformation, but for whatever reason there's a lot of people who love debating whether Messi or Ronaldo got more goals. I guess it's kind of a funny one. There's a community moderation aspect, so I think we also thought that this would be specifically for adding context to misleading or potentially misleading information, but what you can see is that there are some notes that go beyond that towards calling out content that they think is spammy or something. So I think that's just another dimension in which commuted notes is a product that's driven by the people.

**Lenny Rachitsky** (01:43:57):
That's so beautiful, basically they're trying to keep Twitter/X healthy and they're just like, no, this should be taken down, this tweet of spam.

**Jay Baxter** (01:44:05):
Yeah.

**Lenny Rachitsky** (01:44:06):
I love that. Is there an answer on the Messi versus, who is the other soccer player?

**Jay Baxter** (01:44:10):
Ronaldo.

**Lenny Rachitsky** (01:44:11):
Ronaldo, okay. Is there a definitive fact there or is that just unknowable?

**Jay Baxter** (01:44:17):
Yeah, I guess that's an interesting one because it's a case where raters are actually very polarized. I guess it actually kind of fits into the core algorithm where there's some people who are just diehard Messi fans or Ronaldo fans, just like they could be on politics, so we actually specifically modeled that topic, as well as some other topics, so we can estimate people's opinion on that particular debate. It's kind of funny that something like that would emerge.

**Lenny Rachitsky** (01:44:45):
You're saying that's the most controversial topic on X, Ronaldo versus Messi.

**Jay Baxter** (01:44:51):
That's a controversial one.

**Lenny Rachitsky** (01:44:52):
Oh wow, who knew? Okay. Keith, is there anything you wanted to add?

**Keith Coleman** (01:44:58):
Yeah, community Notes is cool itself, but I think what it points to about society is actually even bigger. Society often feels really polarized, you hear people talk about it all the time, no one can ever agree on anything, but actually Community Note shows you people really can agree on quite a lot. Even on super controversial topics related to politics and everything, there's a lot of agreement, that's why notes work.

**Keith Coleman** (01:45:23):
And I think that's a really big reason for optimism about the world, is that while it might feel polarized, there's probably like an 80% set of people that agree on quite a lot of things. And imagine if we could use the same kind of approaches we use with notes, but to find agreement on legislation, or policies, or things like that that people want the government or the world to do, possibly we could get a lot more momentum behind these ideas that the people really want and everyone would be a lot happier. Maybe 10% of the people on the edges wouldn't be happy, but I bet there's a lot of agreement that we are not identifying, and if we did it, we'd all be pretty happy. So I don't know, I think it's easy for people to feel pessimistic about the world, but I think this product is a good reason to be optimistic about the future.

**Lenny Rachitsky** (01:46:12):
What an incredible way to end it. I can also see, Keith, why people want to join you and work with you and work on this team.

**Keith Coleman** (01:46:19):
Appreciate it. If you do want to join, we are hiring an ML engineer. You get to work on these amazing problems with us and have a lot of fun, so we're accepting applications at x.com/communitynotes.

**Lenny Rachitsky** (01:46:32):
Okay, great, I'm glad you gave the URL. Oh man, you're about to get flooded.

**Lenny Rachitsky** (01:46:36):
Guys, thank you so much for doing this. Is there anywhere other than that place to go off, join the team as an ML engineer, is there any other place you want to point people to, either your socials or anything else?

**Keith Coleman** (01:46:47):
I'm KeithColeman on X, please reach out if you have any feedback or want to help us out, whether you may want to work here or want to do something from the outside, we would love to talk.

**Jay Baxter** (01:46:58):
Yeah, I'm @ _JayBaxter_ at X. Yeah, I think in particular, besides just using Community Notes, it would be great to get more substantial contributions, pull requests, collaborate on projects like Supernotes, I think that's the most exciting type of stuff if people do want to contribute.

**Lenny Rachitsky** (01:47:22):
Ship some code guys. Amazing. Guys, thank you so much for doing this.

**Keith Coleman** (01:47:27):
Thanks for having us, Lenny.

**Jay Baxter** (01:47:29):
Thank you, thank so much.

**Lenny Rachitsky** (01:47:33):
It's my pleasure. Bye everyone. Thank you so much for listening. If you found this valuable, you can subscribe to the show on Apple Podcasts, Spotify, or your favorite podcast app. Also, please consider giving us a rating or leaving a review as that really helps other listeners find the podcast. You can find all past episodes or learn more about the show at LennysPodcast.com. See you in the next episode.

---

## [9/15] The creator of WordPress opens up about becoming an internet villain, why he’s taking a stand, and the future of open source | Matt Mullenweg (founder and CEO, Automattic)

**Lenny Rachitsky** (00:00:00):
If you're really open and open source, sometimes you have to stand up the bullies and you have to fight to protect your open source ideals.

**Speaker 1** (00:00:05):
Please put your hands together for Matt Mullenweg.

**Lenny Rachitsky** (00:00:08):
Matt Mullenweg has been making some questionable moves recently. There's a lot going on with Matt and WordPress these days. 20+ years of good sentiment burned in days. You are like a 100% beloved hero of open source and internet and now you're in this, a lot of people don't like you.

**Matt Mullenweg** (00:00:23):
If you were kind of inside baseball with WordPress, it's actually a lot of people who have been unhappy with me over the years. Previously, 1% of the world thought I was terrible and now I feel like it's up to four or 5%.

**Lenny Rachitsky** (00:00:35):
People that don't know what the hell's going on, what's just like the high level overview of what's going on?

**Matt Mullenweg** (00:00:39):
There's a company called WP Engine. By 2018, they got bought out by a private equity firm called Silver Lake. Since 2019, WP Engine has kind of changed a bit. They started using the trademark, they're offering something called WordPress. They're referred to it as like a bastardized hacked up version of it. It's diluting our brand.

**Lenny Rachitsky** (00:00:56):
Why do you think so many people are looking at you as the bad guy?

**Matt Mullenweg** (00:00:58):
A lie gets around the world seven times before the truth has time to get out of bed.

**Lenny Rachitsky** (00:01:07):
Today my guest is Matt Mullenweg. Matt is the co-creator of WordPress, which powers 40% of websites on the internet today, including whitehouse.gov. He's also the CEO of Automatic, which is valued at over $7 billion and owns products like WordPress.com, Tumblr, WooCommerce, Gravatar, and Pocket Casts. There is a lot of drama these days around Matt and WordPress and within the open source community, so I thought I'd have Matt on to address many of the criticisms head-on that he hasn't addressed in other places and also just get the full story on what's going on. We also chat about what incepted him to spend over half his life at this point on open source and creating WordPress. Also, why products like Llama are what he calls 'fake open source' and his perspective on AI and open source. Also, how AI is actually trained on open source code and what that means for the future and his approach for deciding what companies to acquire within Automatic.

**Christina Cacioppo** (00:03:51):
Great to be here, big fan of the podcast and the newsletter.

**Lenny Rachitsky** (00:03:54):
Vanta is a longtime sponsor of the show, but for some of our newer listeners, what does Vanta do and who is it for?

**Christina Cacioppo** (00:04:01):
Sure. So we started Vanta in 2018 focused on founders helping them start to build out their security programs and get credit for all of that hard security work with compliance certifications like SOC 2 or ISO 27001. Today we currently help over 9,000 companies including some startup household names like Atlassian, Ramp and LangChain start and scale their security programs and ultimately build trust by automating compliance, centralizing GRC, and accelerating security reviews.

**Lenny Rachitsky** (00:04:31):
That is awesome. I know from experience that these things take a lot of time and a lot of resources and nobody wants to spend time doing this.

**Christina Cacioppo** (00:04:39):
That is very much our experience before the company and to some extent during it. But the idea is with automation, with AI, with software, we are helping customers build trust with prospects and customers in an efficient way. And our joke, we started this compliance company so you don't have to.

**Lenny Rachitsky** (00:04:55):
We appreciate you for doing that and you have a special discount for listeners, they can get $1000 off Vanta at vanta.com/lenny, that's V-A-N-T-A .com/lenny for $1,000 off Vanta. Thanks for that, Christina.

**Christina Cacioppo** (00:05:10):
Thank you.

**Lenny Rachitsky** (00:05:14):
Matt, thank you so much for being here. Welcome to the podcast.

**Matt Mullenweg** (00:05:17):
Thanks, big fan and long time listener, so happy to be on.

**Lenny Rachitsky** (00:05:21):
I'm a long time fan. I've been wanting to get you on this podcast for so long and this is such an interesting time to be chatting with you, there's a lot going on with Matt and WordPress these days, so it's really interesting, almost good that we waited a little bit to talk so we're going to get into a lot of that stuff. But I want to start with just what is it that you do, Matt? What are all the things you're involved in? Give people a sense of just the things you're working on.

**Matt Mullenweg** (00:05:44):
So first, when I was 19, I co-founded an open source project called WordPress with Mike Little and we started just blogging software, then became sort of a full site thing and then became a platform that really tons of stuff is built on and now it's kind of transitioning into this cool WASM can be embedded anywhere or run locally or make mobile apps. It's really interesting seeing WordPress used as an engine for powering things I would say don't even look like a website, which is kind of wild to me, but that's kind of the beauty of the open source people do things with that you don't expect. End up dropping out of college, moving to San Francisco and then worked at CNET for a year as project manager actually, that's how they hired me.

**Lenny Rachitsky** (00:06:27):
I want to talk about that, but go on.

**Matt Mullenweg** (00:06:30):
And then had this vision where instead of downloading the software and setting up a database and everything, we made a SaaS version of WordPress. I pitched it at CNET, they didn't want to do it, so I was like, "Okay, I got to do this," so I left and started a company called Automatic. And the idea was to essentially compliment the core WordPress software with some commercial services, things that run in the cloud, like Akismet which is our machine learning AI, I guess you'd call it AI now, but anti-spam system, or Jetpack, which is iCloud for WordPress. It does the backups and the real time sync and everything like that.

**Matt Mullenweg** (00:07:04):
So that was 19 years ago, so that's now grown to be over 1700 people in actually 90 countries so we've actually been fully distributed and remote and asynchronous from the start, which I think is one of our superpowers. I actually wasn't the CEO in the beginning, but in 2014, so I guess 11 years ago, I became CEO. The original CEO was... Well, I guess I wasn't the very beginning, but then I hired Tony Schneider from BCL, probably four or five months in.

**Matt Mullenweg** (00:07:35):
And yeah, so that is a very full-time thing and Automatic does a lot of products, WooCommerce, which is open source Shopify, which is now half our revenue. And then we have some really cool apps so like Beeper, DayOne, Simplenote, Rocket Casts or trying to fill up your home screens with open web, open source things that are very privacy and user-centric. So running that company is definitely a full-time job. I still run WordPress.org and the WordPress project, so I'm the lead developer there and so sort of manage all those releases in the community and the directories and all the sort of things we do on WordPress.org and this cool thing called Openverse we took over Creative Commons, which is a way you can find sort open licensed images and audio and video. So basically if you notice a throughput through all these things, it's open source.

**Matt Mullenweg** (00:08:31):
On the nights and weekends or side, a few hours a week I do some angel investing. So I've done over 100 angel investments through an entity called Audrey Capital, which is sort of, if anything's in the sort of WordPress space, I invest in it through Automatic, but if anything's a little more further afield I do it through Audrey Capital and have done some really exciting investments there, everything from name brands like Stripe and SpaceX, but also it was in the seat of Calm or a lot of home automation stuff like Ring, August, smart things. Yeah, just check out Audrey Capital, it's got some fun stuff in there. Daylight Computer, which is one I'm very excited about right now.

**Matt Mullenweg** (00:09:11):
And I guess finally I love San Francisco, so I an a co-owner of a cool grungy jazz club in North Beach called Keys with Simon Rowe. So if Wednesday through Saturday night, you want to see some awesome live jazz, check out Keys.

**Lenny Rachitsky** (00:09:28):
Wow, okay. You said too much, I get it now. Jazz club I was not aware of, I got to check this out. It's called Keys?

**Matt Mullenweg** (00:09:36):
Yeah, keys Jazz Bistro.

**Lenny Rachitsky** (00:09:38):
Okay, cool.

**Matt Mullenweg** (00:09:39):
It's over on Broadway in Columbus, kind of right around there.

**Lenny Rachitsky** (00:09:42):
Amazing. That was news to me. Going back to Automatic, I think people don't get the scale of this thing, so just to mirror back, if you think that even add to what you've said, 1700 people work there, 90 different countries. Also, you didn't share this stat, it was something like 43% of internet websites are built on WordPress, run on WordPress.

**Matt Mullenweg** (00:10:02):
Yeah, so when we started, a lot of websites were built on custom CMSs and there's a lot of fragmentation in the space, but now WordPress has grown to be over 40% of all websites in the world, which is 10x the number two, which right now is Shopify.

**Lenny Rachitsky** (00:10:17):
Right, they're like at 4%. I was looking at that list.

**Matt Mullenweg** (00:10:20):
They're around 4%, yeah.

**Lenny Rachitsky** (00:10:21):
That's unreal.

**Matt Mullenweg** (00:10:22):
It used to be open source was the top three. Unfortunately, Jula and Drupal have fallen behind, and so now it's like Shopify, Wix, Squarespace are the top ones but WordPress is still, because we have this flywheel of open source community, its movements, any open source like Linux or Apache or Wikipedia, it has some positive flywheel effects when it takes off.

**Lenny Rachitsky** (00:10:48):
Awesome, okay. And then there's a few other things you didn't mention, I want to get to this later, but I'll just mention now, you guys own Tumblr, you bought Tumblr, which I don't think a lot of people necessarily know.

**Matt Mullenweg** (00:10:56):
I'm sorry, I forgot to mention that, Tumblr.

**Lenny Rachitsky** (00:10:58):
We're going to get into that, yeah.

**Matt Mullenweg** (00:11:00):
Yeah, running a social network is definitely the hardest thing I've ever attempted. I thought we knew what we were doing because WordPress ran so much of the web, we dealt with, I thought, every content moderation thing you could ever deal with, but social networks are a whole other ballgame.

**Lenny Rachitsky** (00:11:15):
Okay, a couple more fun facts before we get into some other stuff I want to chat about. Fun fact number one is you were super involved in the Bay Lights project. I didn't know this. For people that don't know what the Bay Lights, if you're in San Francisco you definitely know what the Bay Lights project is and I'm sure you love it, for people that don't know what this is about, what is this project and how have you been involved? Why have you been instrumental in to make this a thing?

**Matt Mullenweg** (00:11:37):
Bay Lights, there's two famous bridges in San Francisco, the Golden Gate Bridge, which is kind of the iconic one, there's actually the Bay Bridge, which is the workhorse of San Francisco. It's one of the busiest bridges in the country. And it's really beautiful from an engineering point of view. And so kind of a vision between Ben Davis and artist Leo Villarreal, who's an amazing light artist, actually who started Burning Man was to put, gosh, I forget the number, I think 18,000 LEDs on the side of the bridge on all the cables and create this really beautiful, gentle kind of algorithmic light piece, light art piece.

**Matt Mullenweg** (00:12:18):
And yeah, Ben Davis was dating an artist friend of mine and we were over and having drinks on my patio and we were looking at the Bay Bridge and I had this kind of thing where there's some lights at the top of the Bay Bridge and I was like, "Oh, wouldn't it be cool if those lights were Christmas lights and they could do patterns or something?" It's the lights to keep planes from hitting it. And I was like, "Oh, you could program that." He was like, "Yeah." It was almost like the Social Network thing where a million's cool, but a billion would be really cool. He was like, "Yeah, that would be cool, but what if we put the whole side of it?" And so I was like, "Oh, cool," and sort of made an angel investment in that thing. They hadn't raised anything or had, I don't even think an entity at that point, but I was like to get you started, I forget what it was, 100 or 150k so I gave them that first bit and then it kind of blossomed into a thing.

**Matt Mullenweg** (00:13:07):
And then sort of fast-forward, I don't remember the exact timeline, but they were kind of at a final bit of fundraise and they weren't able to close that last bit and I actually mortgaged my condos and donated the last million, million and a half to finish out that project. The Bay Lights were online for 10 years. The technology degraded, and so the environment's very harsh. So actually we just completed a fundraise and are reinstalling the Bay Lights. They're calling it Bay Lights 360. So now it'll be both sides of the bridge, it'll be visible from also Oakland and the Treasure Island because the first version, the city was very worried about the drivers seeing the lights and it might distract them so we had to angle them so that you could only see it from San Francisco, which was a compromise we didn't love because we love the East Bay and everything else like that too. So new version is coming online hopefully later this year in the fall. And also that turned into a nonprofit called Illuminate, which I'm on the board of run by Ben Davis, who I mentioned previously, that does cool public art stuff around the city. So they're responsible for the Grace Lights, all the JFK Boulevard stuff where that has some murals and the beer garden and all the chairs, that's all Illuminate. So their thing is radical public art. So it's like art that needs to be free and accessible. And I think that's so important for San Francisco. We have great institutions know the SF, MoMA, the Opera, etc, that have huge budgets like 100 million a year and Illuminate literally one 10th of that has created something that millions of people can enjoy.

**Matt Mullenweg** (00:14:40):
And I like to think that anyone along the Embarcadero, you might be going through a tough time, obviously we have people who are struggling with mental health and homelessness and everything like that, but maybe seeing a little bit of art can help raise your soul a little bit. And that's how I think about philanthropy as well. You need to work on the base issues, the fundamentals at the bottom of Maslow's Hierarchy of Needs, and then you also have to work on the things that raise your soul a little bit, so arts. So I like that barbell approach to philanthropy.

**Lenny Rachitsky** (00:15:13):
Elon has a great quote along those lines, "You can't just work on solving problems all day, you need something inspiring to think about and to work towards." First of all, thank you for doing this. If you live in SF, you're like, this makes the city better, just having this around. I didn't realize you were involved in helping come up with the idea itself. I know that you did mortgaged your house to make it possible.

**Matt Mullenweg** (00:15:32):
I can't take any credit for the idea, I was exposed to it. I had an adjacent idea and they had a way cooler one with a real artist and everything like that. So I was just happy to be there. It's like being an angel investor, you can support the entrepreneurs and the people who truly do it.

**Lenny Rachitsky** (00:15:49):
Yeah, okay. And the other funny thing you said is about they were worried about the angle of the lights distracting people. What's funny is when I drive across the bridge, you can only see it when you're driving towards San Francisco looking backwards. So I'm looking in my rearview mirror or in the mirror turning around to the kind and it feels more dangerous, the lights shining in my face.

**Matt Mullenweg** (00:16:07):
They call it impossible works of art. There were like 13 agencies that had to sign off, they were worried the lights would distract birds or seals or environmental reviews, and it was really a lot of public bureaucrats and to make that happen, there was 20 places where someone could have said no and it never would happen. So it's very inspiring to see the city come together.

**Matt Mullenweg** (00:16:31):
Also in San Francisco, I feel like is entering new chapter right now, going from the doom loop to the boom loop. I'm a big believer in the city. So much innovation has come here from food, the burrito, fortune cookies, all these sorts of things are from San Francisco to obviously all the tech innovation that we're all familiar with. It's kind of the city of the future and I don't know what it is in the water from the '60s until now, cultural innovations, things that happen and influenced the whole world, Burning Man, Grateful Dead, et cetera. That all starts in San Francisco. So it's exciting to be here.

**Lenny Rachitsky** (00:17:03):
Let's set the so back, as they say on Twitter. Okay, someone very close to you told me that you're an excellent rapper. I'm not going to ask you to rap, but if you ever want to answer any questions in rap form, feel free.

**Matt Mullenweg** (00:17:17):
Oh man, that would be fun. I've dreamed about being able to do a Q&A and rhyme, but I don't think I'm that talented.

**Lenny Rachitsky** (00:17:26):
Planting the seed, I'm planting the seed. So I want to get into all the drama there in this world and right now, but I want to first lay the foundation of how you got into this and where this all came from. So let's talk about just the origin story of you and open source. More than half of your life, you've been working on open source, you've been working on WordPress, specifically WordPress is such a core community within the open source. What was kind of the origin story of you becoming obsessed and, I don't know, open source-pilled?

**Matt Mullenweg** (00:17:56):
I was a broke kid in Houston, Texas, and my passions were jazz. Houston has actually amazing music programs in the public schools, and so I was very fortunate to go to some of the best civil arts programs, including my high school called the High School for Performing and Visual Arts, where Beyoncé went, Robert Glasper, amazing folks went there. And so music was a big part of my life, and actually economics. So I had this fun teacher, Scott Roman, who participated in our school in the Federal Reserve Challenge, which was run by the Federal Reserve that sets the interest rates and backs the national banking system and everything like that, has this competition for high school students. It ended up being the first academic competition this art school ever won. And yeah, first year we kind of didn't get that far, second year we went all the way to nationals so I got to meet Alan Greenspan, Ben Bernanke was our judge, went to DC. So that was very, very exciting.

**Matt Mullenweg** (00:19:01):
And so being exposed to us having a great liberal arts education, the ideas of Frederick Hayek, Agnes Smith, Alexander Hamilton, Ben Franklin, Thucydides, all these sorts of things, that philosophy really influenced me. And combine that with that, music lessons were expensive, so we couldn't really afford them, so I would barter and trade. I'd build websites for local musicians and exchange for lessons. So these websites, I would start to put software on forums or different things and that kind of exposed me to open source. So my father was also an engineer, he worked for oil companies and things, but his world was all Microsoft, it was all proprietary. And I always kind of grew up in early days of the internet, so was Slashdots and Jeffrey Zeldman talked about web standards and all these things are really kind of the social milieu and zeitgeist that I grew up in.

**Matt Mullenweg** (00:19:56):
So combining all this philosophy I studied, it felt that open source was actually the most important idea of our generation. If the founding fathers were around today, I think they would be open source advocates. If you think about it as more and more of our lives are influenced and actually controlled by the software we use, if we don't have fundamental freedoms attached to that software, we're not truly free.

**Matt Mullenweg** (00:20:24):
So WordPress is under license called the GPL, which has four freedoms, the freedom to use the software for any purpose, so we can use it for anything, whether I agree with you or not, the terms of service is you could do whatever you want with it. The freedom to see how the software works, open up the hood, see how it works, see every line of code, you can audit it. The freedom to change it is the third freedom. And then finally the freedom to redistribute those changes so you can share them. And the GPL has a fun little hack where if you share them, you have to provide those same freedoms to who you share it with. So it's called a viral open source license as opposed to the MIT license or some of the others that aren't.

**Matt Mullenweg** (00:21:03):
So yeah, just kind of decided that this was what I was going to devote my life to. And so that became getting involved with some early open source projects, WordPress was actually a fork of abandoned open source project called B2. So the code base actually started was something that was already out there that I was a user and contributor to kind of volunteer on the forums and contribute code. And then when it was abandoned, myself and Mike were one of four or five different forks that started that, picked it up and tried to continue it for our own use and then later for our larger community.

**Lenny Rachitsky** (00:21:36):
It feels like a lot of people are coming around to exactly your worldview in, say, I was just watching a video of Jack Dorsey talking about how we're just controlled by algorithms and we don't know how they work and we are not in control of our lives. Have you seen that video?

**Matt Mullenweg** (00:21:52):
No, but I actually love that. Also, some people who maybe made their first billion or whatever from proprietary software then come back and it's so cool to see folks like Marc Andreessen or Bill Gurley be huge advocates for open source. I actually remember one of my early meetings with Andreessen Horowitz, Marc Andreessen. I didn't realize that at the time, Tony Schneider and I were sort fundraising and Marc really grilled us, he's like, "How can you build a business on open source? How can you be remote and distributed? Look around Silicon Valley, Google, Microsoft, Oracle, Sun, every great company has had an office. How are you going to build something that can change the internet with people all around the world?" And just had this long hour long debate and we walked out of that, I was like, "Wow, that was the worst meeting ever. They just hate everything we're doing."

**Matt Mullenweg** (00:22:48):
And then the next day they were like, "Hey, we're interested." I was like, "What happened?" I didn't realize that he had this idea where he wanted to attack the ideas and see how we defended it was how they battle tested things. I guess kind of like a Microsoft culture or whatever where you really grill the idea, I just wasn't familiar with that. But it's so cool now that some of these folks that I've learned so much from are such good advocates for open source.

**Lenny Rachitsky** (00:23:15):
Yeah, it's so interesting. I just had the Community Notes team on the podcast, and that's an amazing example of open source, Meta is adopting it from Twitter/X. Speaking of open source, one of the interesting, maybe most common ways people hear about open source these days is AI and AI models, and there's a couple areas here, one is you wrote this really interesting post where you talk about how Meta talks about Llama as an open source project, but you called it a false prophet. What is it about Llama that isn't open source? What are people missing when they see Llama and they're like, "Oh, Meta is amazing open sourcing everything."

**Matt Mullenweg** (00:23:52):
Llama, you can obviously download and run locally and all these sorts of things, right? You don't have to use their SaaS service. However, there's a clause in it that says if you're above a certain threshold of monthly active users...

**Matt Mullenweg** (00:24:00):
... as if you're above a certain threshold of monthly active users. I forget what it is. It's big. It's like 750 million, so it's pretty high. You need a license from them. And so that does not give you the freedom to use the software for any purpose. If at some point you have to ask for permission, you're kind of at the whims of this company who you might be aligned with or you might be an enemy with. And also, how do you define that? So, for example, on WordPress, our products don't have 750 million monthly active registered users, but we reach billions of people per month in terms of visitors. So, is that defined? So, there's just ambiguity there. So, I still think what they've done is amazing and I like that they're releasing it. I was very confused for why they insist on calling it open source because they...

**Matt Mullenweg** (00:24:54):
Actually, Meta has been a huge open source contributor. React. They've had incredible improvements to the PHP engine, which we benefit from a lot. So, they're actually a big open source contributor. I think Mark Zuckerberg really understands and loves open source too. My best guess now, I don't have any inside information here, but I think they're calling it open source because there's some European regulation about open source versus proprietary AI models. So, I think it might be a weird regulatory thing because clearly they understand this isn't open source. When I wrote the blog post, I was just kind of confused, and thought, "Oh, maybe if I get this message out there, they'll change." And then when they didn't, I was like, "Oh, there must be something else going on. I think it might be this regulatory thing." We were actually a big part of, actually many, many years ago, I think it was React that they were doing something with a licensing or a patent restriction on, and the WordPress community actually got Meta to change that and reverse something they were doing to lock it down.

**Matt Mullenweg** (00:25:53):
So, I consider my role as an open source advocate to actually be my primary thing. And it's very much my life mission. I hope to work on WordPress the rest of my life, but also just open source in general. I also support Drupal and Joomla. Anything else that's open source, I'm going to be a supporter of because I think when people choose that versus proprietary software, we're increasing the freedom and liberty in the world. It's incumbent on us to make open source, to make a better user experience, to make a better product so that people choose it, and then the world becomes more free, not less free.

**Lenny Rachitsky** (00:26:36):
It also feels it's important to you to, I don't know, white open source washing, like avoid people using the term when it's not true. And it's interesting in this case, that the thing that makes it not truly open source is the limit. There's a limit where you can no longer use it the way you want. Is that the issue?

**Matt Mullenweg** (00:26:53):
Yeah. And there's actually an open source OSI. There's a formal definition for what makes an open source license. And there's actually many dozens of open source licenses, and sort of public domain licenses and other things. So, it's also their stance that this is not an open source license.

**Lenny Rachitsky** (00:27:15):
Something else that I think is really interesting when it comes to AI and open source, you wrote about this and it blew my mind, such a good point, that the code that these models were trained on was open source code because that's all they have access to. They don't have Windows code, they don't have Shopify code. And what a cool, I don't know, another success story slash... I don't know. I guess, how do you feel about that, that all these AI models are trained on code you wrote in open source community?

**Matt Mullenweg** (00:27:43):
That's beautiful. It's one of the safest things to train on, right, because the license of open source very explicitly allows that. I also like to think about I have some window where my creative output is useful to society. And if you fast forward like 50 or 100, I do believe that the utility for proprietary software eventually approaches zero. When we're sending people to Mars, the operating system of the rockets and the devices and everything like that is not going to be built on the Windows NT kernel, as amazing feat of engineering that that proprietary kernel is. It's going to be built on an open source kernel, Linux or BSD or something like that. And so if you want to be part of something that sort of becomes the fabric of humanity's foundation, like things that allows a Cambrian explosion of things built on top of it, a renaissance of ideas, you want to be involved with open source.

**Matt Mullenweg** (00:28:38):
And so I really hope that more and more people... I'm a little bit of an evangelist here. I'm a missionary, where I really want to encourage more and more people to consider at least making part of their time, even if just a few hours a week, contributing to open source because you could be part of something that is a huge impact. And it's fun, especially if you're a younger developer or designer or PM or whatever, you can't walk up to Facebook and change your home page or say, "I'd like to change this feature," but you could come to an open source project, some of which have hundreds of millions of users. You could go to WordPress or, gosh, Bitcoin.

**Matt Mullenweg** (00:29:20):
Or there's all these things are open source, Chromium, Firefox, and you could actually change a feature or project management things or change the design or improve it. And that's I think really, really special. And sort of the thrill for me of knowing that code I wrote is now executing millions of times per seconds and millions of servers around the world, that kind of thrill, that high is, when I first had my first open source contribution, such a thrill. And I've been sort of chasing that and enjoying that ever since.

**Lenny Rachitsky** (00:29:53):
Say someone wants to actually do this, where do they go? How do they do this? Do they just pick up a project, go to WordPress. org and here's how you contribute? What's a next step there?

**Matt Mullenweg** (00:30:03):
Yeah, pick a project that you use or like. That's obviously a nice one. For WordPress, we have this... It's called make.WordPress.org. That's where we make WordPress. And there's different groups, there's accessibility, there's design, there's the core code, there's plugins, there's all sorts of ways. So, really whatever your talent is, there's people who translate, there's people who do support, there's people who write documentation, there's people who organize events, so whatever you feel like your talent in the world is either that you have or that you want to cultivate. I learned how to code while building WordPress basically. I didn't have too much formal training there. So, it's a great way to [inaudible 00:30:48] your skills as well, and work with some of the best developers and others in the world.

**Lenny Rachitsky** (00:30:53):
This also made me think about AI agents are coming around, Devin and all these AI driven coding agents. Do you have a prediction at when most of the code contributed to the open source projects will be Devin and AI agents such type projects?

**Matt Mullenweg** (00:31:10):
I think Google talked about 25% of their code or characters committed are now sort of AI-assisted, and they're probably on the bleeding edge. I don't know how much of WordPress's code right now is AI-assisted or something like that. But I think over the next five years it definitely approaches maybe a majority. And I'm actually very, very excited, so one of the big challenges that we have as a very open platform is we have this open plugin and theme architecture, so the 60,000 plugins and themes and the way WordPress works is these plugins and themes can modify every single part of the code, so you can really customize everything. However, many of these plugins and themes don't have the same sort of robust security and review process that core has. So, that's where when you hear about security issues with WordPress, it's very rarely in core anymore.

**Matt Mullenweg** (00:32:03):
We haven't had a remote exploit, knock on wood, it's like I think five or six years or something, but in the plugins it can be somewhat more frequent. And so one thing I'm very, very excited about the next year or two is actually more automated scanning because obviously that code basis is so many tens of millions, maybe over 100 million lines of code at this point. It's impossible for humans to review that, so we kind of rely on developers to review that and manage. And of course, we have bug bounties and everything do that, so when things get reported, we fix it quickly. But I can't wait for more automated scanning there, and I think that could vastly upgrade the security of open source. The other thing that's really exciting is right now you see people building apps and stuff and it's just sort of custom generated code, but I think the next generation of these models or sort of the next layer there is because...

**Matt Mullenweg** (00:32:54):
As everyone knows, just writing the code is just one part of it. It's maintaining it that really becomes the life cycle of it. And Stewart Brand's new book is all about maintenance, which I'm very excited about. He's publishing, I think, with Strike. And it's actually kind of open source. He's open sourcing the book, so you can see it being written online. But anyway, to go back, I think that if... And they're starting to do that, is when the open source models you say like, "Hey, build me a website." It actually installs WordPress, and then builds on top of that, and then customizes on top of that. Then you get for free that core engine that's always being audited and updated and getting passkey support or whatever the new things are sort of continuously. And then your custom stuff can be on top of that, which I think is actually a lot more powerful than building something proprietary or custom from the ground up.

**Lenny Rachitsky** (00:33:43):
I love this book concept about maintenance. My sister's partner has this quote that I've always come back to, "Life is maintenance." You basically... Everything you acquire and deal with... You get a generator for your house, you have to maintain that forever now. You get this backpack, okay, now you have to maintain this thing, keep it nice, nice jacket. Everything is maintenance. Everything in your life is just maintenance. And I wonder if that's what the book's about.

**Matt Mullenweg** (00:34:09):
Well, that's why I think technical debt is one of the most interesting concepts. There's so many companies as well that maybe have big market caps, but I feel like they might have billions or tens of billions of dollars of technical debt. You can see in the interface or how their products integrate with themselves through things. And I think about that a lot in our own company. We definitely have some products, almost a little embarrassed coming on because you have such great product people. And we have some variable quality around some of our things right now. If you check out Gravatar right now, I'm actually really proud of it. It's I think a really great user experience, very slick. But there's parts of... Well, I always say, "I'm the unhappiest WordPress user in the world," so there's parts of WordPress and WordPress.com that I'm a little embarrassed and ashamed of.

**Matt Mullenweg** (00:34:52):
We have a really large surface area that we cover with relatively few people, and so there's some parts we haven't looked at in a little while that we need to get around to. And it's a big focus for us this year, is actually kind of going back to basics, back to core, and improving all of those kind of nooks and crannies of the user experience, and also ruthlessly editing and cutting as much as possible, because we just launched a lot of stuff over the past 21 years that maybe is not as relevant today or it doesn't need to be there.

**Lenny Rachitsky** (00:35:21):
That sounds like excellent work for this AI agent of the future that's coming soon. There's one other area I want to mine and that's community, community building, building this ecosystem that you've created around WordPress. It might be one of the most successful, biggest communities on the internet. I'm curious just what lessons you've learned about what it takes to build a successful community, online especially.

**Matt Mullenweg** (00:35:48):
This is probably influenced by economics and jazz. And economics is all about systems thinking. And what are the incentive structures of how you set something up? And then in jazz is all about collaboration. So, if there's something unique I have for your audience, I would say it's don't just build a product, build a movement. And to the extent that we've been successful, I think it's that we give people something to believe in, a philosophy, a worldview. Even silly things, like we had this tagline in the footer of the WordPress.org when we started, it's still there, it says, "Code is poetry." This idea that we're not just writing code, we're trying to create something that can have elements. We name every WordPress release after a jazz musician for the past 60 releases or so. So, those sorts of things bring a little art and soul and some fun into it as well.

It doesn't have to be serious all the time. I think they can give something to believe in and work on and aim towards that's more than just a paycheck or more than just the utility, the base utility of the software. So, it's not just the software, it's also like: how are the meetups? How are people getting together? What events are you running? Are there forums? How do people contribute? Is there office hours or town halls? I do a lot of Q&A. So, what are the things you're doing around the software that's allowing people to get involved, that's inviting contributions, that's allowing people to build on top of it? I've studied platforms quite a bit like Microsoft and others, so our whole ecosystem of plugins and themes is part of what's made WordPress so successful, and the moat that we have.

**Matt Mullenweg** (00:37:33):
The core features of a CMS, you can kind of write with a few developers in a few weeks or something. It's not... It's basically CRUD operations, but to replicate those 60,000 plugins and themes, gosh, no one's done it. That's a huge moat. And proprietary services can create platforms. Shopify has a third-party ecosystem and things like that, but it's never a true platform. And a true platform, it's when your ecosystem makes more money than the core does. And so many times, whether it was the Facebook platform, I'm putting that in air quotes, or the Shopify platform, companies build on it and then they get the rug pulled out from under them because they're too successful.

**Matt Mullenweg** (00:38:15):
And then the sort of thing you're building on decides, "Oh, we want that money or we want that growth." And they sort of change the API or remove your access. There's so many examples of this, especially on, I think, Facebook and Shopify and others, where people got too successful and all of a sudden they knock on the door and they say, "Oh, that's a mighty nice app you have there. I'd love to offer you some warrants where we own a bunch of your company or we're going to shut it off," or those sorts of things.

**Matt Mullenweg** (00:38:42):
And again, you don't have freedom unless you're building on open source. That's why more and more companies and people are choosing... If they're going to build a business on top of something else, if you build it on an open source, you have that guarantee. Even if I grew devil horns and became evil and automatic decided to know whatever, WordPress would still belong just as much to you as it would to me. People can fork the code. They can still own it. They can still build on top of it. So, those things I think are really important.

**Lenny Rachitsky** (00:39:13):
What a segue to all of this drama that's swirling around you these days. I think a lot of people do feel like there's devil horns that have appeared, and so I'm excited to dig into this stuff. I find that every time you go on a podcast these days, if we don't get into this, everyone's just like, "Why is Matt not answering these questions?" Let's get into the hard stuff. So, I'm going to ask you some hard questions. For people that don't know what the hell's going on, they're like, "What are you even talking about?" or just have a sense something is swirling with WordPress and, Matt, I don't know what's going on, what's just like the high level overview of what's going on?

**Matt Mullenweg** (00:39:48):
So ,to set it, you can get WordPress from WordPress.com or us, but also you can get WordPress from dozens of other hosts. The biggest in the world are like GoDaddy, Hostinger, Newfold. It's not the biggest, but it is in the top 10 or something. It has about 700,000 WordPress installs. There's a company called WP Engine. In 2019, WP Engine started as very WordPress oriented and they contributed a lot to the community and everything like that. They were very respectful about distinguishing themselves from core, so people really realized it wasn't officially associated and everything. But in 2019 they got bought out by a private equity firm called Silver Lake. And anyone who follows business, when private equity buys something, there's some the good ones, but there's also many, many stories about how they can really hollow things out, really optimize our profits, become user hostile.

**Matt Mullenweg** (00:40:45):
Actually recently read a story where one of the reasons there was a shortage of firetrucks these LA fires was that the fire truck manufacturers have been rolled up by a private equity firm, and they've been raising prices and their supply constraint and things like that. So, there's literally a shortage in firetrucks right now because of private equity. And of course, if you look at healthcare or other things, there's so many examples of where private equity can really, I think, be one of the darker parts of capitalism. So, since 2019, WP Engine has kind of changed a bit, and they really stopped contributing to core and they started using the trademark in a way that was very confusing in the marketplace. And particularly in the past year, year and a half or so, we're just getting lot... I get a lot of support for requests for WP Engine.

**Matt Mullenweg** (00:41:36):
And when we do surveys, we'd find that 20, 30, 40% of people thought they were officially associated because how they were using our logo and presenting the brand and everything like that was very confusing to people. And as you know, if you don't protect a trademark, you lose it. And also the version of WordPress that they were offering actually wasn't our core vision of the functionality of WordPress. So to save money, they were actually turning off features like revisions. So, a cool part about WordPress that... actually one of my favorite features, is every change to every single post or page is saved forever, just like Wikipedia. So, if you make a mistake, you can always undo it. And of course, as building a great product, that sort of user safety of an undo is so critical.

**Matt Mullenweg** (00:42:19):
Now, obviously you have to store these revisions, so it takes up more database space. Now, it's trivial, it's megabytes, so on modern databases is not that big a deal. But to save money, they actually turned us off, so they broke the undo feature in WordPress to essentially save money. And so you have this thing where they're offering something called WordPress. I think I refer to it as a bastardized hacked up version of it. It's diluting our brand, and then also people think it's official. So, even close friends of mine were like, "Oh yeah, I signed up for this thing. I thought I was supporting you."

**Matt Mullenweg** (00:42:49):
And it's came to a head. So, past 18 months they've also... We contacted them and said, "Hey, you need a trademark license or something if you're going to use this or change how you're doing things," and tried to negotiate something and had many different term sheets over the months offered and different things, and they just kept stretching it out. And I was like, "What's going on here?" And I think part of what was going on is last year they tried to sell the company. So, private equity usually holds things for five to seven years, so they were kind of five years into this. They tried to shop it around and sell it. They weren't able to find a buyer.

**Matt Mullenweg** (00:43:25):
They said, "Well, they don't have any IP, and it feels like they're using your trademark, so they're going to have trouble with you. They don't have a license and things like that." So, while they were negotiating with us, it appears they were also preparing this lawsuit against us. So again, I've been very fortunate in my business career that we've invested in dozens of companies, we've acquired lots of things, by and large, 99% of the time people I've dealt with in business have been ethical, straightforward, honest. I haven't really faced any baldfaced lying or duplicitous behavior. Very, very rarely people who just say one thing and do another or are fraudulent in their behavior, but I think that was happening here.

And so I also just wasn't prepared for it. I was thinking I was a little naive and kind of didn't realize what was going on for a while. So, it came to a head, and at WordCamp US in September, I was like, "Okay, well, if you're still not going to even agree to negotiate, I'm going to give this presentation about how I think both private equity has messed up a lot of open source projects in the past, and how in particular, [inaudible 00:44:46] has done some very bad or evil things." And they were like, "Okay, go for it." So, I did the presentation. I think it was on a Thursday or a Friday. Kind of spicy. People were like, "Oh, can't believe he did that."

**Matt Mullenweg** (00:45:00):
And then on Monday they launched this with Quinn Emanuel, which is kind of the baddest, nastiest law firm, it's like who Elon uses when he sues people, launched this big multimillion dollar lawsuit against both me personally and WordPress.org, so the WordPress community and Automattic. And also they're spending millions of dollars a month on both lawyers and PR. So, they're doing... If you read... Oh gosh, who was the celebrity that they were recently talking about this, like dark PR stuff where they're boosting things on social networks?

**Lenny Rachitsky** (00:45:35):
Oh, Blake Lively and-

**Matt Mullenweg** (00:45:37):
Blake Lively, yeah, yeah.

**Lenny Rachitsky** (00:45:37):
... the other guy.

**Matt Mullenweg** (00:45:38):
So, all that stuff is happening, so there's... And I warned people. I think in the presentation I say, "Hey, there's going to be a smear campaign against me." And internally in the company, I was like, "Hey, they're going to dig up everything that's ever happens. Anything bad anyone's ever said to me is going to all of a sudden become a news item." And that has happened. It's been true. So, right now there is a portion of the internet that does think I have devil horns and everything. Fortunately, this is not my first rodeo. I know a lot of people think like, "Oh, Matt was nice for 20 years, and then got mean." But one thing, if you're really open and open source, sometimes you have to stand up the bullies, and you have to fight to protect your open source ideals.

**Matt Mullenweg** (00:46:19):
Otherwise people could take advantage of it in a way that ultimately can destroy everything you've created. So, this is probably the fourth time the internet has decided I'm the main character or really evil. And the previous ones we don't remember anymore. It's Hot Nacho or the Easter Massacre of Themes or these are the things that aren't even on my Wikipedia page anymore, but those seemed like really big deals at the time.

**Lenny Rachitsky** (00:46:43):
Those are your incidents. Those weren't like historical battles.

**Matt Mullenweg** (00:46:46):
No, no, these are things that, yeah, I was involved in.

**Lenny Rachitsky** (00:46:48):
Cool names at least.

**Matt Mullenweg** (00:46:49):
Including some things I had screwed up, like Hot Nacho was definitely a screw-up on my end very early in the WordPress side, but...

**Lenny Rachitsky** (00:46:55):
Wow. Okay, I'm not going to follow those threads, but those are great names.

**Lenny Rachitsky** (00:48:00):
... at WordCamp, and you said at the beginning of the talk... oh no, afterwards you were like, "I was really nervous to give this talk," and obviously you can see why. Just what finally convinced you this was time? Was it just to go, as you described, scorched earth nuclear? Was it like WordCamp is coming up and this is the moment to go public with this? Was there something else that kind of crossed the line?

**Matt Mullenweg** (00:48:24):
It was a unique opportunity because we were essentially saying that, hey, WP Engine isn't going to be allowed to sponsor WordCamps anymore. They're not going to be a... Because we had, again, up to that point really done everything to bring them in and have to be part of the community. So I really had to also explain to our community, hey, why we're going to be excluding this company that a lot of people saw as doing good. If you go to the WP Engine website, they have whole pages about how much they contribute and give back and how they... they do kind of greenwash or open source wash a lot of what they do. So all their marketing branding was around this positive stuff, and so I was like, "Hey, we need to just explain this case."

**Matt Mullenweg** (00:49:03):
But yeah, again, my defaults and how we've worked with, by the way, every other company in the WordPress space, many of which are much, much larger and make sometimes billions more in revenue than WP Engine, is collaborative. So if there's a trademark violation, usually it's not even lawyers get involved. It's just like there's a email, we have a conversation, we do a call, we talk about it. That's how things get resolved and that's my default. I'm a lover, not a fighter, and that's why this thing doesn't happen very often. I like to say that, yes, if WordPress community or whatever was doing this like every year or every couple months, yeah, you should worry about it, but it kind of happens every like 10 years.

**Lenny Rachitsky** (00:49:46):
So if I could mirror back the issues that you ran into, and I want to go through this a little bit more, the problems you had with WP Engine in this case. One is they were using the trademark both WordPress and WooCommerce without license, and they're just abusing it, confusing people. A lot of people thought WP Engine was actually Automattic and WordPress official. They weren't contributing to the project. They were just making basically a bunch of money and not doing the work off this company they bought and they're just kind of hollowing it out as you described. And then they're also cutting corners, making the product worse, and that kind of reflects on the whole brand of WordPress.

**Matt Mullenweg** (00:50:22):
That's a great summary, yeah.

**Lenny Rachitsky** (00:50:24):
Awesome. I'm curious just which of those three, or is it something even else that most bothered you about this? What's just like, "This is the thing that's eating me"? If I had to guess, it'd be damaging the legacy potentially of this thing you've worked on for most of your life. Maybe it's that, maybe it's just taking advantage of the community. What's the thing that you think is the root of this, just like, "Just this needs to stop"?

**Matt Mullenweg** (00:50:50):
Well, I guess the one thing I'd add to your list was as this was happening they were pretending to good faith negotiate. And in fact, at one point the executive, we were talking about her joining Automattic and running WordPress out of Oregon and when she thought... the VP and she going to sell, she was thinking about what was next. So yeah, a lot of this stuff was, I think that duplicitous behavior also kind of forced us to an edge more than even those other things that you mentioned. There's lots of companies that don't contribute back and it's not as big a deal. But yeah, the legal issue was definitely the trademark thing. So what pushed it to the edge? I think just the magnitude of the issue. They would refer to themselves as WordPress Engine in client meetings and other things. They were very cavalier about how they would imply their association with the project.

**Lenny Rachitsky** (00:51:46):
Obviously, as you can tell on socials, a lot of people are just really upset and a lot of people blame you. There's just, like I said, every time you're on a podcast or on Twitter, people are just like, "Matt, what about this? Why this sucks? Why are you doing this?" And I want to go through some of those things, but just not many people go through... I think you were like a hundred percent beloved hero of open source and internet and now you're in this... a lot of people don't like you. Just as a human, just what is that? How do you work through that? How do you deal with that? What's that been like?

**Matt Mullenweg** (00:52:17):
If you were kind of inside baseball with WordPress, it's actually a lot of people who have been unhappy with me over the years, and when we introduced something like Gutenberg, people hated it. Actually when we introduced a visual editor, people hated it.

**Lenny Rachitsky** (00:52:29):
You've had practice.

**Matt Mullenweg** (00:52:30):
These are huge controversies in the WordPress history. There actually hasn't been a fork or WordPress around all this latest stuff, but there was when we introduced Gutenberg. It's one called ClassicPress where people actually forked the software. So how I would describe it is previously like 1% of the world thought I was terrible, and now I feel like it's up to like 4 or 5%. So it's still not the majority, but as you know, something negative you feel seven times more than something positive. And when people are angry with you, it's kind of like restaurant reviews or whatever, they're more likely to leave a bad review than a good review.

**Matt Mullenweg** (00:53:14):
The people who WordPress, 98% of all the core developers have stayed and contribute and are working on the next version and are supportive and all these sorts of things. And part of the reason these folks are so good is they don't spend all their time on Twitter and Reddit arguing with folks, and also the arguments could be... they're very frustrating because people don't engage in good faith. They don't really change their mind when new facts are introduced.

**Matt Mullenweg** (00:53:43):
And so I've done my best actually because from the open source side I'm really used to engaging with things, and I think that's been one thing I've learned from this is in some forums it doesn't matter how you engage and especially if you have bots or other things running there. I'd leave comments on Reddit and immediately get like 40 downvotes. I'm like, "Hey, this is a article about me and I'm adding a fact to the thing. Why is it getting downvoted? This is very relevant to the discussion." But it's literally hidden. So when you see that thread, you'd have to click like three or four times to see the comment I had left, and so it can really change the perception. And then when you read these things, I think it's just very human nature. Even folks very close to me, if you read a thread and it's all super negative, it's hard to not be influenced by that because we're social creatures.

**Lenny Rachitsky** (00:54:36):
100%.

**Matt Mullenweg** (00:54:37):
Now the good news is I've had lots of credibility weighted support from people like Marc Benioff or other open source leaders or the core people in WordPress, Matias, Mary Hubbard, all the core committers. The international community actually, like just in Japan, they don't care about this stuff. So these are actually, if you look by number of commits and lines of code and everything like that, the folks who actually are most crucial in WordPress. So I feel like that's been a good balance as well for me because there are days where I'm like, "Gosh, am I an idiot?" or it could be really down reading all these things. So that is part of what allows me to balance and get back to that sort of positive, optimistic space that I think you need to be in to do great software and great work.

**Lenny Rachitsky** (00:55:27):
Yeah. The internet can be brutal. Let me go through a couple of specific things that people pointed out because I think you've been on a lot of podcasts and people haven't asked you these questions and I think a lot of people are just like, "But, Matt, what about this? This is really bad." So let me just ask you a couple of things here. One is there's just like a frustration in the community around the instability that this has just caused in the WordPress community. I'll read you a couple quotes. "Real people are receiving fewer projects on WordPress because C-suite are seeing WordPress as unstable because of this feud, and I work at Enterprise and we're very concerned about the stability of this platform and our projects." Just thoughts on that and the impact that it's had on the community.

**Matt Mullenweg** (00:56:05):
Yeah, I think this is until this gets resolved... which by the way I hope it is soon. I think there's no business reason for this to continue. I really hope that they come to a settlement or something. We're ready. They could end this tomorrow if they wanted to. WP Engine could. We can't. We're just defending right now. So it's really incumbent on them. All of our competitors, by the way, are like, "Great. WordPress, the king on the hill, all of a sudden we can use this." And so there's also not just from WP Engine, but also from all the competitors to WordPress, and all the people who would love to capture some market share, they're really leaning into this. So I've seen white papers, I've seen all sorts of things where people talk about this.

**Matt Mullenweg** (00:56:49):
We're actually in the next couple of days going to publish something really cool on the WordPress.org blog though that shows if you actually look at the numbers, like the activity, number of commits, plugin updates, downloads, installs of WordPress since September 20th when this all started, it's quite healthy. And so I'm not saying that there isn't examples of where someone lost a project or something like that. I'm sure it's happened. It's the internet's big. WordPress has so many millions of users and developers and everything that you're going to get some example. But by the numbers things are actually quite healthy, and in some ways it's not that there's no press is bad press, it's raised the awareness of WordPress quite a bit. So people who haven't talked about WordPress in years are now like, "Oh, Let's talk about it." And so a little bit of drama I think, I wouldn't do this all the time, but a little bit can be a good thing.

**Lenny Rachitsky** (00:57:39):
Okay, so one of the most common frustrations I've seen on the internet, people complaining is around the trademark. I don't know all the details, but my understanding is there's kind of a... you moved the trademark to be owned by the foundation and Automattic is exclusive rights to use the trademark. And I think people are like, "Oh, I thought it was the foundation owned it, but maybe Matt still owns it and then you're trying to monetize it through this agreement with WP Engine." Is there anything you can share there that'll make people feel and see your side of the story?

**Matt Mullenweg** (00:58:10):
Yeah, this is totally fair because it's complicated, but people are saying this has been private. This has all been very public and documented on the internet from the beginning. So WordPress.org has always been me personally, and I think because it's... part of the reason we started there is . com was not available when we started. So that's why we started on the .org and things like that. But I think people also assume .org means nonprofit or something, and that's sometimes true but it's not always. It's not a requirement of the .org domain. Then when I founded Automattic and when we did register the trademark that actually was registered under Automattic. So it used to be, for the first five years of the project or whatever, that Automattic just owned everything outright. And again, I had investors and the board and that was under the control of that.

**Matt Mullenweg** (00:59:06):
Now, as Automattic became more successful I was able to consolidate some voting rights and other things and at least later advocate. Also remember, I was like 21 when all this was happening, so I was not maybe the most savvy about legal stuff or didn't always have the best advice. So later as I learned more, I was like, "Oh, I want to actually take this out of the company and create a nonprofit." And so we ended up creating a nonprofit. Now the rules around 501(3)(C) nonprofits run the IRS are actually very strict. So that's also something as people assume, it's like, oh, doesn't the nonprofit run the software, and we applied for that originally and it was denied by the IRS.

**Matt Mullenweg** (00:59:47):
So we actually weren't able to put WordPress.org or the software itself under the nonprofit, but we were able to have sort of an educational purview. So what was eventually approved was sort of running the meetups and other things for WordPress, doing educational stuff. We sponsor a lot of learn to code or running workshops in other countries. We have this cool thing called do_action where we'll do a weekend where we take a bunch of nonprofits and build websites for them and stuff like that. So the nonprofit does a lot of exciting things there and then also negotiated with the investors and everyone at Automattic to actually put the trademark under the foundation.

**Matt Mullenweg** (01:00:31):
Now the compromise there was that Automattic at this point is running WordPress.com. So to continue running that, which at the time had already tens of millions of users and everything, it needed a commercial license. And so the compromise is that the foundation would kind of own the trademark and license it out for non-commercial purposes. I had a license to run WordPress.org because obviously I need that. And then Automattic would retain the commercial license and the ability to sub-license that, so to sell that to others. So this was kind of the grand compromise and create this tripartite structure. I was very inspired by the three branches of government. So there's sort of power in each of those that I think sort of checks and balances each side of it which is on purpose.

**Lenny Rachitsky** (01:01:23):
Wow. Okay, I get why it's complicated. I get why people would be confused. This makes me think about OpenAI had a really strange structure and that got them in a lot of trouble, and it feels like when you're 21 you're like, "Oh, this makes a lot of sense. What a great concept we've come up here," and then all this complexity just adds to a lot of confusion around what's going on. So thank you for addressing that. Another, there's a kind of related question I've seen a couple of times is just why don't you let that .org be run by a community. Why not just give that up to someone else and not just you that's there?

**Matt Mullenweg** (01:01:53):
Yeah, and the frame of that question is kind of interesting because it implies I'm the only person making WordPress which is obviously not true. If you look at the daily commits and activity and everything, it is run by the community. So it's hundreds of volunteers every day that are actually doing the day-to-day work and making the daily decisions and everything happens. So there has been a radical delegation. However, there's ultimately a hierarchy, and I'm the CEO, so I'm like the final final decision-maker.

**Matt Mullenweg** (01:02:24):
And so I think what people advocate for around this governance point of view is like, okay, well, install a board on top of you that ultimately makes decisions for the product or things like that. And there are other open source projects that have this structure. None of them have been successful as WordPress. So I think your audience in particular, is great software ever created by committee or does it more often reflects a vision of a leader or something that can allow us to... and I think particularly WordPress not just remaining relevant but actually accelerating growth over huge technological shifts over the past two decades.

**Matt Mullenweg** (01:03:12):
When we started there was dynamic web apps or DHTML or JavaScript wasn't really a thing, and then the social web and then iPhones and then all this sort of stuff that's changed over time, and we've surfed a lot of these technological changes which is very, very hard to do. Most products do not remain relevant over multiple generational changes like that, and that's been because sometimes we've had to make very unpopular decisions. Gutenberg is a huge part of why WordPress is relevant today, and it's actually an open source project we do. It's the block editor. It's actually bigger than WordPress because it's not just used on WordPress, it's used on every WordPress site, but also like Tumblr, other people. I would actually love if Squarespace or Wix adopted Gutenberg. It's meant to be like a really open source framework.

**Matt Mullenweg** (01:03:56):
But anyway, if we had voted for whether we should do that or not, everyone would've voted against it or the majority would have. It was really a few core people of us in the community, Matias, myself, other core contributors, Ella, Andrew Ozz, that said, "Hey, this is the future and it's going to take 10 years to do and it's going to be a long bet. It's going to suck for the first three or four years, and so everyone's going to hate it in the beginning."

**Matt Mullenweg** (01:04:20):
But then later with iteration, we've had I think now 200 releases of Gutenberg. We do sort of a very strict every two weeks release schedule since it started. It's going to get pretty good, and it's at that point now where it's actually getting pretty darn good. And the next phase of it, actually I'm so excited about, it's going to be collaboration. So all the real time co-editing like Google Docs and Notion has, it's coming out to this open source thing, and with the technological changes, we're actually able to do it peer-to-peer. So we don't need a centralized server. We can use WebRTC and other cool technologies.

**Matt Mullenweg** (01:04:49):
I mean, anyway, I'm going sidelining, but I think that sort of more... and if you look at a lot of great companies, there's a board or whatever but ultimately there's an executive, and some of the most iconic companies of our generations are ones where the executive retains some majority of voting control or other things like that which I've been able to do with Automattic and with WordPress. And I definitely think about succession planning and everything like that, but if or when I'm gone I don't want to pass it to a committee. I want to pass it to someone else who can have a role similar to mine and really sort of try to be a steward.

**Matt Mullenweg** (01:05:34):
There ultimately is a check and balance on that because, again, the community could leave. They could fork the software. People could change. And so you're "in charge" quote, unquote, but you're also at surface. So it's a lot more being like a mayor than a CEO and that you ultimately are accountable to the folks who are contributing and new users and everything like that. So I do feel like there is a balance there. Some of this as well is that there's some people who aren't part of leadership who feel like they should be. So if you look at the Yoast or Korean things, these are folks who actually don't have commit status. They haven't contributed WordPress over the years and serve our normal hierarchy of the meritocracy of how you get the ability to commit code or things like that. They're like, "Hey, I want to lead a release." That's cool, dude, but there's a process. We have different people lead releases over the years, but they kind of worked their way up to it.

**Lenny Rachitsky** (01:06:28):
This makes so much sense to me. That's one of the themes of the podcast, just the power of a singular visionary and leader, founder mode as we've all heard is trending these days.

**Matt Mullenweg** (01:06:38):
You made famous, yeah.

**Lenny Rachitsky** (01:06:40):
I wouldn't say that it was. Yeah, Brian shared it, but then Paul Graham pointed afterwards and then I renamed the title of that episode Founder Mode to [inaudible 01:06:49]

**Matt Mullenweg** (01:06:49):
I really want your [inaudible 01:06:50]

**Lenny Rachitsky** (01:06:50):
If I zoom out, what I'm sensing here is there's people that have this ideal of how something like this should run, but they've never actually worked at a place where a nonprofit board runs it, runs a thing, and have seen what that actually looks like. And so I think there's a big disconnect between the ideal in theory and how does great stuff get built.

**Matt Mullenweg** (01:07:11):
And one of the things I think we've tried to demonstrate with WordPress is actually it's kind of like a open source side and a nonprofit side and a for-profit working in concert. And one of the things people don't necessarily appreciate as much about why WordPress has been so successful is because of Automattic and things like Akismet doing anti-spam or WordPress.com having a free version of WordPress that is introduced over a hundred million people to the software in a way that you could just sign up for free. You don't have to pay for hosting or download it yourself or things like that.

**Matt Mullenweg** (01:07:43):
So that kind of for-profit, nonprofit, open source, working in concert I think is a really interesting model that we're starting to see a lot more companies do. It's actually very exciting to me that some of the things that were controversial when we started open source or distributed work are now the default for so many exciting new startups, and this whole ecosystems of really, really cool open source, like Cal.com for open source Calendly. There's so much cool stuff out there that actually there's a whole generation of younger entrepreneurs that I find very, very inspiring because they're also bringing modern design and web development and everything to open source which is very neat.

**Lenny Rachitsky** (01:08:21):
I anticipate a blog post one day, "I told you so, guys." Open source, remote work, I imagine there's a few more things there. There's one other thing I want to address. I haven't seen you talk about this. It comes up a bunch. It's around, this is very the weeds, but I think it's really important to people, and there's something here for a lot of people, the way you guys forked Advanced Custom Fields. So I think what happened here is you guys forked an existing plugin, I think somebody else's plugin, and then kind of pushed people to this plugin versus the original plugin. What can you share there?

**Matt Mullenweg** (01:08:55):
Yeah, this is very complex. So WordPress.org has kind of like a app store. After WP Engine started suing us, creating millions of dollars of legal fees and things, we blocked their access to WordPress.org. So this plugin they had, Advanced Custom Fields, wasn't able to be updated. At the same time a number of security issues were found in it, including some we reported, and so there had to be an update to it. So we're like, "Okay, we'll ship the update for you essentially." And then we were like, "Okay, I think we need to call it something different because it actually isn't theirs anymore." And they still offer Advanced Custom Fields on their own and people can download it from them, et cetera. So we made Secure Custom Fields which was originally under the same directory listing, so again, because we want all the users of it to get the security updates. This is controversial, and actually they actually got a preliminary injunction, so the judge said "Reverse this." So this has all been reversed by the way.

**Matt Mullenweg** (01:10:04):
There now is a separate fork under separate listing of Secure Custom Fields which actually we have a team on it, developers, designers, and we're creating... just like WordPress was a fork, we've actually forged this. Actually WooCommerce was a forge. A lot of things are forks. So we forked it and now have a new name, new everything that we're doing a lot of product innovation and improving it. So there's a separate project now and separate directory listing for Secure Custom Fields. That's kind of fast-forward to today. They now have access to WordPress.org again. They have updated the plugin. Everything's back to how it was before, and there's now this separate thing called Secure Custom Fields that the WordPress project is officially supporting.

**Lenny Rachitsky** (01:10:46):
So I'm hearing essentially you blocked WP Engine as a part of this, we're just going to simplify WordPress, reduce confusion. They're being bad actors in the space, so we're going to block them, and in that block, there's like a dependency where people couldn't do a thing that they needed to do. So you're like, "And the one that exists, there's a problem with it, so we're going to make that dependence... release a version that you can actually use and fix the security issues."

**Matt Mullenweg** (01:11:09):
That was the intention. I think that there was a lot of perceptions around it that were different, but yeah, that was the goal.

**Lenny Rachitsky** (01:11:14):
Okay. Okay. Great. So maybe just the last question. We talked about just a lot of people see you with devil horns these days. They think you're doing bad things and they don't like the approach you're taking. You talked about there's this WP Engine spending a lot of money on PR and hiring this agency. I guess is there anything else that... why do you think so many people are looking at you as the bad guy? Is it mostly that you think... just where do you think it's coming from? Why are comments always so negative? And we talked a bit about it, but anything more there?

**Matt Mullenweg** (01:11:46):
I don't know if I can say why. I do think one thing I've learned is that a lot of these things we've talked about are nuanced. So one essentially thing I've learned in this process is that it's hard to explain this stuff in 240 characters or the-

**Matt Mullenweg** (01:12:00):
... 40 characters. Some mediums do not lend themselves well to discussing this. And so I tried, but I'm participating less in Reddit or Twitter and trying to do more long form things like this, where you can actually have the context and things can't be taken out of context. Also, I think there's something where social networks sometimes are tuned to promote outrage. And it was very interesting. We ran a sentiment analysis recently. We were kind of looking at different social networks, analyzing all the comments, and we found, actually, that the negative, the sort of devil horn fraction on, what was it, like LinkedIn, Facebook, Instagram was like 8%. It's actually pretty small. On Reddit, it was bigger, I forget the exact number, but on Twitter it was 52%. You're like, whoa, what's going on there?

**Matt Mullenweg** (01:12:53):
And so there's something in the algorithm, and again, we can't see how the algorithm works or what the incentives are, but it can promote the most controversial things. I think that's not a novel perception. There's a lot of discussion around how social media might be creating more fragmentation in society, and I think this is just an example of that, where when you have networks, when people are getting the majority of the information from social networks, and those networks are not designed to provide nuance or balance, or even promote truth necessarily, misinformation can get spread far more than... What's the saying? Like a lie gets around the world seven times before truth has time to-

**Lenny Rachitsky** (01:13:30):
Get out of bed.

**Matt Mullenweg** (01:13:31):
Get out of bed, yeah. There's been a lot of that. So there's actually been a lot of misinformation, untrue things that go viral, and then the untrue thing gets like 700,000 views, and the correction gets like 20,000 views. So there's been some of that happening. When mainstream media has covered this, it's actually been a lot better. So there's been some actually really good articles in some business publications and other things that sort of look at a more nuanced and balanced view. And I think the podcast have been pretty good, but definitely on Twitter, I think you can get a version of all of this that is both, I think, not entirely true and also pretty more negative.

**Lenny Rachitsky** (01:14:15):
Yeah. I imagine people are going to be like, "Lenny, you didn't ask him this thing. Here's the thing he said that I want to learn more about." I'm sure I missed some stuff, but from an outsider's perspective, this all make sense. There's a company, I don't think PE companies are bad innately, but their job is buy a company and make it run more efficiently, and then oftentimes sell it for more. So it makes sense that they buy a company, make it more efficient, cut some corners, don't put a lot of effort into making it awesome, even though I'm sure there's awesome people working in there, trying really hard to make it great. And basically what I'm feeling is they got to a point where this is hurting the ecosystem. They're feeling really dishonest with working with you, and there's a stalling technique. And so, makes sense to me why you just have to stand up and fight back. And it's hard, it's hard to do that. Is there anything else along this thread before we move on to a different topic, anything else you want to share before we close out this chapter?

**Matt Mullenweg** (01:15:15):
Well, if people have more questions, they can come to WordCamp Asia. We're going to do an open Q&A there. We do town halls in the WordPress.org Community. There's a Slack people can get on and ask questions. So there is kind of a lot of open ways to engage, and I'm definitely happy to do that. I'm probably not going to do it on Twitter as much, but when there's longer form opportunities to have a discussion here, particularly if it's more like real time, like this, I'm very happy to.

**Matt Mullenweg** (01:15:43):
And that's why if you look at it, there's actually a big difference. WP Engine has not done any podcasts and no press. They don't respond to journalists, they don't talk about this. And I've done the opposite, where I'm really trying to be out there and engaged. And everyone's like, "Why don't you just let the lawyers do the talking?" And it's like, well, but we have community, and also I feel like we're in the right. So when you're in the wrong, you probably say only have the lawyers talk. When you're in the right, I think you should be out there and telling the story.

**Lenny Rachitsky** (01:16:10):
I remember at the end of your WordCamp talk, you were like, "Any questions?" after this big controversial talk, and I'm curious how it felt. All the questions initially were nothing to do with this. It's what it felt like. You're just like, oh, they already had these questions. They didn't even know what you said maybe, and I bet you're just like, wait, did anyone hear what I just said? Did it feel like that?

**Matt Mullenweg** (01:16:28):
Well, also, that's really like a WordPress community event, so it's a lot of the core developers and things, so they have WordPress questions, so that is something.

**Lenny Rachitsky** (01:16:37):
You're like, hello.

**Matt Mullenweg** (01:16:38):
I've now done hundreds and hundreds of these town halls and QAs, and I really enjoy it because you never know what's going to come up.

**Lenny Rachitsky** (01:16:42):
Yeah, okay. I want to talk about all the companies that you bought and will buy in the future. It's kind of like you're building a little Berkshire Hathaway. I think you've described it that way. It's kind of what it's feeling like. And Tumblr is really interesting. Until I started prepping for this, I didn't even know you guys own Tumblr. I haven't heard this story. Why did you guys buy Tumblr? What is going on with Tumblr? It was like a big deal back in the day. What is the current state of Tumblr? What is the story there?

**Matt Mullenweg** (01:17:10):
Oh, Tumblr is so interesting. At the time, I think it was one of our best competitors. They created this really amazing sort of hybrid of blogging and social networking. And if you kind of zoom back, a lot of things that are now standard on other social networks, even the ability to embed an image with a post, again, it was not supported originally on Twitter and other things. Remember they used to have, what was it, like tweet image, or you have to linked out to other things to post an image to Twitter. It wasn't native functionality, and Tumblr had these multiple post types. You could post a chat, an image. They were, I think, one of the first to support video, so they did a lot of, I think, product innovation under the leadership of David Karp, who's a really amazing entrepreneur and product leader.

**Matt Mullenweg** (01:17:54):
Funny story, both David and I were at CNET at the same time. They had hired both of us.

**Lenny Rachitsky** (01:17:59):
What an alumni group at CNET.

**Matt Mullenweg** (01:18:02):
They could have kept both of us probably. But anyway, the Tumblr, I forget the year, but they sold, I think the same time that Instagram did, for a similar amount, $1.1 billion.

**Lenny Rachitsky** (01:18:14):
To you or to someone else?

**Matt Mullenweg** (01:18:16):
Instagram bought by Facebook, obviously.

**Lenny Rachitsky** (01:18:19):
Right, right. Tumblr.

**Matt Mullenweg** (01:18:20):
And Tumblr bought by Yahoo.

**Lenny Rachitsky** (01:18:21):
Oh, wow.

**Matt Mullenweg** (01:18:22):
Who was at the time, again, Yahoo, we don't think about it now, but I feel a little old. But at the time, Yahoo was one of the internet giants and had recently Marissa Mayer, who was one of the big early people at Google, I think part of creating the API program and everything like that, was the CEO of Yahoo. This was, I think, one of her first big acquisitions.

**Matt Mullenweg** (01:18:46):
Now subsequently, obviously ,we know how Instagram went. I think people were like, "I can't believe you bought this for a billion dollars." And obviously now it's worth hundreds of billions. So that's had a really good trajectory. At Yahoo, I think things became more challenged. So again, this is a little bit of history, but Yahoo then had this thing where they owned part of Alibaba, which then became more valuable than the rest of the company. They had activist investors. I think they had some CEO switches. I think Marissa Mayer leaves or gets fired at some point. There's all this turnover, and I think Tumblr really languished under their ownership. And from what I can understand, the team was actually held back a lot from things they wanted to launch or ways they wanted to iterate. Then Yahoo merges with AOL, which is another kind of early internet thing. That goes for a little while. So again, then Tumblr's just kind of stuck underneath this stuff.

**Lenny Rachitsky** (01:19:43):
Tumbling along.

**Matt Mullenweg** (01:19:45):
Tumbling along. And then that gets bought by Verizon. So fast-forward to 2019. Verizon wants to get rid of Tumblr. And so they're kind of putting it up for sale and had a number of bidders. Automattic ended up buying it for a de minimis amount. I think it's been reported we bought it for $3 million.

**Lenny Rachitsky** (01:20:10):
What a deal. 3 million.

**Matt Mullenweg** (01:20:14):
But obviously, that represented a lot of value destruction over the years. Tumblr had had some tough times. They actually were banned from the App Store at one point for not moderating things well enough and having maybe a little too much porn. Obviously, Twitter [inaudible 01:20:31] porn. They maybe were a little too out there with it, and we're doing a good job filtering it and keeping away from App Store reviewers or whatever. And so Verizon, to their credit though, there were people bidding more. Actually I think a porn company was bidding on Tumblr that would've paid a lot more money. They really were looking for an acquirer that they felt like would be a good steward. From my point of view, I had such incredible respect for Tumblr as a product. And the community, still, despite all of this sort of stuff that had happened, I think at that point still was like, I forget the exact number, but call it 15, 20 million monthly active users.

**Matt Mullenweg** (01:21:14):
So really, sort of active core. And one of the things that's so fascinating is over half of that user base was under the age of 25. And actually had a huge, I think, it was like 25 or 30% LGBT+. I think a very unique place on the internet, where people could have a social network where they could be anonymous, they could put on different identities, they could be someplace their parents weren't, like Facebook or Instagram, really still could take a special spot. So we ended up buying it. Now, people are like, "Oh, you bought it for $3 million." But we bought it sort of taking on all liabilities, including, I think they were under investigation by the FTC, there were lawsuits. There was all this sort of stuff. So it was free like a puppy, not free like-

**Lenny Rachitsky** (01:22:01):
Free like beer.

**Matt Mullenweg** (01:22:05):
Had a pretty big team, I think 185 people. We were taking a lot of burners, burning a ton of cash, and that was 2019. And so, it's been, I think, a humbling experience running a social network. It was very, very different from all the other products that we've done. And I think there's some incredible things about Tumblr and that I'm still very excited about. So where WordPress has primarily a desktop and web user base, Tumblr is obviously like 85% app-based, has a younger demographic. And so part of the vision that now we're executing on is actually we wanted to create a path for people using Tumblr to actually it being powered by WordPress on the back end. So Tumblr users could unlock themes, customization, plugins, et cetera. Actually, we're in the process right now of migrating the half a billion Tumblr sites to WordPress, probably one of the largest data migrations-

**Lenny Rachitsky** (01:22:58):
That makes sense.

**Matt Mullenweg** (01:23:00):
... that's happened in a while. So we're kind of trying to do this in a way that's invisible to users on the front end, so changing at the back end while maintaining the APIs and the interface and everything. So it's a fun engineering project. I kind of posted this kind of call to arms, got a lot of fun people applying for Automattic, and we hired a lot of great folks around this sort of audacious project, this big hairy audacious goal. And so that's where it's at now. I've sort of ran it personally for a few years while we're doing turnarounds, but there's a great team there, but still challenged, still not profitable, so we're still subsidizing it from the rest of Automattic's businesses. Fortunately, the rest of our business have done really well, so we were able to do that, but I definitely want to get to a place where it's sustainable.

**Matt Mullenweg** (01:23:46):
And one of the things we're also experimenting with is can Tumblr have not just an advertising-driven model? I think ultimately the incentives of advertising social networks can lead to the kind of dynamics that you see on the more negative side of Twitter, Instagram, Facebook, et cetera. And so really trying to create a subscription model or a sort of first-party user-driven advertising, where you promote your blog posts or something like that, or you promote a WooCommerce product or something, where it's not a third-party ad ecosystem, which I think has a lot of weird code and malware and lots of stuff I don't love.

**Lenny Rachitsky** (01:24:21):
Wow, sounds like a lot you took on with this acquisition, and I love that you said you ran it initially. So this is a good segue to maybe my last question. I'm curious where this goes. Just how do you... Well, let me zoom out. There's a lot of people these days that are excited about roll of businesses. I'm going to buy a bunch of companies, make them better, make them awesome, save money, and then just keep building this holding company sort of thing. You guys are doing that, and it's working well. What do you look for? How do you decide a company's right for Automattic? What are the factors that are like, we should buy this, we can turn this around and turn it into a big success?

**Matt Mullenweg** (01:24:58):
I don't know if I would do another turnaround like Tumblr again, or at least not for many, many years. It's definitely a different thing. The vast majority of things we acquire, it's simply something that's done well, and we want to accelerate it, or sometimes acqui-hires, where we're plugging it into one of our existing projects, or we're taking the team and putting them on something we're already doing. So it's a really talented team. Tumblr, I think we ended up ultimately replacing 85, 90% of the team as well. So that's just very different. So I do think there are different ways of doing it, but if you look at our other acquisitions like Day One, et cetera, founder's still here, many years later, we're accelerating stuff like that. We brought it to Android, we're bringing it to it to web. It's more of taking something good and making it better. And probably our best example there is WooCommerce, which was a small company, I think 35, 40 people, based out of South Africa, and has obviously grown to... Again, I said Automattic makes about half a billion dollars a year now, and WooCommerce is a majority of that.

**Lenny Rachitsky** (01:26:06):
Speaking of that, actually, you haven't shared the revenue number. I know it's public. Just give people a sense of Automattic's revenue. Can you just share those numbers? Because I think it might blow people's minds.

**Matt Mullenweg** (01:26:15):
Yeah, I think we say publicly it's about a half a billion dollars in sort of ARR revenue right now.

**Lenny Rachitsky** (01:26:20):
Incredible. Okay. I have a question for you. It's kind of a hot seat question. As you talked, I wonder, I feel like people are thinking of this. So you've been talking about PE companies being often bad. You're buying Tumblr. You've talked about laying off a bunch of people, turning it all around. How's that different from a PE company, Matt?

**Matt Mullenweg** (01:26:39):
Yeah, and I agree with you that just because it's private equity doesn't mean it's bad. And also, something people say is like, "Hey, wait, don't you have private equity investors as well at Automattic?" And we do. Now, they own usually a small percentage, sometimes under 1%, and they don't have control of the company. So I think there's a distinguish. Is it a minority investment or a control investment? And with WP Engine, Silver Lake controls the company. And when they control the company, I think there's a spectrum of actions. Obviously, being more efficient is great, and we should all strive for that. And I think every business does, whether it's private equity or our business or things that are founder controlled. You always want to be more efficient. Now, there's some spectrum there where you over-optimize, or you could have dark patterns. Right now on WP Engine, it's very difficult to cancel your accounts. Actually, I think as of today, 45,000 sites have left, so they're, I think, down to 600. Yeah, well, because their customers have realized like, "Hey, this isn't WordPress, this isn't..." or, "They're suing the guy who started WordPress, so maybe we should not support this commercially." So we have the site wordpressenginetracker.com that sort of shows in real time the sites that are leaving. It's kind of an exciting thing to see that number tick up. Actually, maybe a good example as well, even though there's a lot of negativity, you actually look at how people are voting with their wallets. They're leaving.

**Matt Mullenweg** (01:28:07):
So I think you have to judge as well, just look at the track record. So one of the things I'm very proud of with Automattic is we are an acquirer of first resort, and we have founders that have sold to us. Paul Mayne at Day One is a great example that didn't need to sell. They're wildly profitable, could have run it himself for a long, long time, but people choose to join because they feel like we'll be good stewards of it in the future, and ultimately just have to look at the track record.

**Matt Mullenweg** (01:28:36):
So I think don't judge it by what it's called, judge it by the actions over time. And I hope to continue building that reputation for a place that's a good steward of communities and software and everything else for many years to come.

**Lenny Rachitsky** (01:28:52):
Matt, we covered so much. I asked you all the hard questions and more. Before we wrap up, is there anything else that you want to leave listeners with? Any last thoughts, comments, insights, stories?

**Matt Mullenweg** (01:29:04):
Yeah, follow me. I'm at photomatt, P-H-O-T-O-M-A-T-T, on Tumblr, Twitter, Instagram, everything like that. I post a lot about other stuff. I post a lot about AI and open source and other things. Some WordPress things in there as well. I have these life missions to democratize publishing and commerce. We added a new one last year, which is messaging, so it's in beta mode right now, but relaunching in a few months as a product called Beeper, which takes all your Telegram, Instagram DMs, Signal, everything, brings it all into one app. And you can do some really cool stuff like that, and especially when you start to imagine search or AI, local AI around that. So very, very excited about that relaunch. So I encourage people to, you check out the beta now, go to beeper.com/beta to get the new version, and we're going to relaunch that later in the year.

**Matt Mullenweg** (01:29:50):
So yeah, I'm very excited about that. It's kind of fun to be working on something that's at the stage where WordPress was in 2003, 2004. So WordPress is quite mature at this point. WooCommerce is kind of where WordPress was in 2010. And then the Beeper stuff, the messaging stuff is where we were in 2003. So one thing that keeps me excited is working at different stages of this.

**Lenny Rachitsky** (01:30:12):
This feels like a reason to be doing your approach to Berkshire Hathaway is just stay active in early stage stuff and not just optimize established things. So it's beeper.com, by the way, awesome domain name. Photomatt, what's the story of photomatt? You're into photography, I imagine, is the story?

**Matt Mullenweg** (01:30:32):
Yeah, it's a little bit of a pun. So a fotomatt is also, F-O-T-O-M-A-T-T, is a place that you would go to develop your photos back when you would have film and develop things. So originally my username was Saxmatt because I played the saxophone. Sometimes people mishear that. And also, I started traveling so much. There's been years I do like 400,000 miles of air travel, because I go around the world to go to WordPress events and meet the community. And as a distributed company, we do lots of meetups. And so, it became hard to carry my saxophone around. So my method of artistic expression became photography, and that's actually kind of how WordPress started, was actually originally a site where I could share my photos before Flickr, before Facebook and everything like that, sort of use this gallery software, actually open source gallery software, PHP software, to sort of share all the photos I was taking.

**Matt Mullenweg** (01:31:25):
And actually now on my website, I think I have over 38,000 photos I posted. And yeah, it's still one of the things I really love. So it's also a username that was available everywhere, and I still do it. So I'm actually going to the Maha Kumbh Mela, the big 300-million-person gathering at the Ganges River in a few weeks. And one, I'm just excited to experience that, it happens every 12 years, but two, I'm just really excited to take some time to do photography. And yeah, I really enjoy it.

**Lenny Rachitsky** (01:31:56):
You forgot to mention your website and your blog, your WordPress site itself, where you blog. Ma.tt, is the domain, which is amazing. I will point people to one of my favorite ritual you have on your blog, which is you share what's in your bag, you talk about how you travel all this, and I think every year you're like, here's the gadgets I use most and bring with me everywhere.

**Matt Mullenweg** (01:32:18):
It's my most popular post of the year by far.

**Lenny Rachitsky** (01:32:20):
I'm not surprised. You need an Amazon, just buy everything button. Yeah, because basically you're just trying to optimize for the least weight and most utility, right, out of all these gadgets that you're bringing with the on trips.

**Matt Mullenweg** (01:32:33):
Yeah, actually weighing it is something I just started doing this year because my bag actually got really heavy, got like 35 pounds or something. And so, some friends were like, "Hey, why don't we weigh everything and just go through." So now I'm posting the weights.

**Lenny Rachitsky** (01:32:45):
Oh my god. Okay. Anyway, we'll point people to that. Matt, thank you so much for doing this. This was awesome,

**Matt Mullenweg** (01:32:50):
Lenny, thank you so much. And I really appreciate the ability to discuss these things in a longer form. And also just your audience. Oh, I guess final thing I'll say is we're hiring a ton. So you have one of the most incredible audiences in the world. I recommend your podcast and newsletter to a lot of my colleagues. And so, if you're someone who loves this kind of stuff, I think there's a big opportunity at Automattic to have an impact on these things.

**Lenny Rachitsky** (01:33:11):
What roles are you hiring for most and where do people find these roles?

**Matt Mullenweg** (01:33:15):
Automattic.com, A-U-T-O-M-A-T-T-I-C. There's a Work with Us page. You can kind of see how we work. We're fully distributed and can manage that forever. We sort of started that. Another interesting thing is we actually pay the same salaries globally. So whether you're in California or Italy or Nigeria or wherever, we pay global salaries. So yeah, a lot of opportunities, and we're hiring for kind of everything, I would say, but particularly people with great design or product skills is probably one of the areas that you can have the biggest impact at Automattic right now.

**Lenny Rachitsky** (01:33:53):
All right. If you made it this far into the podcast, you should definitely apply. Matt, thank you. Thank you for being here. Bye, everyone. Thank you so much for listening. If you found this valuable, you can subscribe to the show on Apple Podcasts, Spotify, or your favorite podcast app. Also, please consider giving us a rating or a leaving a review as that really helps other listeners find the podcast. You could find all past episodes or learn more about the show at lennyspodcast.com. See you in the next episode.

---

## [10/15] Notion’s lost years, its near collapse during Covid, staying small to move fast, the joy and suffering of building horizontal, more | Ivan Zhao (CEO and co-founder)

**Lenny Rachitsky** (00:00:00):
The way you described the early years of Notion, you described the first three to four years as the lost years.

**Ivan Zhao** (00:00:05):
We try many different versions. The first version, okay, everybody can make and create their software, so let's just build a developer tool that's so easy that more people can do that. We tried that a couple of years and learned that actually most people just don't care. Our realization is actually let's hide our vision, which is everybody can create their software in the form factor that people do care. So what kind of tool do people use every day? Productivity software. It took us two years to realize we need to build a productivity tool. We called it sugar-coated broccoli. People don't want to eat the broccoli but people like sugar, so it gave them the sugar then hide your broccoli inside of it.

**Lenny Rachitsky** (00:00:40):
What other elements do you think are key to you finding something that actually ended up working?

**Ivan Zhao** (00:00:44):
What is the building a product or business. You want user. You want revenue. That's the product business. And building for something you want the world to have is building for your value. You have some taste. You have some aesthetic. There are different energy. You need to create a balance. Too much of yourself. Then there's no users. Then you're just doing our project. And too much for business, you're building a commodity.

**Lenny Rachitsky** (00:01:03):
The way you think about Notion, it's almost like a philosophy of how to work and be versus just a productivity tool. And so I'm just curious how you think about the relationship between tools and human potential.

**Ivan Zhao** (00:01:15):
Tools are extensions of us. And once they extend us, once we shape them, once we bring them to world, they can come back to shape us.

**Lenny Rachitsky** (00:01:28):
Today, my guest is Ivan Zhao. Ivan is the co-founder and CEO of Notion. Ivan is a really unique and also a deeply philosophical founder who doesn't do a lot of podcasts, so I'm really excited to share a glimpse into how he built one of the most beloved and most popular products in the world.

**Lenny Rachitsky** (00:01:45):
We talk about the first three to four years of Notion that he describes as the lost years, how he was able to get into a great school in China by winning a programming contest, the joy and suffering of building a successful horizontal product, plus his approach to staying lean and craft and making trade-offs and also leadership. Also, a wild story about how Notion almost died during COVID because the one database that everything lived in almost ran out of space.

**Ivan Zhao** (00:04:47):
Thank you for having me.

**Lenny Rachitsky** (00:04:48):
I know you don't do a lot of podcasts, and so I'm very honored that you're here. I want to start with the story of Ivan. Your background is quite unique for a founder of a $10 billion plus tech company, and I don't think a lot of people know it. For example, you grew up in a small town in China. And the way you got out of there, the way you got into tech is pretty interesting. Can you just walk us through that early years of Ivan and how you got out there?

**Ivan Zhao** (00:05:16):
Yeah. I think a small town in China, the definition, it's actually 4 million people. It is called Urumqi. It's in the northwest desert part of China. So I grew up there and then I moved into... My mom took me to Beijing, the capital of China. And that's actually how I got into programming, coding, because I'm from somewhere else and in order to go into good school in the capital, you need to win some kind of competition. And there's different paths. You can get at math or you can get at programming like Information Olympiad. I was really into computer games at the time so of course I picked the programming one so I can play computers all day long. And I win some competition and got me to a good school. So that's how I got into programming.

**Ivan Zhao** (00:06:05):
Later then, I moved to Canada. When I moved to Canada, got into college, did not study computer science since I already knew how to code, but a lot of video games. Did a lot of art actually, art and science. By the time I graduated college, I realized most of my friends are artists. They need to make their websites, get web portfolio made. And I'm the only nerd in my art friend circle so I made three or four websites and realized, "Oh, actually people don't know how to create with the software media, computing media." So that got me into want to create a product like Notion today which it allow more people to create tools, create software for their daily work and life.

**Lenny Rachitsky** (00:06:49):
Okay. So going back to get into a great school and to leave the small town, not so small, you had to enter a programming contest. And you placed first or second or how well did you actually do in this one?

**Ivan Zhao** (00:07:05):
Second in Beijing.

**Lenny Rachitsky** (00:07:07):
In Beijing, okay.

**Ivan Zhao** (00:07:08):
Pretty big. Beijing is a big city.

**Lenny Rachitsky** (00:07:11):
Okay. Incredible. Another stat or a story I heard is that you learned English by watching SpongeBob SquarePants. Is that real?

**Ivan Zhao** (00:07:18):
Yeah, it's real. I moved to Canada pretty late, 16 years old, and what I learned is in China you can learn English but it's typically just grammar and doing exams. What you're missing is the context, the culture. So you have to watch SpongeBob or Simpsons to get a sense of humor essentially. You can understand jokes. Watching cartoons, it's probably the easiest way to do that.

**Lenny Rachitsky** (00:07:45):
That's amazing. And there's another seminal moment in your path. I don't know if it was this point or later, but the Douglas Engelbart paper ended up being a very meaningful moment for you.

**Ivan Zhao** (00:07:57):
So while I was in Canada in last year of school working on trying to building website from our friends and building a creative tool for them, and then you just look into the history of a creative tool for software, for computing. Eventually arrived at 1960 and '70s. So you realize the first generation of computing pioneers, which is around San Francisco, Stanford areas, South Bay, they actually had the best ideas. For them, people like Douglas Engelbart, Alan Kay, Ted Nelson, those first generation pioneers, for them computing, there shouldn't be a separation between builders and users. It's the same medium. Engelbart's original paper called Augmenting Human Intellect, when I read that paper, it's like holy shit. If you are making software, if you know how to code or design, this is the highest leverage thing you can do for other people. So giving them the ability to use computing to augment their problem-solving ability or their intellect, that just got me obsessed with this problem and I want to start a company like Notion.

**Lenny Rachitsky** (00:09:05):
It makes me think of Steve Jobs's famous line of how the computer is a bicycle for the mind.

**Ivan Zhao** (00:09:10):
You know what? Steve Jobs is actually at fault of this in some strange ways. So the story is... Actually, the fact. It's not a story. Xerox PARC has working on the first-generation personal computers called Xerox Alto. Alan Kay was one of the main persons behind it. Alto runs down the system called Smalltalk, which is there's no separation between users and users' app. There's no thing called application. Everything is malleable. You can change the tools. So when Steve Jobs, the famous story is when he went to Xerox PARC to in demo with Alto, he does not... It's the first time he see graphic user interface, one of the first time, and it's also they present them with this Alto system that everything could change. But he did not see the power of it. Even when people would demonstrate like, "Hey," Steve Jobs say, "I don't like this direction of scroll bar direction. When you scroll up and down, it shouldn't scroll the opposite reverse direction." Then people just instantly change the scroll bar direction for him.

**Ivan Zhao** (00:10:16):
That's the power of the original Smalltalk Alto system. He only saw the graphic user interface. He did not see the underlying object or the environment power. As the generation of Steve Jobs and Bill Gates made PC, personal computing, popular and they stuck with this an application framework rather than the Smalltalk object framework. Then that has all the apps we have today and has the SaaS route we have today.

**Lenny Rachitsky** (00:10:43):
That vision of how products should be sounds very familiar and we'll talk about that later of how you think about Notion, but let's assume to the beginning of Notion, when we were chatting earlier, the way you described the early years of Notion, you started Notion in 2013 and some over 10 years ago at this point, you described the first three to four years is the lost years of Notion. And I think this is actually a really big deal for founders to hear about because there's all these companies these days, you hear these stats, they had 100 million ARR in two years, in under two years now. And you don't hear a lot of stories of companies of your scale and success that took three to four years to find product market fit essentially. What went on during these lost years as you described them and just how did you stick with it? That's a long time to stick with something that isn't working.

**Ivan Zhao** (00:11:32):
Because the goal is always building a computing tool. It's like what product is this? It's really hard to shape the product. The vision is, the dream is there, but the product is very... There's so many paths. We'll try many different versions. The first version to take, okay, everybody can make and create their software. So let's just build a developer tool that's so easy that more people can do that. We tried that a couple of years and learned that actually most people just don't care. The majority of people, they wake up, they have report due, they need to get their job done, they don't care creating software to optimize whatever they're doing. They don't care. So we give to our friends, give to investors. It did not resonate with people.

**Ivan Zhao** (00:12:22):
But we really want to build that tool so we just keep going and our realization is actually, let's hide our vision, which is everybody can create their software, in the form factor that people do care. So what kind of tool do people use every day? Productivity software. So that's why it came to Notion today. If you use Notion, Notion are more understood as the productivity suite, but our intent, and if you use Notion, more you discover intent, which is that it has a no-code developer power into it and you can create almost any kind of productivity software using Notion itself. That took us two plus year to realize. So actually the world is not like you. The world are not developer, designer mind. That the world is they only care what's in front of them and they're so noisy.

**Lenny Rachitsky** (00:13:16):
There's a quote that this makes me think about where you said, "The first version of Notion was more about what I wanted than what people wanted."

**Ivan Zhao** (00:13:23):
It's very much so because sense of maturation is you don't see the world just from your perspective but from outside your perspective. At tech, we were young. Took us multiple years. It hit your head straight into the wall to realize that. People just don't care.

**Lenny Rachitsky** (00:13:40):
I love the way you phrased that, that you have to hide your vision behind something that people understand and know how to use and...

**Ivan Zhao** (00:13:47):
We call it sugar-coated broccoli. People don't want to eat the broccoli but people like sugar, so give them the sugar then hide the broccoli inside of it.

**Lenny Rachitsky** (00:13:55):
Wow. The other thing I've heard is that you threw away your code every time, so you rebuilt it many times. You threw away the code each time.

**Ivan Zhao** (00:14:03):
That's true. Actually, it took us four year to get somewhere. First two year is that you build too much like developer product. Nobody cares. It took us two year to realize we need to build a productivity tool. Then it took another year to realize to build this out, but in the middle of that I realized we built on the wrong technical foundation. So eight, 10 years ago, there's computing before. Right now, all the web app runs on React. Before React wins, there's a competing technology called Web Component from Google. And it makes sense. Web Component feels like a Lego-like, the building block-like, and we're betting on that technology. And then we realize because it's so new, it's just so unstable. It don't know where the bug come from. It's from your source code or from the underlying libraries? Then we have to restart the company, rebuild the whole thing. Otherwise, we're going to run out of time. So we set a code base. We set a company so we can build on our own more orthodox technology foundation.

**Lenny Rachitsky** (00:15:10):
How did you actually stay solvent all this time? A lot of people want to keep working at an idea. Oftentimes they need to pay the bills. How practically were you able to keep working for three to four years? I know there's a story of your mom loaning you some money during that time.

**Ivan Zhao** (00:15:25):
Well, Chinese mom always can help, and I'm a single child. Yeah, actually my mom helped me kickstart the company because I'm Canadian. In order to move to US, you need to register a company. So my mom helped me with the initial and raised the money. I returned the money to her. Then we run out of the money so, "Hey mom, can I borrow that just to bridge us?" Which she did. I'm really grateful for that. How we bridged? How do you last here so long? Because the thing you want to create does not exist, which what is called Notions. It's a Lego for software. It doesn't quite exist. There's a Lego for Lego. You can see that in furniture exist, but Lego for software at the usable mass market adoption level doesn't quite exist. And you just want that thing to exist. And I grew up with Legos. It's the only toy I ever wanted, and I want the same feeling of creativity and playfulness to the toy that people can use every day. And my co-founder, Simon, feels the same way. Lego is the only thing he wanted for every Christmas.

**Lenny Rachitsky** (00:16:36):
Have you guys seen Magna-Tiles though? I have a one-and-a-half-year-old and Magna-Tiles are quite delightful. I think it's like a pre-Lego. The children can play them.

**Ivan Zhao** (00:16:45):
Magna-Tile?

**Lenny Rachitsky** (00:16:46):
Yeah. It's like they're little magnetic plastic planes and then you can build much bigger things really quickly. It's more for babies, but I'm having a blast.

**Ivan Zhao** (00:17:01):
Oh, I see it. It's like... Uh-huh.

**Lenny Rachitsky** (00:17:02):
It's a different version of Legos. I like that you're in real-time looking it up. You're like, "Okay, our new vision Magna-Tiles for software."

**Ivan Zhao** (00:17:08):
Now, most people know, "Oh, Magna-Tile." Idea is the same. Modular, right?

**Lenny Rachitsky** (00:17:13):
Yeah, creativity. Okay, back to your story. So there's also a moment where you moved to Japan. Just what was that about? Is that just escape and disconnect?

**Ivan Zhao** (00:17:20):
Yeah, that was during one of the rebuild phases. During the... We know what the product should look like. It should be a productivity software with a Lego power hiding inside of it. We build on the wrong technical foundation. And if we continue to build on the wrong ones, we're going to run out of money. Company won't exist. So we decided to lay off everybody. At that time, the Notion was five people. The layout I brought back to me and Simon, two people. And morale obviously there was really low. You have to say goodbye to your teammates. And so we have the idea, "Let's just go somewhere that we've never been to change the scenery a little bit." And Japan is always top on our list.

**Ivan Zhao** (00:18:00):
So the funny thing is if we... And we subleased our apartment and office. We're actually making money living in Japan and then San Francisco. So we did that for a while. We actually travel around the world for a while just to change it up, me and Sam just coding every day and design every day. That's some of the happiest moments. Birthday every day.

**Lenny Rachitsky** (00:18:27):
I saw a stat you're coding 18 hours a day. Here's the quote I heard, :We just code, code, code. Then hey, let's go for food. Then we go eat, go back to work, and do it again."

**Ivan Zhao** (00:18:35):
Because me and him working so well now. Even back then, it's like you know what each other other people are thinking and you can just cross through the problem space really quickly. The technical product space, design space, and just non-stop of shaping stuff.

**Lenny Rachitsky** (00:18:55):
So maybe just to close out this thread, for people, for founders that are either struggling and just can't find a thing that's working, "I've been working on something for a long time," I'm curious what advice you'd share for sticking with it. And I'll share things I've heard you say so far and I'm curious if there's something you'd add. One is you just believe this needs to exist in the world and you need to really feel this, "I need this to be a thing." I think there's an element of staying lean, like you've let everyone go and it's just you and Simon again. There's also this element of disconnecting almost and just going to a different location and just like, "Let's just reset." What other elements do you think are key to you finding something that actually ended up working?

**Ivan Zhao** (00:19:36):
I'm lucky and Simon lucky that high is never too high, low is never too low for us, so somehow it wasn't feeling too down. Whenever I feel down, I just go to sleep and next day I'm just reset. So that's lucky for me. Definitely don't be afraid to reset. I think courage is quite important because oftentimes you're working on things don't matter, but momentum just took you there. Your first point of building something you want the world to have. What is the building a product or business? You want user. You want revenue. That's a product business. It's almost like a sports. The market is the arena. Then you'd want to optimize the scorecard where it's building for winning. And I grew up playing sports. I like to compete so I like that.

**Ivan Zhao** (00:20:34):
And building for something you want the world to have is building for your value. You have some taste. You have some aesthetic. You have some values. You want the world to have more of that. They are different energy. I realize actually fairly recently, they're really different. Depends on which day I wake up, I might be in different mood for things, but building for value it's more lasting and more fulfilling. Looking in the thing we're building today and looking back, I find most proud of thing I create something authentic to myself and happen to be also useful for others, and that just keeps you going. And that feels like a more durable energy source for all those dark years, loss years during Notion, and still every day for me.

**Lenny Rachitsky** (00:21:24):
It's interesting you say that because also there's this aspect of it wasn't working initially because you're building it for yourself and not for people, but what I'm hearing is it's still important to build a thing that you are still excited about but also have you go back and forth. Here's what the business needs and here's the thing I'm excited about.

**Ivan Zhao** (00:21:42):
Yeah, really a cue. Almost like a therapist, right? It's true. You're building too much for your own self and value without realizing at the end of the day, if you're building a product and tool has to be used by others, you need to create a balance. Too much of yourself, then there's no users. Then you're just doing our project. You're just doing a research project. And too much for a business, you're building a commodity. So where's the spectrum? Yeah, it's never ending spectrum. It's interesting.

**Lenny Rachitsky** (00:22:17):
Yeah, okay. So I'll summarize some of those things you shared of just how to stick with it and stay with an idea and not give up. So I love that you said just get sleep, very Brian Johnson of you, just like, "Get some sleep when it's a real down day. There'll be another day tomorrow." Really simple but...

**Ivan Zhao** (00:22:32):
It's like a daily personal physical reset. You can reset your code base. You can reset your mental model.

**Lenny Rachitsky** (00:22:39):
Okay. And then there's also, I love these points. Don't be afraid to reset, as you just said. Tobi Lutke was on the podcast. He said the same thing. "Just be comfortable with some cost. I have done all this already and I will throw it away and start again and that's okay."

**Ivan Zhao** (00:22:53):
Yeah. I think it's not just a self-help way to say don't be afraid to reset. That's like, that's okay, that's fine. I think the more interesting point here, it's like you can create progress through better abstractions. And that thing compounds faster, can catch up to all the things you build much quicker than you ever thought. Or humans are not thinking, not good at thinking in terms of abstraction or exponentials. We're thinking in terms of linearly. If you just reset it and you find a better way to do it, you can get all the thing you have to some cost recovered really quickly.

**Ivan Zhao** (00:23:29):
So actually going back to the computing pioneers part, small talk, one of the first system and a huge influence for Notion was really tiny code base and inspired by Lisp, which is another programming languages and probably a hundred lines of code or something. The kernel of things could be really small, but just like math. It can compound. It can have complex behavior that unlocks so much value and things for you. But if you just find those right, you can catch up to all the things you did. You are free to lose really quickly. So I think that's the kernel of why reset is so powerful.

**Lenny Rachitsky** (00:24:10):
And we're seeing exactly what you're describing in LLM advancements these days. All these companies have been working on this for so long and then they've cracked an abstraction of how to think about scaling these systems. And now just people launch them and are immediately where the companies that have been working this for decades are today because they are building off these abstractions as you described in these.

**Ivan Zhao** (00:24:34):
Trying to caught up the US really quickly.

**Lenny Rachitsky** (00:24:36):
With DeepSeek, yeah. The point you also made about momentum, be weary of momentum taking you in direction and moving in a different... not being stuck to that direction is exactly the way I think the chain of thought models network actually where generally LLMs are like, next word, next word, next word, next word. And if they ever make a wrong turn, they're stuck. They keep going from that path. And these chain of thought models are now good at just like, wait, let me rethink. Is this actually the right path or should I start again? So I feel like AI has almost figured out exactly what you're describing.

**Ivan Zhao** (00:25:06):
Interesting.

**Lenny Rachitsky** (00:25:07):
Oh, man. Okay, last question about the early years. Everyone's always wondering what does product market fit feel like? You worked on it for three to four years. What was the moment? What would it look like? What was different when you're like, "Okay, this is going to work"?

**Ivan Zhao** (00:25:20):
I think going back to me and Simon, high is never that high, low is never that low, it never hit us as a binary state. Just like, "Oh, good. We have people who care about this thing we make now. Oh, good. People reach out to us who are paying us." And it's a very gradual ramp. Maybe that's why early days when it's really the lost eras, it doesn't feel too low because it just... Even for Notion today, it feels like it's so small in terms of where it could be. It just they keep going, right? It's a less of a milestone way to thinking about things. It's more just like, "Can we do the same that's in our head and better than we did last week?" way of thinking about things. So there's a such movement that product market, boom, milestone achieved. Didn't feel that way.

**Lenny Rachitsky** (00:26:13):
I've heard that from a lot of founders actually. Was there a moment in that point of just like, "Oh, this is different," or, "Maybe it's going to work this time"?

**Ivan Zhao** (00:26:23):
I think for a while, okay, once we start revenue, product grows faster now. Investors start knocking on the door was like, I remember one day it's like there's a dog food, dog treats sent to our entire office. So first of all, office wasn't public, the address. And the dog treats, why do people want this so much? So that was a moment I paused a little bit and I guess there's enough attraction for investors.

**Lenny Rachitsky** (00:26:55):
And the dog treats were trying to... It was like a gift to be like, "Hey, you should talk to us. We're sending this fun gift."

**Ivan Zhao** (00:26:59):
Yeah, because of the way how we just hire someone in the office as a dog. Then I think we post on Twitter or something. And I said, "Why did this show up to our office?" Someone really hustled into where we are in our office address and follow us on Twitter.

**Lenny Rachitsky** (00:27:17):
Did you end up taking their money?

**Ivan Zhao** (00:27:19):
Not the first time, yeah.

**Lenny Rachitsky** (00:27:20):
Okay, later. Okay, it's long game.

**Ivan Zhao** (00:27:24):
No.

**Lenny Rachitsky** (00:27:25):
Awesome. So I've never heard that before. Sign product market fit as VCs are starting to... You start getting a lot more messaging and cold outreach from VCs.

**Ivan Zhao** (00:27:33):
Actually, I had one of our investor, it's really helpful because all those years you just like there's no feedback loop. You just go for it. Then the feedback loop gradually show up. Then for a while it's, oh, VCs start knocking on doors. So I should talk to those people. The people like what we're doing. I did some meetings, quite a few of meetings. Maybe it doesn't... I realize and one of the members is saying, "Ivan, what are you doing? You clearly don't need money. You're just trying to feel good to do external validation about this." And I said, "Oh, that's so true." It doesn't help us make a better product and the truth is with what customer tell us. Then we just went back to building. I went back to hardcore building, no meeting modes. That's where the dog food story came about and realized, "Oh." It's interesting.

**Lenny Rachitsky** (00:28:30):
You mentioned this investor, they said it was really helpful. Is you want to give them some credit or do you want to keep-

**Ivan Zhao** (00:28:35):
Oh, Shana Fisher. She's in New York.

**Lenny Rachitsky** (00:28:38):
Okay, cool.

**Ivan Zhao** (00:28:39):
Yeah, she's like another therapist, right?

**Lenny Rachitsky** (00:28:43):
**Ivan Zhao** (00:30:20):
Yeah.

**Lenny Rachitsky** (00:30:20):
It's still in the bank. You're nodding, if you're on YouTube. You didn't have a salesperson until you hit over 10 million ARR. You hired your first PM at 50 people. You've always kept the team generally really small. Why is that been important to you? It's very cool now. Everyone's like, "Of course, that's how it should be." But for the past decade, that has not been the case. You've always been that way. Why has that been so important?

**Ivan Zhao** (00:30:44):
I think going back to the abstraction system way of problem solving, I think we're lucky that me and Simon and Akshay, we have the skillset you probably can run a whole company, which is a couple of us, I can design, I can do marketing, storytelling, close sales deals. So you realize you don't need a lot. But when you can do a lot at the same time or hire people who can do that, naturally keep the company small. And you all know you're doing product management. The overhead is actually more from internal communication. It's really hard to get people's mind to be aligned on things, to see the world in the same way. And the part that you do need people, maybe you can solve better through systems, through better tools.

**Ivan Zhao** (00:31:39):
Notion itself is a meta tool to build other tools. So we pretty much run everything on Notion. We use the same mindset to build our company. And accidentally, that keep our headcount low, keep our company profitable, which then puts you on a positive treadmill of you don't have to go for the next 18, 24 months to find money. You can just focus on building.

**Ivan Zhao** (00:32:03):
And also because your team's small, we have this internal Notion called talent density. We don't try to track number of people but we try to track how talent-dense, revenue per employee we are. And people want to work with either more talented people. So it's a positive company group.

**Lenny Rachitsky** (00:32:27):
I wonder how much of this is actually from being around for so many years without success of, "We just have to stay very lean and save our cash because otherwise, we'll die." Do you think that was a formative experience to inform how you want to operate or is that always something?

**Ivan Zhao** (00:32:42):
No, I wouldn't say we're, Notion, is a cost saving-first company. I like fancy chairs. I like furniture. But we're not wasting money. I think it's more just from a taste or approach to problem solving. I just believe better system is much better than brute force through people.

**Lenny Rachitsky** (00:33:05):
When people hear this idea of staying lean and staying small, it sounds great or we're going to be super efficient and lean and smart with our money and down dense. It's very hard to do and it's very hard not to hire more engineers, more designers. What advice do you have for folks that want to operate this way? What has allowed you to actually be successful while staying lean and not having as many engineers as competitors, many designers as competitors?

**Ivan Zhao** (00:33:29):
I think just understand abstraction or system is a better curve than pecan curve, right? Linear. We internally help other people and understand this. Internally we use the metaphor that Notion's a small bus. The bus, the smaller the bus, it's easier to turn corners, easier to accelerate, easier to maneuver. The bigger the bus it is, bigger the boat or bigger the bus, slow down. And as a leader in the company, you decide who sit around you on the bus seats. That dictates how fast our overall bus moves, dictate your work and life experience at this company because you pick your roommate, you pick your seat mates. That metaphor clicks with people inside a company and overall help us optimize to make the bus link.

**Lenny Rachitsky** (00:34:18):
I've never heard of that metaphor before.

**Ivan Zhao** (00:34:22):
It probably came up somewhere but... Or it does not.

**Lenny Rachitsky** (00:34:25):
Small bus. So along these lines actually, so I visited the office recently and I noticed that it's just a very cozy vibe. And I learned that you had a rule of no shoes in the office for a long time until the last office, that you all ate around one table for a long time, that you try 30 different shades of warm white on the walls before you chose. Why is that important to you? Why is it so important to be so thoughtful about the office experience?

**Ivan Zhao** (00:34:53):
Maybe there are two dimension part of it. One is the pragmatic part. You just want office to be a pleasant experience to be at. Therefore, most office, the top light feels like hospital. You're just like, "Oh, man." And the white is so pale and the floor is so dark. Why use some kind of cream, make floors more friendly colors? And don't use top light. Top light is evil. So just the office feels cozy so people spend more time. You feel more creative, more at ease in the office space.So the vision work we have is should feels like artist studio or should feel like your home. And that's why most our office furniture are home furnitures. It just feels cozy. That's more so people spend more time, feels more creative, juices flow better.

**Ivan Zhao** (00:35:42):
The other world just like at least personally for me, it hurts the eyes if you just see ugly things. It's more from a value aesthetic front. It's like we talk about ergonomic chairs. Does it hurt your back when you sit on bad chairs? But you have more visual improve from, at least for me from the eyes. If the chair looks ugly, the wall looks ugly, it hurts. So it's better to not have thing that hurts.

**Lenny Rachitsky** (00:36:05):
You also have a really interesting naming convention for your conference rooms.

**Ivan Zhao** (00:36:08):
Yeah, that's true. Yeah. We name our conference room after timeless tools in history. So there, I'll give you an example. iPhone's obvious one, original Macintosh, various different form of chairs, Lamy's 2000 pens, Toshiba rice cookers, and other ones because they're inspirations. They're just like at the end of the day, we're creating a tool. We're creating a meta tool. A lot of people to create tools, software tools. And Toshiba rice cooker changed how people eat rice in Asia for a hundred million, tens of a hundred million people. The Sony transistor radio is the first one to shrink something small and useful for people. And those things change people's life and last for decades. What it's like to create a software product like that? I want to inspire my team to think that way. Because software, and especially tech, it's every six months, every 12 months cycle. We don't think enough about creating something that lasts. I care creating something that at least the form factor lasts longer than 18 month.

**Lenny Rachitsky** (00:37:21):
There's a quote that you tweeted once that I think of as you talk about this from Steve Jobs. "The problem is that there's just a tremendous amount of craftsmanship between a great idea and a great product." I don't know if you remember tweeting that, but just what do you think of when you hear that?

**Ivan Zhao** (00:37:36):
Yeah, I think the key word here is craft. Internally, our company philosophy called crafts and values. Craft is your skill set, your taste. Value is your personal value and how do you see the world. Craft is interesting word. It's like about apply your value to some technical know-how and to make more clever trade-offs to create something new and useful and just keep doing that.

**Ivan Zhao** (00:38:09):
My wife often refer me as a wood cabinet builder. That's how at least my mindset training towards building Notion is like, "Oh, can I make this wood cabinet more beautiful and more useful and feels nicer on your hand?" And that's like you have aesthetic direction towards it and you have your technical know-how to actually make things happen. Then you need to do permutation and trade-off in your head where on paper and to get there. That to me, that's craft. And building product, to me at least, to me feels that way. Building business feels that way. Building company feels that way.

**Lenny Rachitsky** (00:38:44):
It's interesting that so much of this conversation is this and the way you think about building this company is this balance between practical, useful things people need and business and practical stuff, and then the value of building something you're proud of and craft. And there's always this trade-off almost of speed and quality, and I know that's an important element for you. Just thinking about trade-offs between decisions, so talk about just trade-offs, just how you think about making a trade-off.

**Ivan Zhao** (00:39:13):
Yeah, I think this is quite relevant especially for product makers and business makers is there's no free lunch. You don't get something for free. You have to give up something. Then what do you give up? It's essentially you give up the right thing that market or your user wants at that given space and time. It's just the craft of building a business or building a product. And that the market is so dynamic, especially now with AI. The optimized function for the market changes so then you need to make new trade-off and new technology emerges. I always feels like AI language model feels like a new type of wood. It feels like aluminum. It's a new type of material. So you can make... Mass air travel wasn't available until aluminum become cheap enough that people can make airplanes that support this at cost. And it's like computer wasn't there until semiconductor becomes... It's like require new technology to unlock new way to making trade-offs, and then you need to balance the technology trade-off with human behavior trade-off.

**Ivan Zhao** (00:40:35):
As a human, ever since we got out of Africa, we're set, right? That's a constraint. It's invariable. And every generation pick up some new things but after you're 16 years old you don't want to learn new things. So those are there are the people trade-off, technology trade-off. There's some macro. There's a different dimension of things just cooking together that come together as a product more as a business than what is that? And I think a product maker, business maker's job is to find that sweet spot of all the multiple dimensions, then create something has a right to exist. At least it's more durable to exist.

**Lenny Rachitsky** (00:41:16):
And I'm hearing there's this thread of just like with new technologies, what is now possible. And I know you guys are doing some cool stuff with AI that I'm going to get to that is unlocking some cool new ideas. But before I get there, I want to talk about just you as a leader. At this point, you've been at this for 12 years, something like that.

**Ivan Zhao** (00:41:31):
Like that, yeah.

**Lenny Rachitsky** (00:41:32):
And if you don't mind me saying, you're a soft-spoken leader, which is you're not like the archetype of what people imagine is like the CEO of a 10 billion... And I'm sure you guys are valued much more now. I don't even know. That was probably an old valuation. I think it's great for people to see leaders like you that are not necessarily the classic archetype of CO, and I imagine there are things you've had to work on and build and lean into that aren't natural to you to step into this role of this increasingly growing, high-scale business. What are some of the areas you've had to most build and learn to do that didn't come naturally to you?

**Ivan Zhao** (00:42:10):
I guess you've never been in a business meeting or brainstorm session with me. You're not there.

**Lenny Rachitsky** (00:42:15):
Haven't seen that side of Ivan Zhao.

**Ivan Zhao** (00:42:18):
Yeah, I wouldn't say I'm the most soft interaction person at work. It's actually the reverse is true because I grew up in China. People way more direct. People just say what they want, say what they think. And you move to California, you move to US, you move to the West, you felt wow, everybody says everything's wonderful, everything's nice, but that's not true. I would say Notion's ethos probably more like a East Coast rather than West Coast, so somewhere in between. It's more direct.

**Ivan Zhao** (00:42:51):
What do you want to learn? A bunch of things. I think the early days is we talk about that the world's not like you. The world don't care about you so you have to shave off the idealistic part of you to go something that's like the world actually cares, the sugar coat of broccoli. You have to hide the broccoli within something, the sugar pills. So that's one. That's more self. That's more myself.

**Ivan Zhao** (00:43:20):
As company grows, you realize... I'm pretty good at storytelling. So that's a one-to-one influence. But as a company grows, you realize you need to be one-to-many storytellers. That's a skill. The one reason I try not to do podcasts and all those things, oh, it's actually it drains energy in a different ways. I prefer just building product and brainstorm sessions. Then you realize it's a necessary craft for me to pick up in order to change the shape of the company, the business I'm building. I treat it like a craft. There's some things skill that's in the video game. You need to pick up something to unlock something else and to make new demand, you trade off with yourself and the business. That's fun though. Every 12, 18 month, Notion's like a new company or at least they require different skill set coming from me. So I need to pick up new things. And it's an infinite game and infinite games are more fun.

**Lenny Rachitsky** (00:44:19):
I love this idea. I love that you keep coming back to this idea of there's the ideals and the values and the vision and what you're trying to do, and then you have to find the way to frame it and package it so that people actually understand and want it. And that's how you get in.

**Ivan Zhao** (00:44:35):
Yeah. It's like human minds are resistant to change, and how do you land in people's head? Through my best word marketing and positioning are for. So you need to find the sweet spot to get in. And you also be truthful. It's not just deceiving. So deceiving is not truthful. You can fool other people once or twice, then there's no future. It has to be actually tied back to something genuinely the value creating or the exchange with the other person. So yeah, it's a craft. Storytelling is the vast dimension of making trade-offs.

**Lenny Rachitsky** (00:45:15):
I love this word, trade-offs. Comes up again and again too. It's so interesting that there's these threads that have come up again and again in our chat. Along that journey of becoming this leader that you've become, what would you say has maybe the biggest surprise or most unexpected part of the journey of something you've had to learn to do or something that didn't turn out the way you expected? Just as a personal growth story.

**Ivan Zhao** (00:45:36):
If you use the product in the past three years, you realize Notion product, you realize, "Hey, we actually ship bunch of things not so great." Two years ago. Actually last year, 2024, is the year that I can say we ship good stuff at good velocity and good quality and align with our values. We get lost there for a year, a year and a half shipping something not according to our value, not according to my value. Notion, we call Notion is Lego for software. We ship non-Lego pieces into our product. We're still there. We're still cleaning up part of it. That's a realization. It's like going back to the value part, it's like if you create this thing called a product or business, you attract people are value aligned to it. Then if you're trying to optimize too much on this competition revenue side of things, forced to introducing something anti-your-value, then the system, it's like there's organ rejection with your employees, with your customers.

**Ivan Zhao** (00:46:42):
I'll give you a concrete example. For a while and still is, project management is one of the most important use cases for Notion. And you can get a better project management tool just by hard coding things like sprints, milestones, all those things into your product, or you can do it in the way the Notion are being, through Lego pieces. What are the sprint? Sprints are clusters of a task that group together. So it's a new Lego. So introducing Lego is much harder, slower. You can instead we hard-code a sprint concept into the product. And this doesn't quite fit. And took me at least a year, a year and a half to realize that's not the way we should continue building Notion. We should go back the original Lego way of building the product. So we changed quite a bit internally. Now, it feels good now. Building according to your values is the meta point, at least for me.

**Lenny Rachitsky** (00:47:43):
Okay, I got to follow this thread. What is it that you changed that allowed you to come back to your first principles? Was it like you step... Is it founder mode was the answer? Is it people, personnel shift? What allowed you to change the way things were going?

**Ivan Zhao** (00:47:58):
I would say all of that above, but especially just release the sprint product through our community and customers. Then it's like what is this? It's like underpowered compared to other competitor products to doing product management and it doesn't work well with the rest of Notion like I said. And if you talk with engineers, they'll say, "Okay, there's this part of Notion you have to touch the code base. That's just weird. That's your hardcore too much into it. From all the dimension technical front, calling a customer. And when you use the thing it just doesn't feel right." So there's another saying that if you build in a Lego way inside Notion in the code base or product, the system work for you. If you're building non-Lego way, the system work against you. So in some sense, we're creating a tool that has emergent behavior, inter-channeling that emergent behavior to unlock more values.

**Lenny Rachitsky** (00:48:52):
So I'm hearing as you launched it, it just didn't go well. Everyone's just like, "What is this? This isn't feeling good." And there's a moment of realization of I see. Here's what we did wrong here and we should come back to this original abstraction vision of what we're trying to build.

**Ivan Zhao** (00:49:04):
That took nine months, a year to realize sometime.

**Lenny Rachitsky** (00:49:11):
Along those lines actually, people come on this podcast and they share all these stories of things are going awesome all the time. And this was a great example of it didn't. I'm curious if there's another story of let's say a crisis that you all went through when things were looking pretty bleak for Notion along the journey of building Notion.

**Ivan Zhao** (00:49:31):
Yeah, one of the bleakest one, it's when we... During COVID, we just couldn't scale up our infrastructure. For the longest time, Simon's really good at don't do premature optimization, so for the longest time, we Notion runs on one instance of Postgres database. And then we find the beefiest machine. We keep scrolling, find a beefier future machine to scale our user base, but then we're running off even the largest instance there is for Postgres. So there's a doomsday clock that when we're going to truly run out of this space to store everything in Notion and Notion got a complete shutdown. So we stopped building any new features, all hands on deck, almost every engineer in the company trying to solve that problem. Eventually we did, but it was a close call.

**Lenny Rachitsky** (00:50:21):
How close are we talking about?

**Ivan Zhao** (00:50:23):
If I recall correctly, probably in weeks running out of the time. And then as you approach the limit of what Postgres can do, behavior becomes sporadic. You really don't know which day going to hit you. But we just need to go as fast as you can to become sharding problem.

**Lenny Rachitsky** (00:50:39):
Yeah, I was going to ask, so the solution is sharding the database?

**Ivan Zhao** (00:50:39):
Yeah, sharding.

**Lenny Rachitsky** (00:50:40):
Okay, cool.

**Ivan Zhao** (00:50:43):
Don't do as late. Yes. Don't do premature optimization but plan ahead a little bit. Don't go late.

**Lenny Rachitsky** (00:50:49):
How long did you have from when you launched this doomsday clock to time running out? Was that a few months?

**Ivan Zhao** (00:50:54):
Maybe a bit longer. Yeah, in the month, less than six but more than three, something like that.

**Lenny Rachitsky** (00:50:59):
The bittersweetness of COVID just ramping up certain businesses.

**Ivan Zhao** (00:51:03):
People just run like they have to use online productivity software, collaboration tools.

**Lenny Rachitsky** (00:51:08):
Yeah, blessing and a curse. Speaking of a blessing and curse, this is a great segue to where I wanted to go in the final area I want to spend time on which is building horizontal software and building software that bundles together a bunch of different stuff. Notoriously hard to build a horizontal platform that does a lot of things when there are often point solutions that are very, very good at that one thing. And it's interesting. If you look at the timelines of companies that have built horizontal products, they all take a long time to build and finally find product market fits. It's actually a really common pattern. And when we were talking about what would be fun to talk about, the way you described it is the joy and pain of building horizontal products. So let me just ask broadly just what have you learned about what it takes to successfully build a horizontal platform type of product?

**Ivan Zhao** (00:51:56):
First of all, no regret. And second, I wouldn't want to build anything else because going back to the value, Lego for software doesn't exist and Lego is a horizontal thing. So that's the thing we want to build. We always want to do that. So we did not start to optimize for business but we're optimized for that vision.

**Ivan Zhao** (00:52:19):
Learning-wise, I think segmentation is quite important because people can use a Lego for different things. Only hardcore Lego fans care about Lego bricks. Most people care about Lego boxes. And they actually want the Lego box to be ready-made. When you unpack the box, the set is there for you, right? That's what we're learning a lot, especially move up market. There's this term that took me a while to learn. It's called solutions. You need to be a solution for enterprise customer, you need to sit somewhere on a P&L to optimize for their business where due third risk. That's Lego box. It's not a Lego brick. Segmentation related to that. So you need to shift your mindset as you more towards B2B, more towards move out market. I wish we have done earlier. For the longest time, I've stalled too much in the Lego brick mindset, now in the solution Lego box mindset.

**Lenny Rachitsky** (00:53:14):
That's such a good metaphor. I feel like even if you're not building Legos for business, just this idea of what is the box that you are selling to people, how's it being positioned? How do you picture it? What are the value? Props such a good metaphor.

**Ivan Zhao** (00:53:29):
If you're building vertical software and naturally your vertical is the box, right? So you know you have 1 or 2% of your selling to. Pretty straightforward that your market constrains you and no judgment. People like you, you can go that way, but then you just hit the wall off the market. The advantage of building horizontal, there's no wall, at least for in our space. We, Notion, go after entire software market, but then you need to create a wall yourself. So to make your go-to-market distribution, to create the spot in people's mind, your customer's mind more clearly for them and for your go-to-market teams. That's why where solutions is one of my favorite word internally to rally the sales team or the product team. You think that way, but then you need to hold in your head, make sure you're still building bricks behind the scene. Otherwise, you pigeonhole yourself into the best spot, like what we did with project management sprints features.

**Lenny Rachitsky** (00:54:25):
So speaking of that, so I don't know if you know this. I ran a survey recently where I asked my readers what tools they use most, what tools they love most. And it went out to my entire subscriber base. We've got 6,500 people filling out the survey. And Notion more than any other company placed very highly in many categories. For example, I have the notes here, it was the second most popular project management tool after Jira. It was the fourth most popular docs. Which is interesting because you think Notion would... Notion is known for docs and it's interesting, that was the lowest one actually. And then it was third in CRM, just behind Salesforce and HubSpot.

**Ivan Zhao** (00:55:02):
Yeah, we did not intend to build CRM, but what is a CRM is relational database. That's why we give people that brick. That's a relational database and they can build CRM themselves. I think the good advantage is if a customer use Notion, they can address those three, four use cases in one place. Especially for our startup mid-market companies, their need for each of the vertical use case is not as complex so they can have all the information in one place, good for their teams, good for AI actually. That's a huge market change that's like we did not expect until recently. And save their costs, which is more and more people care about the bundling purchase nowadays. And our approach for that is, yes, we're number two in project management, number what? Number four in CRM, but we're interested in more bricks to make us number... Move up the categories in ranking. So it just takes time, but that's our approach.

**Lenny Rachitsky** (00:56:05):
Yeah. Well, it's working whatever you're doing there. So say someone is trying to build a horizontal tool like yours. There's a lot of founders that are trying to build something that can do a lot of things really well. Do you have any advice for that first use case? Just figuring out something that initially works like you're talking about segmentation, is there something there of like, "Do this if you want to find any success with a horizontal tool"?

**Ivan Zhao** (00:56:28):
First, I wouldn't recommend it.

**Lenny Rachitsky** (00:56:30):
But you wouldn't do it differently?

**Ivan Zhao** (00:56:32):
I wouldn't do it differently myself, but I wouldn't recommend it. It's a problem. The problem space too large to have best practice, but I can share something that's relevant for us. Notion, we always want to build a meta tool, a tool to build the lack of our software. We somehow stand up upon document notes as one use case. And that just gave us a large top of the funnel that there's a 1 billion plus people use this use case every day. So that fuels our growth. We call our internal strategy called B2C2B. All those consumers, personal user use Notion for the most simple way you can use a computer or your phone, which is note-taking or document-sharing. And then they realize, "Oh, Notion can do more of that." There's relational database power, you can do tasks, you can manage track other things. Then they bring Notion to work.

**Ivan Zhao** (00:57:24):
Half our B2B customers coming from prior personal users, and most of them are using Notion for notes and talk in the first place. So pick. Well, at least we stumble upon a use case, a horizontal use case to give us a large top of funnel that help us grow our more verticalized enterprise use cases, and that's the reason where we ship a calendar product last year because which other category of software has 1 billion plus users? There's document notes, there's calendar, there's email, right? That's why we're also working on the email product right now.

**Lenny Rachitsky** (00:58:01):
Yeah, man. Watch out, everyone. And then you mentioned AI and it's such a good point that AI is best when it has data. And the fact that you have all of this stuff already in there gives you a lot of really interesting opportunities to leverage AI.

**Ivan Zhao** (00:58:17):
We definitely did not expect language model. It's such a gift for everybody building tools, right? Complete change the material you can work with. One realization, it's you have a surface area that people spend daily work with, especially during writing and managing your tasks and project. It's really easy to slice the language model writing AI capability into it. So that's the first part we built. That realization is AI is so good at reasoning and understanding and searching things, and we can do a much better job of finding and searching things if all the information are together. That's what we realized. AI is really good with bundled offerings. AI is really good with horizontal tools. So that's the second phase, we call it. The first product was our AI writer product. Second product is AI Q&A or connectors. Please look at all the information in Notion and give your answer.

**Ivan Zhao** (00:59:18):
And then we also need to work with the external connector because there's things that are living in Jira, living in Zendesk that other customers still rely on. So we need to build AI connectors. But more and more information coming back to the Notion core. I would say the third one, which is even more fascinating, it's for the longest time and it's still is one of the biggest weaknesses of building for Legos, it's hard to piece together. It's not everybody can put together a Lego set from scratch. There's always the builders and user with the Legos. But guess who is really good at piecing things together, assemble things? Especially things like since Sona 3.5. AI is so bad at writing code. Coding is just assembling things together. So now we're looking at holy shit, we spent the last five, six year building all those Lego blocks for knowledge work. If we're just putting AI coding agent on top of it, you can create any kind of knowledge, customer software, customer agent for whatever your vertical use cases you need. So that's the most fascinating approach for me, and we did not expect this at all.

**Lenny Rachitsky** (01:00:33):
Thank you, AI. Is there anything else along the lines of building horizontal products and bundling that you think is interesting to share or important? Otherwise, I have one last question I want to ask you.

**Ivan Zhao** (01:00:43):
I think market is like waves. There's... Who said this? There's two-way to build business, bundling and bundling, right? There's too much of a zig and the zag. Actually, my favorite version of this is there's a classic Chinese literature called Romance of Three Kingdoms. It's great novel. It talked about the three kingdom era of China and the opening sentence of this novel, it's, "Empires long united must divide, long divided must unite." That has always been bundling, unbundling. It's one of my favorite book to read when I was a kid, but business works same way. When there's too much, you can see this.

**Ivan Zhao** (01:01:29):
It's like before computers, everything works on paper. Our knowledge work are done through papers is fully democratized medium. Then PC happens during the '80s. The first era is a piece there actually are so many applications. There's early database software, dBASE, it's quite famous. It started dBASE 2 because it gives them credibility. Oh, they have been stick around for some time. So that's the first unbundling phase of software computing. Then Microsoft bundled everything back into one suite in the '90s. Then the SaaS unbundled it. Now, we're at the tail end of SaaS. There's so many verticalized SaaS average company to use almost a hundred tools. It's madness. So there's more the market shifting towards more a bundling approach. And with AI and with the macro, so there's more value to be created through bundling, at least for now. The market could shift again. So understand this trend, I think, helpful to see should you build a vertical solution or should you build horizontal solution because it does different things.

**Lenny Rachitsky** (01:02:40):
I love that story. Okay, so last question. Something that one of your early investors, Finn Barnes, suggested to ask you. I'm curious where this goes. There's this, and you've touched on this a number of times, just the way you think about Notion, it's almost like a philosophy of how to work and be versus just a productivity tool. And so I'm just curious how you think about the relationship between tools and human potential and humans and how we live in the world.

**Ivan Zhao** (01:03:08):
The tools are extensions of us. That's why our office room named as timeless tools. They extend us a little bit. And once they extend us, once we shape them, once we bring them to world, they can come back to shape us.

**Ivan Zhao** (01:03:29):
One of my favorite quotes like the Marshall MacLean quotes, "We shape our tools. Then after, our tools shape us." I think that's probably too philosophical for building product or business, but there is a sense thinking what are you bringing to the world that will come back to bite you or shape you? And are you extending the part, the so-called good part of human nature, or are you extending the part that might be more zero-sum, might be more negative, right?

**Ivan Zhao** (01:04:05):
For me, what is Legos? Lego is creativity. Lego is beauty. Software to me feels like lacking both. It's definitely lacking a lot of creativity. It's so rigid. So I believe both are human nature that worth amplifying. You can build another business that amplifies a different part of human nature. There was Sequoia famously invests in seven sins or seven human natures of human because they're so powerful if you just latch onto them, you can create a business, you can create a product. But at least I prefer to amplify creativity and beauty in the domain of software. To me, that's aligned with my values and I think can at least shape the market, shape our user of our product towards the better part of themself.

**Lenny Rachitsky** (01:04:58):
It must feel so good to have a product that is so aligned with the way you want to see the world and actually working and growing at this rate and scaling and becoming this, I don't know, part of the ether of the world.

**Ivan Zhao** (01:05:11):
It feels good. Yeah, it feels good that some of the most heartwarming thing is still it never gets old when you walk by coffee shop and see people using Notion. Oh, it feels good. And it feels good that we see people in our community can create a living selling Notion template, Notion apps, that they're not a software engineer. And going back to the original mission of when people create software, I think that's one of the most fulfilling thing, at least as a maker of tools can experience.

**Lenny Rachitsky** (01:05:39):
That last point, I think people don't realize, so people are making millions of dollars selling Notion templates on the internet like at Etsy and other places.

**Ivan Zhao** (01:05:48):
Consulting templates, yeah, and they're not programmers. I think I would say that's the heart of that because their domain expertise, they're YouTubers or creators. They have lifestyle brand. They know certain things but they're not makers of software. Then they can use Notion, package their workflows and expertise into Notion and templates and make limit with it. It's awesome, all that.

**Lenny Rachitsky** (01:06:13):
Yeah, millions of dollars is it's crazy. Ivan, before we get to an abridged lightning round, I'm curious if there's anything else that you wanted to touch on think might be useful for folks to hear before we get to a very exciting lightning round.

**Ivan Zhao** (01:06:27):
I think people in tech, I wish more people look beyond tech to steal good ideas. It's like Tech Hacker News Twitter are so focused on the now and what's in front of it, what happened six months ago, versus humanity. If you just read books in other industry, you can look sideways. If you go back to history, there's a massive amount of patterns and shapes and trade-offs you can steal from and you can make what's in front of you much more interesting. You could give you... People figure out clever patterns in whatever domain in the past. You can just take in front of you. And I wish more people do that. I think it would be a very interesting way for product makers, business maker to solve the problem in front of them by stealing outside of from the domain of tech and business. So at least it's very inspiring, very useful for me personally.

**Lenny Rachitsky** (01:07:26):
It makes me think of the quote, "Good artist copy. Great artists steal."

**Ivan Zhao** (01:07:30):
Great artists steal, yeah. Well, Steve Jobs stole that from Picasso or something who stole from former artist probably.

**Lenny Rachitsky** (01:07:37):
Well, this is actually an amazing segue to our very abridged lightning round. And the first question is... And by the way, welcome to the lightning round.

**Ivan Zhao** (01:07:43):
Oh, okay.

**Lenny Rachitsky** (01:07:45):
The first question is just what are a couple of books that you find yourself recommending most to other people? Could be along the lines of what you just described or could just be generally.

**Ivan Zhao** (01:07:53):
I think the domain that are interesting the most is the complex system domain. You can look up the term. I think more and more people talk about this, but thinking a system, complex system when all the different things merge together, it creates emergent properties. Talking about ants, talk about beads, talk about life itself. It's just so fascinating how do with few primitives, few Lego bricks, you can create a thing called life. That thing just, it's sugar for me. So I love reading in that domain. And this is really helpful for create product, at least a horizontal product because you're trying to channel the energy, smaller parts to create something that the sum is much larger than its parts.

**Lenny Rachitsky** (01:08:43):
Is there a specific book that comes to mind or is it just generally that's a cool area?

**Ivan Zhao** (01:08:48):
That's a cool area to your attention to.

**Lenny Rachitsky** (01:08:50):
Next question. Do you have a favorite recent movie or TV show you've really enjoyed?

**Ivan Zhao** (01:08:55):
I like to watch old documentaries. Maybe this is another area or category too. There's quite a few on YouTube. People make really good documentary in the '80s, in the '70s. That's like all the old BBC ones, they're just excellent and they have a strong opinion in them. It's no longer just general education thing. They have a direction. They have a taste. Go look it up. Oh, yeah. One is a really good one to get started called Connections. I think it's called but the gentleman's name is Burke. It's about how different things from different domains inspire other domains, and usually he used 30 minutes or 60 minutes to chain together a bunch of connection of stories. It's really good for technologists to watch. Highly recommend.

**Lenny Rachitsky** (01:09:49):
I feel a very consistent pattern throughout all of these answers and your entire conversation of just emerging properties, connections, Legos, building abstractions.

**Ivan Zhao** (01:10:00):
Yeah, I think I did Enneagram. My Enneagram, it's 7 and 7. 7 is, it is actually perfect with what we just talking about. 7 is creative, finding connection, see the forest and tree. 8 is they call Challenger. It's like competitive AR optimizing. So true energy accessing me.

**Lenny Rachitsky** (01:10:23):
Oh, wow. It's all makes sense. I got to take this Enneagram. This comes up a bunch on this podcast.

**Ivan Zhao** (01:10:27):
Right, yeah.

**Lenny Rachitsky** (01:10:29):
Final question. Do you have a life motto that you often think back to, that you often repeat in your head of just like when times are hard or just to keep going with something you're working on that you find useful?

**Ivan Zhao** (01:10:41):
I like to think things as a craft. You just make it better. Make for yourself. If it's unique enough for yourself and useful for others, things will follow.

**Lenny Rachitsky** (01:10:51):
Ivan, thank you so much for being here. Two final questions, working folks find you online if they want to follow-up on anything, and then how can listeners be useful to you?

**Ivan Zhao** (01:11:00):
Probably find on me on Twitter, Ivan Z-H-A-O. It's helpful give us feedback about Notion, about our product. That's the best help.

**Lenny Rachitsky** (01:11:12):
What's the best way to do that? Is it like DM, Ivan, or is it-

**Ivan Zhao** (01:11:14):
Yeah, just DM me.

**Lenny Rachitsky** (01:11:15):
Okay.

**Ivan Zhao** (01:11:15):
DM me. Yeah, that's probably the best way.

**Lenny Rachitsky** (01:11:19):
Okay. Oh boy, here you go. And then you guys are hiring. Anything specific you're looking for? Anything people should know if they're like, "Oh shit, I want to go work here"?

**Ivan Zhao** (01:11:29):
We're trying to hire misfits. So if you think you're a misfit, if you're exceptional at many things especially, you want to build Lego for software, you want to take interesting spin on AI with Lego for software, then DM me.

**Lenny Rachitsky** (01:11:45):
Amazing. Ivan, thank you so much for being here.

**Ivan Zhao** (01:11:47):
Thank you for having me.

**Lenny Rachitsky** (01:11:49):
Bye, everyone.

**Ivan Zhao** (01:11:50):
Bye.

**Lenny Rachitsky** (01:11:53):
Thank you so much for listening. If you found this valuable, you can subscribe to the show on Apple Podcasts, Spotify, or your favorite podcast app. Also, please consider giving us a rating or leaving a review as that really helps other listeners find the podcast. You can find all past episodes or learn more about the show at lennyspodcast.com. See you in the next episode.

---

## [11/15] Building Lovable: $10M ARR in 60 days with 15 people | Anton Osika (co-founder and CEO)

**Anton Osika** (00:00:00):
... Lovable is your personal AI software engineer. You describe an idea and then you get a fully working product. The reason is to enable those who have had such a hard time finding people who are good at creating software that's been their absolute bottleneck and let them take their ideas and their dreams into reality.

**Lenny Rachitsky** (00:00:19):
You guys hit 4 million ARR in the first four weeks. You hit 10 million ARR in the first two months with just 15 people. You're the fastest growing startup in all of Europe. How did you decide on Lovable is the name. It's so sweet.

**Anton Osika** (00:00:31):
The best word for a great product is that it's lovable. A lot of jargon that I like to use to emphasize what we should be striving for is building a minimum lovable product and then building a lovable product and then building an absolutely lovable product. So I took that jargon with me in the company name.

**Lenny Rachitsky** (00:00:47):
People would wonder just what jobs will be more important, what skills will be less important?

**Anton Osika** (00:00:51):
Doing a bit of everything. Being a generalist, I think much more important than it used to be. If I'm putting together a product team today, I would really obsess about getting as many skill sets as possible for each person I hire.

**Lenny Rachitsky** (00:01:03):
What have you done that has allowed you to grow this fast with so few people?

**Anton Osika** (00:01:07):
People love the product. That's the driver of the growth.

**Lenny Rachitsky** (00:01:15):
Today, my guest is Anton O-C-K. Anton is co-founder and CEO of Lovable, which is essentially an AI engineer that takes an English prompt and codes a product for you in minutes. You can then talk to it, iterate on the product, and then launch it to the world. It's one of the fastest growing products in history. The fastest growing startup in Europe ever, and as Anton describes, their goal for Lovable is for it to be the last piece of software that anybody has to write because it'll be able to create all future products for us. They launched just a few months ago in the first four weeks, hit 4 million ARR in the first two months across 10 million ARR, all with just 15 people. Absurd. In our conversation, we covered a lot of ground, including a live demo of Lovable, how their team operates, how they hire, what has most enabled their team to scale this quickly with so few people, pro tips for using Lovable, how it all started, how he recommends you build product teams going forward with tools like this existing, what skills will matter more and less going forward?

**Anton Osika** (00:05:19):
It's a pleasure to talk to you, Lenny, great to be here.

**Lenny Rachitsky** (00:05:22):
I don't know how you have time to do this podcast. Your life must be insane these days with the pace at which you guys are scaling, just how much is changing in AI every day. So I just extra appreciate you making time for this. I think you said it's 10:30, your time, is when we're doing this.

**Anton Osika** (00:05:39):
I'm a bit tired, yes. Mostly from the crazy pace of everything, but yes.

**Lenny Rachitsky** (00:05:45):
This is going to be an invigorating conversation and you're not going to be able to sleep.

**Anton Osika** (00:05:50):
Yes. I'm sure. I'm sure.

**Lenny Rachitsky** (00:05:51):
Okay, so for folks that are maybe a little bit familiar with Lovable or not at all familiar, what is Lovable? What's the simplest way to understand it?

**Anton Osika** (00:06:00):
I'd say Lovable is your personal AI software engineer. You describe an idea and then you get a fully working product from the AI. And what this means is that entrepreneurs actually today, they turn their ideas into real businesses. We have a lot of designers and product managers that create the first version of their product ideas to show to the teams, and some of them become founders because of the empowerment from this, but also developers themselves, they actually writing code or creating products much faster. The reason, it's pretty obvious for me, but I'll spell it out, the reason why we're doing Lovable is that I don't know about your mom, but my mom doesn't write code and-

**Lenny Rachitsky** (00:06:52):
Same.

**Anton Osika** (00:06:55):
... almost all my friends throughout my life reached out for help. Like, "Anton, I need to build something. How do I find a great software engineer?" And we are building for this 99% of the population who don't write code. Currently, if you're technically inclined, you get much further, but over time, naturally the way to build software is by just talking to an AI. And that's how we see it.

**Lenny Rachitsky** (00:07:21):
I love the way that you guys describe it and you didn't mention it, but I think it's building the last piece of software ever. How do you phrase that?

**Anton Osika** (00:07:28):
Yeah, we say we're building the last piece of software.

**Lenny Rachitsky** (00:07:31):
The last piece of software. Okay, we're going to do a live demo, but first of all, can you just share some stats on the scale of this business at this point because it's quite absurd.

**Anton Osika** (00:07:42):
Yeah. So we launched Lovable less than three months ago, and now we have 300,000 monthly active users and 30,000 of those are actually paying and it is growing at the same rates, just almost only through organic word of mouth.

**Lenny Rachitsky** (00:08:02):
Okay. And I'll share a couple stats in terms of revenue, just so folks know this, and we'll have this in the intro too. I think you guys hit 4 million ARR in the first four weeks. You hit 10 million ARR in the first two months with just 15 people. You're the fastest growing startup in all of Europe, and you guys had to rewrite your entire code base recently and you couldn't ship any new features for a while. Is that right?

**Anton Osika** (00:08:26):
That's right, yeah. People were saying like, "Oh, you're shipping so fast." And we were all quite frustrated because we wrote our service in this kind of scripting language and then as we started scaling, we were just, no, we have to throw everything away and rewrite it in a more performant way.

**Lenny Rachitsky** (00:08:43):
Okay, before we get to the demo, last question, you shared there's some companies that have started based on Lovable. I didn't even know that. So what are some examples of companies slash businesses that have launched off of Lovable and now are actually companies?

**Anton Osika** (00:08:56):
I mentioned designers using Lovable and one of our early users, Harry, he started shipping real web apps to his clients instead of just shipping designs. And then he went on to say, okay, wait, I'm going to start an AI startup. And his company, he launched on Product Hunt and everything and making money is just like, let's anyone upload their photo library and then the AI parses and categorizes it. And if you go to launched.lovable.app, this is an app built with Lovable is again a product Hunt version where you can see a lot of businesses or small SaaS featured there.

**Lenny Rachitsky** (00:09:38):
Okay, cool. So we're going to come back to some of this stuff, but let's get into demo. I rarely do demos on this podcast, but I'm finding that I think it's really important for people to see these products in action because in a large part, this is the future of product building and a lot of people hear about, "Oh yeah, AI's coming," and I don't think a lot of people actually see what the latest tools are capable of. And so I love showing these sorts of things on this podcast.

**Anton Osika** (00:10:05):
So Lenny, I was thinking, did you ever consider making a copy and build your own Airbnb?

**Lenny Rachitsky** (00:10:14):
I haven't, but go on.

**Anton Osika** (00:10:17):
How about you do that?

**Lenny Rachitsky** (00:10:18):
Let's do it. Let's do it. Okay, so we're going to make our own Airbnb.

**Anton Osika** (00:10:18):
Okay.

**Lenny Rachitsky** (00:10:21):
Okay, cool.

**Anton Osika** (00:10:22):
So I just put in the first prompt for an Airbnb clone.

**Lenny Rachitsky** (00:10:27):
Okay. And what is the prompt, just for folks that aren't watching?

**Anton Osika** (00:10:31):
Two words, Airbnb clones. That's the prompt.

**Lenny Rachitsky** (00:10:33):
Okay.

**Anton Osika** (00:10:34):
Just start simple and then what you get is that the AI says, okay, I can go through what does a beautiful Airbnb clone look like and it goes through a bit of design decisions and then I'll zoom out to see more of it. We have this just UI that is... I mean it has all the nice things you would expect from Airbnb clone where you see different categories and you can see two listings from Airbnb with login buttons and everything. So far it doesn't have the functionality of Airbnb, it just has the UI. I would now ask for an improvement on some of the functionality. Like if I'm switching category, I want to see different listings, let's say. But if you have any thoughts on what we should build next, let me know.

**Lenny Rachitsky** (00:11:25):
Okay, and so you had this preloaded, so you didn't see how long it would take, but how long would this normally take for it to just write all this code and have it for you?

**Anton Osika** (00:11:32):
The first prompt takes 30 seconds.

**Lenny Rachitsky** (00:11:34):
30 seconds? Okay. And it's like a very good copy of Airbnb. I love that you didn't have to show it a design, you just tell it Airbnb and it knows. Okay, so your question is would I want to add to my own version of Airbnb? I've always wanted to explore buying the place that I look at just like, Is this for sale? So what if we see what that would feel like if you're just a way to buy a listing.

**Anton Osika** (00:11:59):
Okay. Okay. So how about we add, I mean prompting is important here, so let's be specific, but we would ask, add a button on the listing which has purchased this Airbnb home. Is that it?

**Lenny Rachitsky** (00:12:16):
Perfect.

**Anton Osika** (00:12:19):
Okay, so, add, and [inaudible 00:12:19]. I'll be even more specific. It will pop up a model to purchase the listing.

**Lenny Rachitsky** (00:12:32):
Perfect. And I love... So I think something as you're typing, I'm just going to share thoughts as you're doing this. So the site that you asked this AI engineer to build, it's actually a functioning website that you can browse around. It's not just a design, obviously there's no actual listings here, there's no actual houses here. Say you were trying to actually build Airbnb and you wanted to start adding actual homes that plug into this, how does that sort of step work?

**Anton Osika** (00:13:02):
So as you say, this is just kind of the mockup UI, but it's also interactive. If I want to add login and add listing management, then we will connect something called the backend. So where data is stored, where user's log information is stored, and I can show you how to do that. First let's just try out where we got with this short prompt.

**Lenny Rachitsky** (00:13:29):
Let's do it.

**Anton Osika** (00:13:31):
Adding the purchase listing and it didn't do exactly what I wanted. I said, add a button... Or I didn't say what a button should say, but it says book now, and if I click book now I get a booking confirmation. So the AI was like, okay, it didn't... It was probably surprised by you wanting to buy the listing since it's Airbnb. So it still says book the listing, but it shows a pretty model where I can click confirm and pay. And then it's says booking confirmed.

**Lenny Rachitsky** (00:14:05):
I'll just say real quick, I love that this is actually a really good example of why being a good product manager is important. A lot of wasted time happens when you're not clear about the problem you're trying to solve and why you're trying to solve it and all that kind of stuff. So it's really cool that this is a use case where you have to be really good at explaining what it is you want. And it's interesting, you don't have to tell this AI-why. Humans want to understand, "Why is this important." Mostly you need to be very clear about what it is you're doing and I love that's a really strong PM skill. Your PM's really good at that. So we have to...

**Anton Osika** (00:14:39):
Hey. Explaining exactly what you expect and what you're not getting is even more important with AI than with the humans. So I'm going into hooking up more of the actual functionality, but first I'll actually show you something. What's the fastest way to change what went wrong, it's created buttons that say book now and I want them to say, "Buy now." And what I could do is select this item and say change it to buy now. But what we just realized is that you can actually edit this, this is a fully functioning product, but you can edit it visually like you do in Squarespace and Wix and so on. So I'll just change the text to buy now and then it instantly changes. It actually changes deep down in the code base, but it's very fast to do that.

**Lenny Rachitsky** (00:15:35):
So I think people listening to this and seeing this, if you're not aware this is the cutting edge of tools like this, no other tool out there lets you generate code from an AI engineer and then actually just change a small element of it of every other tool that I'm aware of. You have to ask the agent, do this for me, and then you hope that it does the right thing. So this is a huge deal which you just showed. Right?

**Anton Osika** (00:16:00):
Yeah. Now it says buy now.

**Lenny Rachitsky** (00:16:01):
Okay. Like that's amazing. Okay, and that's something you just launched?

**Anton Osika** (00:16:04):
Yeah. Great. We just launched this a few days ago, but I won't go into for building the full functionality, but what it looks like is that you connect an open source backend as a service and that's called SuperBase. And I have this instance to connect to that's completely empty, just like one click to set that up and now it's connected to the backend. It's just automatically generating some code and explaining what I can do next. And what I would do now is, say, let's add login, let's say let's add login.

**Lenny Rachitsky** (00:16:41):
And where is it actually hosted on the backend and everything in general?

**Anton Osika** (00:16:45):
So everything can be one click deployed and then it's running. It's hosted by a cloud vendor, which is hosting, I think a huge chunk of the internet, it's called Cloudflare, and the backend is hosted by also good cloud writer, which is called SuperBase.

**Lenny Rachitsky** (00:17:07):
Amazing. Okay, let's wrap up the demo, that was... Unless there's anything else, was there anything else really important that you wanted to show?

**Anton Osika** (00:17:14):
No I'll just explain what I would do next. I would say, okay, let's add login. Let's make the listings editable by the users so users can upload listings and then this is going to take a bit more time, but with patience and good prompting skills, you're going to get to a full working Airbnb.

**Lenny Rachitsky** (00:17:33):
That was a really good piece to add. So basically this is getting to a place where it actually is not so different from actual Airbnb. People can log in, they can add their home, you can add internal tools to add listings for your, say, sales team, ops team. Basically it just will allow you to build a marketplace that looks a lot like Airbnb. Amazing. Okay, thank you for the demo. I think for a lot of people they're like, "Yeah, yeah, I've seen this kind of stuff," for most people, like, "Holy shit." It's unreal what... It's almost like we're taking for granted now. You can ask an app to build you a whole website and that costs probably like a few pennies. It took like five minutes versus it would've been tens of thousands and weeks and weeks and months to even build just a prototype.

**Anton Osika** (00:18:23):
I mean, these tools as we see here, they're already very good, it looks really good as well, but mainly I would say they're getting better very, very fast. And I'd say one of the bigger bottlenecks is now they're not integrated into the current way that you have your existing products and so on. But since they're getting better so fast, I think the best thing for people who are interested in this or interested in just being a part of the future economies, get your hands very dirty with these tools because being in the top 10% in using them is going to absolutely set you apart in the coming months and years.

**Lenny Rachitsky** (00:19:05):
So let me follow that thread. So say you are magically able to sit next to everybody that is using Lovable for the first time and you could just whisper a tip in their ear to be successful with Lovable, what would that tip be?

**Anton Osika** (00:19:20):
It takes a lot to master using tools like Lovable and being very curious and patient and we have something called chat mode where you can just ask to understand like, "How does this work? I'm not getting what I want here, am I missing something? What should I do?" Is the best way to be productive is also one of the best ways to just learn about how software engineering works, which is you don't have to write the code anymore, but it is useful to understand how software engin- or how building products works. So I think that's the patience and curiosity is super useful. The second part that we spoke about is that being, if I would sit next to you, I would probably say like, "Hey, you are not being super clear here." For example, don't say it doesn't work. Just explain exactly what you're expecting and which parts are working and which parts are not working. And that's something that a lot of people don't do naturally.

**Lenny Rachitsky** (00:20:25):
I love that when you have an engineer you're working with that does a very expensive mistake to miscommunicate something, to just forget about a feature, to forget a better requirement, and here it's... You do that and then 30 seconds later you're like, "Oh okay, sorry, that was wrong." And then you could just try again.

**Anton Osika** (00:20:41):
That's true. It might be more costly with humans.

**Lenny Rachitsky** (00:20:45):
Okay, and so the first step is chat mode. So your advice is chat with the... What do you call it? Do you call it an agent? What's the term for the thing that you were talking with?

**Anton Osika** (00:20:57):
Yeah, Lovable is an agent.

**Lenny Rachitsky** (00:20:59):
Just Lovable?

**Anton Osika** (00:20:59):
Yeah.

**Lenny Rachitsky** (00:21:00):
Okay. So you're talking about Lovable by the way. How decide on Lovable as the name? It's so sweet.

**Anton Osika** (00:21:06):
I think it's all about building a great product. That's what I want more people to be able to do and the best word for a great product is that it's Lovable. A lot of jargon that I like to use to emphasize what we should be striving for is building a minimum Lovable product and then building a Lovable product and then building an absolutely Lovable product. So I took that jargon with me in the company name.

**Lenny Rachitsky** (00:21:36):
That is great. Absolute Lovable product. ALP is the new MVP. Okay, so we talked about this, the scale you guys have hit at this point, I imagine it's far beyond 10 million ARR. Do you share that at this point or are you keeping that private?

**Anton Osika** (00:21:51):
We don't anchor on the numbers, but I could probably do a two X tweet about this quite soon. Yes.

**Lenny Rachitsky** (00:21:57):
Okay, so it's far beyond 10 million ARR at this point. It's one of the fastest growing startups in history, the fastest growing startup in Europe. I want to zoom us back to the beginning. What is the origin story of Lovable? How did it all begin? What was the journey to today?

**Anton Osika** (00:22:14):
I think I was not impressed by what people were doing with the large language models [inaudible 00:22:21], especially after I was using them way back. But when ChatGPT came out, they were starting to get really good at taking a human instruction and spitting out code and then people in my team, I was the CTO at a YC startup, they felt like, "Oh, Anton, you're exaggerating. This is not going to change anything in the coming years." So I wanted to prove a point and I created an open source tool called GPT Engineer where you write something like create a snake game and then it spits out a lot of code, a little of different files and then opens the snake game. And then I tweeted a video about that and GPT Engineer is to date the most popular open source tool to showcase the ability for large language models to create applications and it's at like 50 something thousand GitHub stars and dozens of academic references.

**Lenny Rachitsky** (00:23:21):
And I know that I'll just add that it GitHub shut you down because they thought it was some kind of attack, like how many stars you're getting, how many people were using it,

**Anton Osika** (00:23:29):
Right. Yeah. So that came later. That's with Lovable. So this is Lovable. Lovable, earlier was always creating new projects on GitHub when someone used Lovable and we asked them, "Is it fine? How was the limits here?" They said, "Oh, there are no limits." But once we started creating 15,000 projects per day, so there were a lot of usage. Then some engineer when was on call, maybe they woke up in the night and they saw their servers were taking too much load because of us. So then they shut off down completely and we got this email that said, "Oh, you broke some kind of rules and we didn't know what was going on."

**Lenny Rachitsky** (00:24:13):
That's similar to a story I heard when ChatGPT was originally being trained, Microsoft servers blocked it because they thought it was some crawler and it was just actually the very first version ChatGPT being trained on data. Anyway, keep going.

**Anton Osika** (00:24:29):
So I built this tool called GPT Engineer and I was thinking about we're seeing the biggest change humanity will ever see, I think, where before you had the manual labor being taken over by machines, but now it's actually cognitive labor being done better than humans by machines and what's the best way to have some kind of positive impact here? It's not to make engineers more productive, which there's a lot of companies using AI to make engineers more productive, Microsoft did with co-pilot and so on. But it is to enable those who have such a hard time finding people who are good at creating software that's been their absolute bottleneck and let them take their ideas and their beliefs into reality. So enabling more entrepreneurship and innovation by building the AI software engineer for anyone. And then I grabbed a previous colleague of mine who has also been a founder, Fabian, and I said we should build something like GPT Engineer but it has to be for the people who don't write code and that's the story.

**Lenny Rachitsky** (00:25:43):
Okay. And then that became lovable? There's the shift from open source into a product that anyone can use but also pay for. Makes sense. Okay, so from that point I saw a stat that you started making a million dollars in ARR per week and once you launched lovable, is that true?

**Anton Osika** (00:26:00):
Yeah, so launched, we actually called the first version of the product like GPT Engineer app and it was very different in some ways and we launched that under a waitlist and said like, Oh yeah, we have this waitlist and we got a lot of feedback and iterated. Finally, when we thought the product was really good we said okay, now we have a Lovable product. And it was mainly on the AI that we did a lot of improvements, once we launched that, that was 21st of November, so that's almost three months ago. We just hit 1 million ARR in a week and then it kept growing at that pace. It still growing at even faster than that pace.

**Lenny Rachitsky** (00:26:43):
Faster than 1 million ARR per week. Holy shit.

**Anton Osika** (00:26:48):
Yeah.

**Lenny Rachitsky** (00:26:48):
Okay, that sounds like product market fit to me. You said that you did a lot of work on the backend. I saw you tweet about this that you guys figured out some kind of unlock on scalability, like a new scaling law that allowed you to build something like this. What can you talk about there that on the technical element allowed you to build something new and the successful?

**Anton Osika** (00:27:08):
There are many scaling laws I would say when you build AI systems and this one in particular is about when you put in more work, the product reliably gets better and better. And what you've seen generally when you have AI building something is that it can get stuck in some place. It is super good in the beginning and then it gets stuck. What we did was to painstakingly identify places where it got stuck and there is different approaches but address different ways how we do it but address the places where it gets stuck, tune the entire system quantitatively and having a very fast feedback loop to improve it in the areas where it got stuck. The most important areas, it still does get stuck sometimes, but that's the scaling law and we're still early in that scaling law, I would say.

**Lenny Rachitsky** (00:28:04):
And so when you talk about things getting stuck, it's like the AI agent just saying, I don't know what to do from this point or they introduce some kind of bug. Is that an example of getting stuck?

**Anton Osika** (00:28:13):
Yeah. It introduces some kind of bug and then it's not smart enough to figure out how to get out of that bug.

**Lenny Rachitsky** (00:28:20):
I see. And this is a common problem people have with tools like this is they get to a certain point and then it's like, "Well I don't know what to do. I'm not an engineer, here's a bug it's running into or the infrastructure's built the wrong way." And so it sounds like one of the paths to solving that is what you're describing is you make the AI smarter to avoid more and more of these places they get stuck. Another is people just learning how to get AI unstuck. This is something when we had Amjad on the podcast from Replit, he said that this is the main skill that he thinks people need to learn is how to unstuck AI when it runs into a problem. Just thoughts there, I don't know anything along those lines come up as I say that.

**Anton Osika** (00:29:04):
This is something that is a problem today and the frontier of where this is a problem is very rapidly receding back. So what we did was we identified the most important areas, so specifically adding login, creating data persistence, adding payment with Stripe. Those are the things that we made sure it doesn't get stuck on, for example. And the places where it gets stuck today is currently something that you can use being very good at understanding and getting unstuck, but in the future it won't be so important. This experience just going to not get stuck.

**Lenny Rachitsky** (00:29:48):
And I know you're not talking super in-depth about this because this is one of your unfair advantages, this kind of stuff you figured out. So I'm not going to push too far. I know you want not everyone's into exactly the same stuff. So I want to zoom back to the pace of growth that you guys have seen. One of the big stories, everyone's always looking at you guys of like 15 people, 10 million ARR in two months. That's absurd. I don't know if it's ever been done in history. If so, it's maybe a couple other AI startups recently. How have you been able to do this? What have you done that has allowed you to grow this fast with so few people?

**Anton Osika** (00:30:24):
I'd like to take credit of having done everything end to end in the product, but we are building on top of taking the oil here, which we have discovered oil, which is are the foundation models and then what we've done is that we're obsessed about what's the right way to present this to a user. What's the interface for the human to get as much out of this as possible? Packaging together, I showed you in the demo how you can add authentication and making this work seamlessly together as a whole. That's what we've done. And then people love the product. That's the driver of the growth. For getting awareness, we've mainly been posting what we've shipped on social media, that's how people know about us.

**Lenny Rachitsky** (00:31:17):
So building in public is how people usually describe that. So I think it's like you guys have the advantage of the demos are just like, "Holy shit, you can do that." And then you guys share the numbers that you guys are growing at. So it's innately interesting and shareable, but I imagine most people have something interesting to share. I guess is there anything that you think you did that other companies maybe haven't done that make the product so lovable?

**Anton Osika** (00:31:43):
I mean the team is everything in building a great product, so I just give a big shout-out to the team that has written the code. I haven't written much of the code recently, I would say. You want people who can ship really fast and have good taste for what this simple, what's the right abstractions and I think that's what we've done differently and have this obsession for us making it better and better and better.

**Lenny Rachitsky** (00:32:17):
**Anton Osika** (00:33:32):
We have set up lovable so that we can change lovable with itself. We have done that. There is a lot of hyper-specific things in terms of running a separate... We spin up a dedicated computer for each user. It doesn't do everything. Lovable doesn't do everything. So we use the tools that are for developers, not for the 99%, most of the time. And everyone uses AI all the time in writing code. It's also in great course for experimentations.

**Lenny Rachitsky** (00:34:10):
And are there tools like Cursor and stuff like that? Like any tools you can share right now?

**Anton Osika** (00:34:14):
Yeah. I think Cursor is the one that almost everyone uses in the team.

**Lenny Rachitsky** (00:34:19):
Yeah. Okay, cool. I did a survey recently on tools that my listeners and readers use in cursor. 17% of all people that read my newsletter use Cursor already, which is absurd and you guys are in there, too. Okay, so kind of along these lines, there's obviously other competitors and companies in this space, so everyone's always wondering, you, Bolt, Replit, Cursor is a different kind of thing. What's the simplest way to understand maybe how Lovable might be different from say Bolt and Replit, which I think are probably the closest.

**Anton Osika** (00:34:49):
The packaging for non-technical people is what we aim for and I showed you in the demo that you can edit the text, you can change the colors and so on instantly without having to go into the code editor and without having to wait about 30 seconds for the AI do the full change. So that's the big way that we think about packaging it. And then for making sure that this can be used as productively as possible in a larger team. Something that's different from I think all the other tools is that it is synchronized with GitHub and that means that you can use Cursor, or the people in your team that want to be more low-level, they can use Cursor and while the people who don't want to mess and set up their local file system and commit to GitHub and so on, they can use Lovable's.

**Anton Osika** (00:35:48):
Not getting stuck is I think the most important thing for people. And that's why we entered this space late, we haven't done the same type of marketing as many others and we still, from the people that I talked to, ranked as the one that works most reliably.

**Lenny Rachitsky** (00:36:06):
I love it. Okay. So this point about how you can just use Lovable to build a lot of it for you and then get into Cursor to edit and tweak is a really big point. And you're saying other companies aren't as good at that. I don't know if any other does that.

**Anton Osika** (00:36:06):
Yeah.

**Lenny Rachitsky** (00:36:23):
I don't think they let you do that. Amazing. Okay. And then what's the vision for Lovable? What's the end state of this? Is this everybody can build anything they want sort of thing? What's the simplest way to understand where you're going in the next, I don't know, five, 10 years.

**Anton Osika** (00:36:37):
I have to say. So we're building the last piece of software and it is inherently very hard to predict how the world looks like in five years. These days it's very hard. But the last piece of software, how I see that is that it's almost instant to go from what you want to change in the product or what product you want to build to having it fully working end-to-end, integrated with any of your existing systems or integrated with the very powerful third-party providers. Already today you can just ask, add and chat with OpenAI and then you get the chat with OpenAI in your product. But that's just working perfectly is something that's coming in the coming two years, I would say. And then after that there is a lot of things in building a product that is not just the engineering side, right? And I think an AI can be very useful in aggregating and understanding your users.

**Anton Osika** (00:37:42):
So, if you use the analytics tools, you know that there is something quite common which is to see how users have interacted with the product. AIs can do that on an absolutely massive scale and propose changes to a human to say, "Oh, yeah, that sounds like a good change to make it a bit more intuitive." And it can also automatically run A-B tests so that you can see the data, all these improvements to the product. I think that's on the horizon as well, quite.

**Lenny Rachitsky** (00:38:15):
What's interesting about this in one way is people wonder just what jobs will be more important, what skills will be less important? Let me share a thought I have and then I want to get your take and see where you go with this. It feels like what is getting more valuable is being good at figuring out what to build and then knowing if the thing you had built is correct and good and ready. So it's like discovery, ideation, idea, part of the step of launching a product and then it's like taste and craft. Just like is this the thing? Is this going to solve people's problems because the building now is being done more and more and it's interesting, it used to be the reverse engineering was the hardest, most valuable skill and now it's figuring out what to build.

**Lenny Rachitsky** (00:38:59):
You could sit there and you could just tell it what to build and a lot of people get to your screen I'm sure and they're like, "I don't know what to build, I don't know what people want." And it's like that's the thing now. So just reactions to that and thoughts on what's skills will matter more and less.

**Anton Osika** (00:39:13):
I mean if you're a founder or you want to build something. Yeah, I totally agree that figuring out what are pain points and seeing there are often currently solutions, some kind of solution to everything. How can you make this 10X better somehow figuring that out is super important when you have an existing product. Then I think taste and tasting what is good is even more of the important part. The engineer skill set is still going to be important because that helps you understand what are the constraints, so what you can build and I just think a lot of software engineers are probably a bit scared now like, "Okay, am I out of a job? What's going to happen?" But they should see themselves as the people who translate the problems that are stated a human, probably, to technical solutions, but they do have to abstract themselves up a few steps, not just looking at in their tech stack like oh I can just do the front end changes. Engineers or technical people are very good at understanding what are the constraints technically and they should see themselves as that translators.

**Lenny Rachitsky** (00:40:30):
Is it almost like you want to learn the eng manager skill of overseeing engineers versus the actual engineering skill or you think it's still going to be really important to learn how to code and be really good at that?

**Anton Osika** (00:40:44):
I mean doing a bit of everything. Being a generalist is I think much more important than it used to be. And if I'm putting together a product team today, I would re-obsess about getting as many skill sets as possible for each person I hire. They should know how architecting a system works, preferably they should know the sign, they should have product taste, they should know how to talk to users. I think everyone should know a bit of all of that, preferably.

**Lenny Rachitsky** (00:41:17):
Easier said than done. It's hard to find people that know all these things. So let's segue to hiring and how you hire. How many people do you have at this point? Is that something you share?

**Anton Osika** (00:41:27):
Yeah, now we're at 18.

**Lenny Rachitsky** (00:41:29):
18. Okay. Wow. I love that you... It sounded like you're about to say, "Oh, we have a hundred people now." No, 18. Okay, so you went from 15 to 18. Okay, great. So what do you look for when you're hiring people? The way I saw you describe it on Twitter is you look for cracked engineers, the best crack team in Europe, things like that. I guess just specifically what are you looking for when you're hiring?

**Anton Osika** (00:41:52):
I think the most important thing is that people care a lot and they're not just like, "Oh, I'm here for a job. I'm here for being just a passenger on this journey," but everyone should really care about the product, the users and care a ton about the team, how the team works together and that you're always contributing to making the team work more productively together and that care or preferably obsession gets you a very long way and then you do often want to have absolute superpower in some dimension to be able to understand and do as many possible things as possible, have this generalist brain that quickly learns any skill but be super, super good in one dimension. And for us that's mostly cramming as much out of AI, out of the large language models and understanding the entire parameter space of what you can change to make our product perform better.

**Lenny Rachitsky** (00:42:58):
So, how do you actually test for these things? Some of these things describe, I think everyone's looking for, they care about the user, they want to collaborate well. Because you have 18 people building in a company that's growing more than a million ARR every week. That's an absurd scale and the people you have found are clearly world-class and I think a lot of people are going to want to hire the type of people you're hiring. So when you're actually interviewing, how do you suss out some of these things like their AI cramming skills, their team building collaboration, what do you actually do?

**Anton Osika** (00:43:32):
I ask people what they've done before and these people that I'm describing, they have often done something where they care a lot about what they've done before and dig into details about the technical things that they did. And then we do the normal thing of showing a very hard problem that is a bit unorthodox that someone hasn't seen before preferably and see how they think through thinking research through that. Then something that I think is more uncommon is that we do, I pretty much always have people join the work simulation for at least a day, often a full week.

**Lenny Rachitsky** (00:44:13):
Awesome. Okay, so work trial. That's awesome. So basically they work with the team for at least a day. You said sometimes a week, and I love this point you made about they care deeply about something they previously worked on and you look for just obsession with the thing that they built last or something they worked on. What percentage are engineers of these 18?

**Anton Osika** (00:44:39):
So 12 at least write code at least part-time.

**Lenny Rachitsky** (00:44:44):
12 out of 18. Okay, cool. When we were setting up, you're like, "Oh, our engineer's creating content now." I think that's a cool example of how people do a lot of different things. Also. Okay, so I have your job posting that you shared once of the actual job description. I'm going to read a few lines from it. It's very inspired by Shackleton, right?

**Anton Osika** (00:45:07):
Yeah.

**Lenny Rachitsky** (00:45:07):
Would you agree? Cool. I love it. By the way, did you write this or did you have AI write this job description where you create an engineering job description? In fact, let me read it to you. I don't even know, you may not know what M you're referring to. I'll read a few lines here. "Long hours, high pace, candidates must thrive under a high urgency under AGI timelines approaching, difficult mission ahead, honor and recognition in case of success, those seeking comfortable work need not apply." And then there's a few other things, "Collaboration with other exceptional minds, purpose larger than any normal engineering role, generous share in the venture success." Amazing.

**Anton Osika** (00:45:44):
Thank you.

**Lenny Rachitsky** (00:45:44):
Thoughts?

**Anton Osika** (00:45:45):
Yeah, so I did get some help with the formatting of this, but then it was mostly me doing the exact phrasing of the different sentences.

**Lenny Rachitsky** (00:45:56):
So good. And I love that to some people it's going to be like, "Holy shit, I'm not signing up for this." But to a lot of people, the people you want is like, "Yes, this is exactly what I want to be doing."

**Anton Osika** (00:46:07):
Great.

**Lenny Rachitsky** (00:46:07):
Amazing.

**Anton Osika** (00:46:08):
Yeah.

**Lenny Rachitsky** (00:46:08):
Okay, cool. So it feels like one of the elements of hiring here is, create a really good filter to be clear about just how intense this is so that the people that want that are the ones drawn to you. Okay. And then you're also, you're in Sweden, fastest growing startup in Europe ever thoughts on building in Europe slash Sweden versus the US slash San Francisco?

**Anton Osika** (00:46:34):
Yeah, so this ambition level that you're talking about in the job ad is more uncommon in Sweden and I think that is the biggest unlock that someone like me, sees that this is the time in human history when you have the most impact for a worked hour and that's why we have to be super ambitious, just up the ambition level and then we can maybe retire and have AI take care of most things in society and inspiring people to be this ambitious in a place where the average ambition is lower but the talent, the raw talent is much more available, is a great recipe. I think that's a great recipe. And that's, I think it's some kind of advantage there. It is a bit of a double-edged sword but it's some kind of advantage.

**Lenny Rachitsky** (00:47:34):
So what I'm hearing is there's incredible people in Europe, they're just not, they're harder to find and what I'm hearing is the key is how do you suss them out and get them to want to talk to you?

**Anton Osika** (00:47:49):
Most people in Europe, they haven't thought that, "Oh, going on an extremely ambitious mission is what I want to do." So that's figuring out who those are is a big part of it.

**Lenny Rachitsky** (00:48:01):
Awesome. Okay. I want to talk about prioritization. I imagine all these things that I just shared about just how ambitious this mission is, how much you're doing the last piece of software, you must have a bazillion things that people ask you to build that you want to build. What's your approach to deciding what to prioritize and actually build?

**Anton Osika** (00:48:21):
Just top line? I think identifying what is the biggest bottleneck, what's the biggest problem and iterating fast on saying, "Okay, this is the biggest problem, let's really, really solve that problem." And then picking in the next one and not overthinking, not dreaming out the long roadmap, that's my [inaudible 00:48:41]. There's a very, very simple algorithm. Understanding what is the, mostly the biggest problem is not always a simple problem I think. Yeah, so we spend time as one should on talking to users, reading up on what people are writing. We have the feature board for where people do a lot of requests, as you say. And then when we pick one of the problems, we are quite engineering-led. For a product like ours, it's hard to be have product managers that are not engineers say, oh, this is what we should do now because the right solution to the problem might be entangled in things that are technical details.

**Anton Osika** (00:49:32):
They might be entangled in technical details of like, "Okay, yes, this is the biggest problem, but we should have this larger technical initiative that's going to solve all of these problems." So it's quite engineering-led compared to many other product companies.

**Lenny Rachitsky** (00:49:48):
As it should. I'd be worried if you guys had a product manager at this point, that wouldn't make no sense right now. I imagine the answer is it's chaos and there's no actual defined process, but just what does it look like generally? What's kind of the cadence you guys operate on? How do you take an idea to build it, spec it, launch it? Just what does that look like if you have something?

**Anton Osika** (00:50:10):
If you look back three months, we mainly said, "Okay, let's do this weekly planning." We do have a big jam board where we have all the main problems and then we have kind of ranked them which else do we focus or when we focus on next or this week? And then we have a demo where we say, "Okay, these are things we ship this week." So to get everyone on the same page, we do have a bit more of a roadmap now, where we say we are going to make so sure you can support custom domains. Next, they're going to add collaboration after that. And the biggest problem now or the biggest initiative now that solves the biggest problem is making the system more agentic and that has a bit of a longer roadmap, but we still do the cadence of weekly planning. These are the things we're focusing on. This week, it's mostly... There's a good word for this that I would want your help with, but polish, we were fixing the bugs and polish this week and that was the planning on Monday.

**Lenny Rachitsky** (00:51:21):
That was actually this week was polish, polish week. I love that. How far is this roadmap that you are now having?

**Anton Osika** (00:51:28):
I mean it's clear over the coming month, but it stretches out three months, but in one month it's probably going to look a bit different.

**Lenny Rachitsky** (00:51:39):
Okay. And then what are the tools you use just for folks that want to understand the latest tools? So you said FigJam, what else is in that stack of tools?

**Anton Osika** (00:51:46):
I mean we do so many things in our company in Linear because it's just amazing product. So we do talent application tracking in Linear and after going through and this thing, lot of the other custom-made tools for that Linear and then FigJam.

**Lenny Rachitsky** (00:52:06):
So simple. How soon until one of your engineers is an agent engineer, an AI Engineer, do you have a sense?

**Anton Osika** (00:52:15):
I love to dig into what does that question actually mean? I think we've been talking about, Oh, AI that would require something playing chess, that's AI. If a computer can play chess, that's AI and now that's like, Oh no, that's a chess program and which always shifting this forward and forward. I think anything that a human doesn't do is just a smart computer system, right? When is an software engineer and agent, I think it's always going to be just we're building in... Lovable is just an interface that humans interact with to create the software that they want and then how we solve that, we said going to be an agent under some definition. Yeah, sure. I think so, but that's less important to me.

**Lenny Rachitsky** (00:53:15):
Okay, I like that. Let me ask this, you guys are moving super fast, scaling like crazy. You described a little bit about your process, weekly planning, FigJam board of ideas and now there's a roadmap that you're kind of thinking out in the future. Is there anything else that you found helps you move this fast that gives you a lot of leverage over the small team you have to ship quickly and move fast that you haven't already mentioned?

**Anton Osika** (00:53:40):
We work from the office most of the time. I think it's pretty nice. Then you can say like, "Hey, I think we're thinking wrong about this thing," or, "Shouldn't we actually do this other thing?" And especially I think lunch, eating lunch together is a pretty productive hour where you're cross pollinating. I mean people are constantly thinking subconsciously as well about how to solve these different problems and which the most important ones are. And then being in office has this focus or most of the time usually focus, but you also have this high bandwidth where everyone has to be down structured communication.

**Lenny Rachitsky** (00:54:18):
I love that. The answer to the CEO of a company that's one of the most advanced AI tools in the world is one of your answers to how to move fast is lunch together. I love that. That's so human and so it makes all the sense in the world, but I love that that's still a part of this.

**Anton Osika** (00:54:34):
Yeah.

**Lenny Rachitsky** (00:54:36):
Okay. You talked about this kind of on the same thread you talked about if you were to start a team, like a new product team today, say you were head of product somewhere or head of RPM, VP of product somewhere building a new product team, scaling a product team, what would you do going forward that's different from what people have done in the past in terms of who you're hiring, how you're structuring them, that kind of thing? Just what do you think people should be thinking as they build product teams going forward? Knowing tools like Lovable exist and all the other stuff that's going on.

**Anton Osika** (00:55:13):
I mean everyone should be excited about using AI. I think that's a pretty big ones. And then the team working really well together is, like the lunch, you have to sit down and solve problems together. The bottleneck for most products these days is not going to be as much on engineering, but having good taste, good intuition about your users. And that, engineers and everyone preferably in the team should have that willingness at least to want to go through that motion and listen to the users and truly understand what they care about.

**Lenny Rachitsky** (00:55:59):
Well it's kind of like the background of most of the engineers and people you've hired. Is there anything in common? Are they just super impressive humans generally, like champions of programming contests, stuff like that? I don't know. What are some attributes of the folks you've hired so far?

**Anton Osika** (00:56:19):
I think raw cognitive capability is the strongest, the strongest correlate of being at Lovable. But there is this startup mindset that I think is also very strong. Being much more interested in moving very fast and iterating fast, then having a lot of structure, a lot of process and thinking about the business as a whole. More than thinking about my specific profession, my specific craft that I'm seeing myself wanting to dig into on me.

**Lenny Rachitsky** (00:56:58):
Amazing. Okay. So smart, very smart entrepreneurial, acts like an owner, isn't just like, this isn't just a job. But they feel like they actually have agency. Okay, this is great. There's something you said kind of along these lines that I think is important that one of the things that gets you excited about what you're building is giving people super powers and especially people that don't add a code, basically 99% of people. Is there anything along those lines that you think is important to share?

**Anton Osika** (00:57:27):
It's very clear to most people who have been engineers or been founders, that there's so many that have failed in their endeavors because they didn't have someone that know how to solve the technical parts. And now that we're close to having people know that this exists and they solve everything, it's going to be an Cambrian explosion of entrepreneurship and better software product. We're not going to settle for all the annoying bad technology that we use today. And everyone who has an idea is going to say, "Okay, I'm going to build this thing and show you that this is the best version of the product or what our company should be doing," instead of having long meetings or writing up documents. So it's going to be empowering across a lot of different professions and places in the world.

**Lenny Rachitsky** (00:58:33):
What's next for Lovable? What's the next few things they might launch as this episode comes out?

**Anton Osika** (00:58:38):
As I mentioned this agentic behavior, and when I say agentic, what it means is that you give more freedom to the system to decide what happens next. It might want to write a test, run those tests and say like, "Oh, the test failed, let's fix those." So that's one of the big unlocks for getting further faster. And then there's some more obvious things that you want to do to go all the way to easily go all the way to making money with Lovable. And that's like how do you set up so that it's hosted on your specific domain? How do you collaborate there seamlessly with your team and making that is here so that those are just obvious things and something we're thinking about is to help founders succeed after they built their first version. And how do they get more users? How do they get feedback? How do they get the word out if they build something useful?

**Lenny Rachitsky** (00:59:42):
I was just going to say that and that's exactly where my mind went is everyone's going to be building all these things. No one's ever going to get any traction with these tools. No one knows how to find users, get anyone to basically go to market. And growth is a whole different skill. So that is so cool that you're thinking about that. How do we run some paid ads for you? How do we think about SEO? How do we think about word of mouth, reality referrals? That is very cool. Okay.

**Anton Osika** (01:00:06):
Yeah, we already have playbooks that we help the people building with how do you do those things that you can find up on our blog?

**Lenny Rachitsky** (01:00:15):
Interestingly, this makes me want to buy some meta stock because all these apps that everyone's building, they're going to be running paid ads on Facebook and Google. Oh my god, what a good business those other guys get. I want to come back to, you said that you can work on your existing code base. This is actually a big question for a lot of people. They see all these tools, they're all amazing for prototypes and concepting. You talked about how you can actually do this within your existing code base, use Lovable.

**Anton Osika** (01:00:41):
Let me correct you there. You cannot use it on any existing code base.

**Lenny Rachitsky** (01:00:46):
Got it.

**Anton Osika** (01:00:47):
We kind of have a research preview of importing your code base, but what you can do is if you start in Lovable, then you can have engineers editing it in whatever tool they want to use for editing it.

**Lenny Rachitsky** (01:00:59):
Okay, cool. That's great clarification. So I guess just for people, because most listeners here are not building something brand new, they're working within an existing product. So you're saying that that is coming, you can use Lovable in the future in some form with your existing app and product?

**Anton Osika** (01:01:15):
Correct.

**Lenny Rachitsky** (01:01:16):
Wow, that's huge. Okay. Because basically most people, so that's going to be a big deal. Okay. Final question. We have the segment on this podcast called Failure Corner, where most people come to this podcast, they show all these stories of success and everything's going great, and here's all the things always winning. You guys, this is a good example, just up into the right, the fastest growing product ever. What's an example when something totally failed in the course of your career and what did you learn from that?

**Anton Osika** (01:01:49):
I am a bit hard-pressed to find something that totally failed, but I think there's a bit of a product lesson where I was the first employee at an AI startup here in Stockholm called Summer Labs, and the premise was just, okay, so humans learn in different ways. If you personalize, then you get two standard deviations more effective learning. So there are a lot of products like education software that helps you learn that is not personalized. And we were building an API to personalize learning and the AI and so on, it was pretty good.

**Anton Osika** (01:02:34):
But the thing that we were doing in the end was to say like, Okay, here's this product. Someone has to build a product or some way to learn or be it like English thing Duolingo, and then the people that have that product have to use this advanced AI API to start making it personalized. And it is very hard retrofitting like, oh, you have to switch out the engine and put in this AI. And the big learning here is that it didn't work very well for the company. I mean, the company wasn't super successful in this. The big learning is that you have to start with how is this product working end-to-end and then add AI or think where should we add AI? So that was a big learning for me that you really want to see what does the big picture of the user, what's the big picture of how do you think the user experience should be? And then add something with AI to solve specific problems. And now Summer Labs is doing great, but it's not on top of that product specifically.

**Lenny Rachitsky** (01:03:49):
I think it's a lot of people hear this and they're like, of course, but I think it's so hard to actually remember this point when you have some cool tech and you're like, "Holy shit, everyone needs to try this. They're going to love it." And then you don't realize no one actually cares if it's not solving a problem for them. There's a lot of novelty products that everyone want to use for a little bit and then forget instant, I don't actually need this often. And so what this makes me think about is, there's all these product lessons for what is likely to help your product be successful. And an app like a tool like Lovable can help you do this because if someone is building something, you can guide them, Okay, what's the problem you're solving for somebody? How many people have this problem? How much does this matter to them?

**Anton Osika** (01:04:38):
Maybe we should add the Lenny mode. It activates in Lovable, it activates this product coach. That would be infinite questions, like, "No, no, wait, hold on, why are you doing this?"

**Lenny Rachitsky** (01:04:50):
Absolutely. Let's take a step back. Everyone's going to be like, "[inaudible 01:04:55], get out of my way."

**Anton Osika** (01:04:57):
[inaudible 01:04:57].

**Lenny Rachitsky** (01:04:56):
Yeah, exactly. What's your experiment plan? I think there's actually a big opportunity there to save people. There's a play around with this thing and then there's like, okay, but really is this anything people actually want?

**Anton Osika** (01:05:09):
I love it. Can we call it Lenny mode? Is that fine with you?

**Lenny Rachitsky** (01:05:12):
100%.

**Anton Osika** (01:05:13):
Awesome.

**Lenny Rachitsky** (01:05:14):
Let's do it. I'll license you no cost.

**Anton Osika** (01:05:16):
Sure.

**Lenny Rachitsky** (01:05:17):
Okay. Okay. We made a deal here. Let's do it. Okay, Anton, is there anything else that you wanted to share? Anything you want to leave listeners with before I let you go and go to sleep?

**Anton Osika** (01:05:28):
I think, again, the world is changing quickly and it's very fun. You should see that's like have fun in all of this change, and the best thing you can do for your current profession or if you want to have a new job is to be in the top 1% in knowing how to use AI tools. So go out there, use Loveable, use other AI tools, and become... Make sure to understand or try to understand as much as possible in how to use them productively. That's something I tell all my friends generally, and I love the audience to know as well.

**Lenny Rachitsky** (01:06:06):
Okay. Well, I got to try to make this even more specific for people. How do you know if you're in the top 1%? What's a heuristic almost slash how do you get there? Is it just use it a hundred times a day? What else? What can you recommend?

**Anton Osika** (01:06:19):
Yeah. I think if you spend a full week on trying to reach an outcome, the best way to learn is I want to do this thing and then I want to use AI to do that thing. And you've spent a full week, you are in the top 1% in the global population. And if you surround yourself with friends who have this obsession or they also care a lot about this, then you'd be quickly in the top 0.1%.

**Lenny Rachitsky** (01:06:47):
So what I'm hearing is find a problem that can be solved, find a problem, a pain point for yourself or someone, and then end-to-end fully solve that problem. Spend a week getting from idea to a thing that somebody's actually using and you're in the top 1%.

**Anton Osika** (01:07:03):
Yeah. I think... At the top, yeah, the top 1% by just spending a full week and asking AI if you don't understand. So making sure that you understand.

**Lenny Rachitsky** (01:07:15):
Yeah, that's the thing people forget. You just ask. Would you ask the chat feature of Lovable in this case or would you go to Cloud or ChatGPT to ask for advice?

**Anton Osika** (01:07:24):
I mean, my recommendation here, if you're in product is to use Lovable to build software and learn that AI tool and then you should use ChatMode and ChatMode, I have to add, is something you activate in your user profile. It's not launched in the main product, so it's in labs, but if you add that flag, then you can use ChatMode. If you want to learn some other AI tool, then you should ask that tool or ask Cloud, ChatGPT about how that topic, that domain works.

**Lenny Rachitsky** (01:08:02):
Okay, amazing. Where can people find you? Where can they find Lovable and how can listeners be useful to you?

**Anton Osika** (01:08:09):
Lovable posts updates, and memes on Lovable underscore dev on Twitter, we post things on LinkedIn as well, and there are a lot of things coming out and changing in how we build software, so you can follow Lovable underscore dev and you can follow me at AntonOsika at Twitter. I'd love more feedback on where people see this is a huge change for them. There are a lot of people posting about that on Twitter, but we have a Discord where you can share like, "Oh, this is how I use Lovable. It was super useful to me." And feedback.lovable.dev can ask for new features. There's a lot of people asking and uploading what features you want next. And that's super useful. That's the most important thing for us. We just want to solve people's problems.

**Lenny Rachitsky** (01:09:04):
Amazing. Anton, you're doing incredible work. What a journey. I'm excited to have you back someday when we see more chapters of this journey.

**Anton Osika** (01:09:12):
I have a lot more to learn.

**Lenny Rachitsky** (01:09:13):
As do we all. That's why people listen to this podcast. Anton, thank you so much for being here.

**Anton Osika** (01:09:18):
Thank you so much, Lenny.

**Lenny Rachitsky** (01:09:19):
Bye everyone. Thank you so much for listening. If you found this valuable, you can subscribe to the show on Apple Podcasts, Spotify, or your favorite podcast app. Also, please consider giving us a rating or leaving a review as that really helps other listeners find the podcast. You can find all past episodes or learn more about the show at Lenny'sPodcast.com. See you in the next episode.

---

## [12/15] Inside Bolt: From near-death to ~$40m ARR in 5 months—one of the fastest-growing products in history | Eric Simons (founder and CEO of StackBlitz)

**Lenny Rachitsky** (00:00:00):
The rate you're growing is absurd. You're in this cohort of companies that are just growing at rates that we've never seen in the history of startups.

**Eric Simons** (00:00:05):
The company was on the verge of going under when we launched Bolt, and what ended up happening is, in the first two months it went from zero to 20 million of ARR. And we've already crossed 30 million of ARR, with the current rate we're on, our forecast for the year is we want to get to 100 million of ARR.

**Lenny Rachitsky** (00:00:22):
This is just non-stop wild shit. How is this possible? What has allowed you to grow this much, this fast, with such a small team?

**Eric Simons** (00:00:30):
Most importantly, it's been the people. It's rare to find startups where you have the core group of five, six, seven people that have been there for five years plus.

**Lenny Rachitsky** (00:00:38):
You basically were building a tech first, and then looking for a problem to solve later, which is often what people tell you not to do.

**Eric Simons** (00:00:44):
I think that's the hard thing about being an entrepreneur. There are periods of time where you have to make judgment calls that are not going to be the consensus view. You got to have confidence in your convictions on how to best play the hand.

**Lenny Rachitsky** (00:00:54):
A lot of people see these stats, and they sometimes don't see that there was also years and years of work before that.

**Eric Simons** (00:00:59):
It was kind of like, Bolt's this overnight success, seven years in the making.

**Lenny Rachitsky** (00:01:05):
Today my guest is Eric Simons. Eric is co-founder and CEO of StackBlitz, which makes a product called Bolt, which is currently neck and neck with Cursor for being the fastest growing product in history. They're currently the number one most popular web AI code app with over three million registered users. Two months after launching last October, they hit 20 million ARR. At the time of this recording, they're approaching 40 million ARR. The story of Bolt is wild. They actually started the company seven years ago, and were about to run out of money and shut down. But they realized the tech that they'd been building for the past seven years, called WebContainer, was perfectly suited for building AI products in a browser. So they launched the product with a tweet, and as Eric describes it, it was an overnight success seven years in the making. If you'd like to better understand the cutting edge of AI coding apps, and where things are going with AI and product building, this episode is a must listen.

**Lenny Rachitsky** (00:02:02):
If you enjoy this podcast, don't forget to subscribe and follow it in your favorite podcasting app or YouTube. Also, if you become a yearly subscriber of my newsletter, you now get a year free of Perplexity Pro, Notion, Linear Granola, and Superhuman. Check it out at lennysnewsletter.com.

**Eric Simons** (00:04:52):
Thank you for having me. Yeah, I'm stoked to be here.

**Lenny Rachitsky** (00:04:54):
For folks that are not super familiar with Bolt, what is Bolt?

**Eric Simons** (00:04:58):
It's really simple. You go there, there's a text box, and you tell it what you want to build. Whether it's a web or a mobile app, and so, it's kind of one of these text to app building tools that have become pretty popular over the past few months here. And it's not just building a static site, or something like that, but you can actually build full stack, real software with databases, and hosting and et cetera, just from prompting. And in a ridiculously short period of time, it's not like you're spending hours and hours or days, putting this together. You can get results in like, a minute.

**Lenny Rachitsky** (00:05:36):
Let's just share some numbers about the scale of what you're building. The rate you're growing is absurd. You're kind in this cohort of companies that are just growing at rates that we've never seen in the history of startups, and you guys are at the edge of that. Share some numbers about how things went when you launched, and where they're at today.

**Eric Simons** (00:05:53):
Yeah, when we launched, the company was on the verge of going under when we launched Bolt, our company StackBlitz. We'd been around for seven years building web-based development, environment stuff. And so when we launched this we were like, "This would be amazing if this added a 100K of ARR over the next couple of months." And what ended up happening is, in the first two months, we went from zero to 20 million of ARR. And I think we're on month four, or four and a half or something like that at this point, and we've already crossed 30 million of ARR, and we're on the verge of crossing 40. By the time this comes out, it appears that we're going to be at 40 million ARR. So it's just, the scale of the growth of the revenue has been nuts.

**Eric Simons** (00:06:41):
And of course, that correlates with insane user growth, as well. We've added three million registered users just in the past few months here, and monthly active users is around a million, I think at this point, per month. So it's just been... I've never seen anything, I've been doing startups for 15 years, I've never seen anything like this. Everyone I've talked to, our investors, or et cetera, there's not a lot of corollaries to what's going on here. And it's kind of extraordinary, because our company wasn't doing AI stuff six months ago. We had no AI products, and just out of nowhere we, from almost death of the company to being the number one buy traffic revenue, et cetera, like AI Cogen app, that's totally web-based, in the world. I think the only other startup ahead of us is for Cogen, just in general, would be Cursor, and the option revenue at this point.

**Eric Simons** (00:07:45):
And so anyways, it's been a heck of a ride. And our team's like 15, 20 people, so it's just dealing with, we're going to be closing on 100,000 customers, and our support team's like three people. So we're trying to scale as fast as we can. So it's just kind of mind-boggling, just the scale of the demand, and how we've had to turn things around to match the demand as best as we can.

**Lenny Rachitsky** (00:08:14):
Mind-boggling is an excellent way to describe what you just shared. A million monthly active users, you're at 40 million annual recurring revenue, five months into the business. Is that right?

**Eric Simons** (00:08:26):
Yeah, single digit. Yeah, single digit weeks. That's the current track rate that we're seeing for the thing. Yeah.

**Lenny Rachitsky** (00:08:33):
I think, are you guys are the fastest growing startup in history?

**Eric Simons** (00:08:37):
I mean, I think it depends on probably where you peg the number. Because yeah, we're here to just build great products, and just push the limits of what's possible with the technology. And I think that we do our jobs well, kind of crazy things can happen, but I mean, the current track rate we have, we're going to be exceeding the forecast for Q1 with the current rate we're on, and our forecast for the year is we want to get to 100 million of ARR. And now, I think there's been a couple, that would either be on par with Cursor, or ahead of them, or something like that.

**Eric Simons** (00:09:19):
And I think there's going to be more things like this, too. I don't think that... It just, there's something really, I think a lot of people are in disbelief about it too, where they're like, "This is, okay." And this is from when we were at, got to four million ARR, five million ARR in the first month. I would talk with people and they're like, "Okay, yeah, but that could go to zero." And then it went to 20 the next, "Ah, it could go to zero," but now we're closing on 40.

**Eric Simons** (00:09:40):
So from my view, I was also very skeptical, as this. I've never seen anything grow like this, right? And so part of me, for like a month I was kind of waking up waiting for the day where it just was like, "Okay, it's over." You know what I mean? This crazy thing happened, and now it's not. But that data just hasn't come. And you see this happening with Cursor, you see this with a lot of these other AI startups. And the value proposition is real. The free market is filled with rational actors. People are coming to these tools because it is solving problems, they're able to do way more for way less cost, than it would otherwise. And that's why I said, I think we're going to see more of this, whether it's in coding or other verticals, or whatever. In a sense, it's almost like maybe the new normal, as AI just continues to get better. But, anyways.

**Lenny Rachitsky** (00:10:41):
Let's get to a demo of Bolt, so people can actually see what this looks like in action. And as you go through it, if you can even point out stuff that is different from other products in the space, say Lovable, VZERO, Replit, that other folks have heard about, that'd be useful.

**Eric Simons** (00:10:57):
Awesome. Cool. Yeah, so this is Bolt, you just go to bolt.new. Things that I think are really interesting about Bolt. One is, it's just dead simple. Whether you're logged in or logged out, it's the same UI, it's extremely simple, it's just a text box. And I think that the biggest difference between Bolt and the other stuff out there, it's actually subtle. It's not like something you'd necessarily see in the UI, but it's how fast it is, and how reliable it is.

**Eric Simons** (00:11:23):
And this is because of how we are actually doing the compute, because what's going on here is when you type into, whether it's Bolt or another product, it has to spin up a dev environment to actually make that application. So there needs to be some operating system somewhere that's running it. Everyone else runs those things on cloud servers, which those can take minutes to boot up, and they often will run into issues, and then you can end up literally stuck and have to contact support to get it done, and get it unstuck.

**Eric Simons** (00:11:49):
With Bolt, and for the past seven years, what our company's been doing has been building an operating system that runs inside of your browser locally, using your CPU. So we have a very permissive free tier, and it's insanely fast, and it's insanely reliable.

**Eric Simons** (00:12:02):
So if I want to, just as a quick example of this say, "Make a clone of Spotify," and just hit enter. This thing's already getting to work, and already, on the right here, this is a full dev environment. This is an actual operating system, running inside of my browser. And I can run commands on it, et cetera. And really, what you're seeing down here, this terminal and kind of what's backing it, this is what took us really five, six, seven years to build, and make so reliable. There would not be a Bolt without this technology called WebContainer, that allows us to run an operating system in the browser.

**Eric Simons** (00:12:38):
Because what's going on here is, our AI agent for Bolt has bidirectional communication with this operating system. It's writing code, it's running the dev server for this thing, it's going to go ahead and spin this up. You can see how fast this is, in a matter of 60 seconds I said, "Make me a Spotify clone," and now we have one. And it looks pretty darn good.

**Lenny Rachitsky** (00:12:55):
That looks really good.

**Eric Simons** (00:12:58):
And that's one of the other aspects around Bolt is, this technology we made for the operating system side, the guys that have been working with us for the past five-plus years on it, before this they were actually doing machine learning AI stuff. And so when it came time to write the agent for Bolt, we had just an incredible amount of in-house expertise on how to actually merge these two different technology sets, to have this really reliable experience that produces really beautiful, really functional stuff. So that's based on what's really cool about the Bolt experience.

**Eric Simons** (00:13:32):
The other thing is, a lot of these products it's like, you can make something, but often you want to actually have a URL where you can share this. Having, maybe even attach a domain to it, or whatever have you. So with Bolt, we actually have built-in integrations with production grade hosting providers like Netlify, and for databases with Supabase.

**Eric Simons** (00:13:53):
So if I go and just click the deploy button here, this is actually going to run a production build of this project we made here. And again, this is doing this entirely inside of my browser, so it doesn't cost us anything to do this. So again, you can do this for free, and it has gone ahead and deployed this on a real URL, on Netlify. This is live, I can share this with anyone, and if I want to buy the domain spotifyclone.com, and point it at this, I can click this link here. That will kick me into Netlify, I can attach this to my account, buy a domain, point at that thing. And then from there on out, whenever I'm prompting Bolt to make changes to this application and hit deploy, that goes live on my public website.

**Eric Simons** (00:14:36):
So this is the simplest way to build a web app that's ever existed. That was one of the key realizations I had, a couple of weeks into the thing, I was seeing people use this for personal use cases. Like medical donation sites, or weddings, or whatever. And I was like, "Don't these people know that Wix or Squarespace exists? Should I tell them?" And then it hit me. Those things are so complicated to use. I don't know if you've ever seen just the UI of these things, but they're crazy complicated, and that's just for building a static website. There's no way you could actually build a functional app. And that's like, with Bolt, if we were to sit here for another 30 minutes, we would have streaming. You'd be able to make playlists of different MP3 files, or whatever. You can just keep prompting this thing to keep adding functionality.

**Eric Simons** (00:15:25):
So that's, I think, some of the cool core experience of both here. I can show you something cool that we just launched, if that would be of interest.

**Lenny Rachitsky** (00:15:32):
Let's do it.

**Eric Simons** (00:15:32):
So this is like web apps, right? Web apps are amazing, but often you want to have a native app. And it's hard to build web apps, it's even harder to build native apps, that can actually, that you can then go put in the app store. And so we partnered up with a company called Expo, and their entire business is making, basically, React Native tooling and this ecosystem that makes it super easy to build beautiful apps, and actually get them in the app store.

**Eric Simons** (00:15:57):
And so right here, I'll zoom in a little bit, we have this little, "Build a mobile app with Expo." So if you click that, we kind of instruct you on how to just prompt mobile apps into existence. So yeah, let's make another Spotify clone that's an actual native mobile app. Let's say, "Make a Spotify clone," go ahead and hit enter. And what this thing's going to do is actually, again, spin up a operating system here, where it's going to boot up the Expo tool chain and actually go and make a mobile app for us.

**Eric Simons** (00:16:28):
And what's cool about this is, we could actually preview it just in the browser here, but once this thing's done and it boots up, it's going to show a QR code, we're going to be able to scan it, and in real time actually basically have a test flight of this native application that we can try it on our phones, and as we keep prompting you'll see it making changes and stuff. This is kind of the first time that, you don't have to be technical to make production grade web, full-stack web and mobile apps. At this point I've done nothing that requires developer knowledge to do any of this stuff.

**Eric Simons** (00:17:10):
And I think that's what a lot of people are really excited about with this, and you know, the majority of our audience are people that are not developers, that are using this. They're PMs, they're designers, they're entrepreneurs. Because these are people that have always been great at building products, but previously, the only way that they could get their ideas into coded software was through a developer's fingertips. And now, they can deal with their own, through prompting.

**Eric Simons** (00:17:37):
So you can see here, we've got this little QR code. I'm going to go ahead and scan the thing.

**Lenny Rachitsky** (00:17:43):
I'm going to do it, too.

**Eric Simons** (00:17:43):
Cool.

**Lenny Rachitsky** (00:17:44):
By the way, I love that you had just enough things to say until it finished. That was pro.

**Eric Simons** (00:17:50):
Just as I planned, you know? So on my screen it's booting up, it's bundling the JavaScript of this thing, and it's beta. We just launched this last week, by the way. So if you can kind of see on my screen here, I actually have this Spotify looking app, right?

**Lenny Rachitsky** (00:18:07):
Wow.

**Eric Simons** (00:18:07):
That, you know.

**Lenny Rachitsky** (00:18:09):
That looks like, exactly like Spotify.

**Eric Simons** (00:18:11):
It looks exactly like Spotify, right?

**Lenny Rachitsky** (00:18:11):
It's good.

**Eric Simons** (00:18:11):
Yeah.

**Lenny Rachitsky** (00:18:14):
We're going to be sued right now, so let's be... You're doing too good a job with this. No, that's amazing.

**Eric Simons** (00:18:24):
Yeah. So it's pretty cool, right? And so what's cool is that, and as you keep prompting on your device, it'll just keep reloading. Without you having to kill the app, you can actually see the functionality getting added. And so, in this use case that you and I have right now, it's like if you and I were building an app together, we could be on other sides of the planet and you could actually be not just seeing a screenshot of the thing, but actually touching it and feeling it, and putting it through its paces.

**Eric Simons** (00:18:47):
And so a lot of product teams, I mean, this is just changing how people do product development. It's faster to do this than design a whole bunch of Figma frames, necessarily. Right? So.

**Lenny Rachitsky** (00:18:59):
We're going to spend a lot of time on that. Okay, this is incredible, this whole episode so far is you just blowing my mind and I imagine listeners' minds, just over and over and over. I don't even know where to go with all this, sometimes.

**Lenny Rachitsky** (00:19:09):
You made a really important point, that you worked on this for seven years before you launched Bolt. A lot of people see these stats, zero to 40 million ARR in five-ish months, and they sometimes don't see that there was also years and years of work before that. And the reason that you guys have been so successful is all the work you did that allowed, that built this WebContainer technology, it sounds like. Is there anything there that's worth sharing, you think, of just that part of the journey? I know we'll go through the origin that all, where Bolt came from, but I guess just that WebContainer component specifically. That feels like a huge deal.

**Eric Simons** (00:19:44):
A hundred percent it is, yeah. And I would say this is, surprisingly to me, it's still one of the contrarian viewpoints of our company. Because over the years it was like, when we first... And that, the WebContainer was the bet, that we made the company on. Just to be clear. StackBlitz was a browser-based, deep technology play on, "Can we make a web assembly based operating system that can boot in a browser, in like a hundred milliseconds, and run full on development tool chains?" That was really it.

**Eric Simons** (00:20:21):
And we'd gotten the idea for this, and the insight that this might be possible, because back when my co-founder and I came out to the Valley, he and I grew up down the street from each other in Chicago, we wrote code together at 13, and been building stuff ever since. And we came out to the Valley in 2012, and we just had the good fortune of bumping into Dylan Field and Evan Wallace when they were building Figma, in the early days. And that was, I don't think a lot of people know that Figma was also a browser-based deep technology play. Their first pitch for Figma, they didn't have a design tool. Their first pitch was this 3D ball dropping into water, inside of a browser town.

**Eric Simons** (00:21:00):
And the pitch basically was, "Browsers have this new capability called WebGL," the predecessor to WebAssembly, "and with these things, for the first time, you could actually create a graphics rendering engine, that you could then build a design tool on top of. But you're going to have to write that rendering engine from scratch, because nothing exists that can just compile into WebGL, or whatever. And if you want the performance you need, et cetera, it's going to take us years to do, but if we do it, we think this will change everything for design."

**Eric Simons** (00:21:35):
And obviously, we know how that story has panned out now. And back in 2017, 2016, 2017, Albert, my co-founder and I, saw the same sort of story begin to play out, but for web development and development environments. And specifically there was some stuff that landed in browsers like WebAssembly, shared memory, service workers, these different APIs. And we were like, "Oh, wow. It should be possible, theoretically, to write an operating system in WebAssembly that could run Node.js, and NPM and all the tool chains on top of it, that you need to do web development."

**Eric Simons** (00:22:12):
And that would be huge, because setting up developer environments, it's a pain for beginners. A lot of people churn out. The first thing you do when you learn how to code is not even learning how to code, it's how to set up your computer to even start writing the code. If you go join Netflix, or any of these other fan companies, the first month or two is you being onboarded, to run that stuff on your computer and set up your environment. And we're like, "If we could just have that be something, you click a link and it just boots in your browser, that'd be huge."

**Eric Simons** (00:22:42):
It's also, if you look at the other productivity apps that have really worked on the web, they've all had this compute model, right? Figma, when you open a Figma document, there's not like some cloud VM that gets spun up for you to render the documents. You're dragging things around. It's using your CPU and your memory to do the work. Same thing with Google Docs. That's the only model that's ever scaled to a billion users. And so, when you look at Cloud IDEs, like Cloud 9 was the first one, back in 2009 or so. The way these have always worked is that your browser's basically doing nothing, when you go to that. Every user that gets connected, there has to be a cloud VM that gets spun up for them, and then your browser's just taking your keystrokes, sending it to the server, and then sending back the results of it. And that's how all these other AI code, text to app sort of tools work. They're all using cloud VMs.

**Eric Simons** (00:23:34):
And the problem is, on a small scale it can work, but as you scale it up, I mean there's not even a 100 million VMs to rent, on the planet. But there are a billion devices that you can run this stuff on. Because that's kind of what we've seen with Bolt where, if you want to build a product that's going to be able to scale to that size, you have to look at all factors and go, "We have to build, make sure the technology provides the best experience, zero latency, transient cost." There's a permissive free tier, because the other problem with the server is, you end up, if you have a free tier, people are mining Bitcoin on it, they're DDoSing people using your servers. So inevitably, you have to nerf these things and roll them back. But if it's all done on the end device, it doesn't matter.

**Eric Simons** (00:24:18):
So anyways, WebContainer was the key piece, and what we struggled with, it took us four or five years or something, to build WebContainer. What we struggled with for the years after that was just how to build a product around it, because developers loved it, but they weren't using it in ways that they would pay money for. And as much as the nerd side of me wished that that would be enough, that it was like, building cool technology was enough. It's like, "It's not. We're here to build a venture scale company." And so that was kind of why we were high at the end of the journey, where it was like, we're taking shots on goal. And at some point, this got a connected bat, right?

**Lenny Rachitsky** (00:25:04):
There's a lot of really interesting lessons from this journey, that I think are counterintuitive. One is, you basically were building a tech first, and then looking for a problem to solve later. Which is often what people tell you not to do. And it worked out, in this case.

**Lenny Rachitsky** (00:25:19):
The other interesting takeaway here is, it feels like it's a similar moment to when AJAX came out and then everyone's just like, "Wow, you can build new things here." So it feels like there's a lesson here of just, "If there's a new technology that has enabled, something big that we think may, let's just work there for a while, and see if something comes up."

**Lenny Rachitsky** (00:25:35):
And then I think the other lesson here is just, as a founder, just survive as long as you can. Because you may find something that works.

**Eric Simons** (00:25:43):
All great points, all great points. Because you're dead right. And fortunately, my co-founder and I had, we had built a lot of unsuccessful stars before this. We spend most of the, or 20 times, churning through ideas on things. So when we had conviction, I was like, "This seems like a technology that will be important." It seems like, the web is the most ubiquitous... The pitch or the theory in our head was like, "The web is the most ubiquitous platform in the world, but yet it has no, you can't use the web to build the web."

**Eric Simons** (00:26:08):
Every other platform, Mac has Xcode. Windows has Visual Studio. The web had nothing. And we were like, "At a minimum, Google should probably buy this thing from us. It seems like it should probably be part of Chrome," at a minimum. And we thought, "Hey, this could be a huge enabler." The vision of just making it as easy to build full stack applications as using Canva, it just seemed really compelling.

**Eric Simons** (00:26:43):
But when you do that sort of risky deep technology play, you need to... And we were very good about this, like the previous company Albert and I did, we bootstrapped it all the way through to acquisition, so we understood and we were living hand-to-mouth, to bootstrap that thing. So we understood out of it how to have a low burn rate, and take a lot of shots on goal, and make every dollar stretch beyond what anyone would think is reasonable or possible. And that's how we played our hands with StackBlitz. We didn't raise money for the first two or three years of the company's life. We were bootstrapping it. When we did raise money, we barely spent it. Largely because it was like, "We need to just take a lot of smart bets, and it doesn't make sense."

**Eric Simons** (00:27:32):
And I would just say generally, until you see pull, just people pulling the product out of your hands, you don't want to be spending money. You should be like, default, no. And when you go and buy software, you should be going, "We're a tiny startup. Can you sell it for half?" Everything you buy, just keep the burn rate as low as possible, because you need as many shots on goal as you can possibly get. Because you have no idea. I think just generally, for startups, that's the right way in my view, to approach it. Unless you're seeing, again, immediate demand and pull, or whatever.

**Eric Simons** (00:28:02):
But yeah, I think that'd be, maybe the extra context I'd add on top is, I think that we ended up doing a good job of being extremely conservative. During a time in which, during 2020, through 2020 and 2021, which were times where exuberance and growing headcount was like, KPIs of companies. And were things that were being... With lot of emotional force of like, "Hey, you guys ought to be doing this." And I'm glad that we didn't heed the advice, because if we had tripled the company and kicked up the burn rate, there would be no Bolt. We would've gone out of business a lot of time ago.

**Eric Simons** (00:28:48):
So I think that's the hard thing about being an entrepreneur, I think is you kind of have to... There are periods of time where you have to make judgment calls that are not going to be the consensus view. Maybe years later, it'll become the consensus view, but you got to have confidence in your convictions on how to best play the hand.

**Lenny Rachitsky** (00:29:15):
There's so many great lessons here, I think just this idea of just staying alive. Dalton came on the podcast, he's a partner at YC Ones, and he just had this phrase, "Just don't die." And that's exactly what you guys did, seven years of just trying it until something worked, and I love that you actually were planning to shut down the company right before you launched Bolt. And I know you launched it with just a tweet, right? That was the launch moment?

**Eric Simons** (00:29:35):
Yep. Yeah.

**Lenny Rachitsky** (00:29:37):
Maybe talk about that moment of just, after launch, signs that, "Okay, this is working. Something's different."

**Eric Simons** (00:29:43):
Yeah, yeah. So day one it was like, there's great reception to the tweet. We were like, "Wow, this is one of the biggest things, launch day reception we've ever seen." And I think on the first day, I think we added 60K of ARR, or something. Which was like, I mean, crazy. Again, we were at 600, so we added 10% in a day. And I remember our dev ops engineer, he was the one who would flag me. He was like, "Guys, we got 60K today. This is crazy." And I was like, "Yeah, yeah. But this is launch day."

**Eric Simons** (00:30:12):
There's the tech crunch, peak of initiation, in the classic startup-

**Lenny Rachitsky** (00:30:17):
**[Inaudible 00**:
30:17]-

**Eric Simons** (00:30:17):
... star, yeah. I was like, "Listen, guys." I'm trying to temper enthusiasm for the team. I'm like, "This is great. Got a lot of work to do." And then the next day we added 80K, or whatever it was, and it just kind of kept going. And all the while, the product we put out, we built a thing in 90 days. We built Bolt in 90. So there's a lot of things that were missing in the product. Like, basic stuff, basic stuff. And which, again, we cut the right corners on the thing to get it online, but we had this just growing influx of people using it, going, "How is there not a mobile responsive view? How are chat messages not," we got to 20 million of ARR without a mobile responsive view, by the way. Just throwing that out there. It was like the iPhone not having copy and paste until iPhone 5, or whatever. That was that, this was that for us, it was like, no mobile. You looked at it on mobile, it was terrible.

**Eric Simons** (00:31:13):
But there was stuff like that, so we had to just... And then, we're a small team and so, we were completely unprepared for just the growing traffic. And there was a whole bunch, I mean, the list of problems that were happening every single day was nuts. I mean, to start, we had never had a plan on stackblitz.com for more than $9.00. We had one price, nine bucks. And so when we launched Bolt we were like, "Again, we don't think, hopefully people like this, but nine bucks doesn't get you a lot of inference." And so people burn through nine bucks in 48 hours. And they're like, "I want to buy more. How do I buy more? Why won't you take my money?"

**Eric Simons** (00:31:56):
So it was like, within the week we rolled out just completely new pricing plans, where you could upgrade, which ended up, has kind of now become the standard. All the other guys in the space have copied this. Where prior to Bolt going online, Copilot, all these previous AI things, everyone wanted this Netflix model where there's one price, it's like all you can eat, or whatever. And the problem is, if you do that, you want the inference cost to be kind of low, because you're expecting people to use it a lot. And so you can't do these agentic experience things, it would be too expensive.

**Eric Simons** (00:32:29):
And what we ended up stumbling into is that, "Okay, actually, people are willing to pay more. People want to pay for more inference, because we've crossed this threshold where you can get a very tangible ROI." You know that this is providing a tremendous amount of value to you. So anyways, that was one thing, and the servers were just melting. Anthropic ran out of GPUs for us. Dario emailed me, he was like, "Listen, we don't have anything more to give you." At the times, where we're like, "How do we deal with..." It was just bananas, for weeks. It felt like in 300, when they're surrounded by 10,000 people, and our team is just doing everything. There's 15, 20 people, just doing everything. My chief of staff and I were doing customer support 95% of the day. Anyways.

**Eric Simons** (00:33:22):
So yeah, it was a crazy wild time. I mean, it still is. We've had a little bit more time to grow into this. And usually, I mean as a company, to grow into even 20 million ARR, you get a year at least or something, to kind of staff up.

**Lenny Rachitsky** (00:33:39):
Often, decades.

**Eric Simons** (00:33:40):
Yeah. So that was as hard, we'd go to people and kind of be like, "What do we do?" And the playbooks we get back are, take six months, or a year, or something. It's like, "This isn't going to work." Which is funny, this is what it's all about. I mean this is, at least for me, that level of intensity, it's challenging. Fun challenges, you know?

**Lenny Rachitsky** (00:34:08):
Wow, okay. This is just nonstop wild, wild shit. So you mentioned that your team was about 20 people through all of this, you guys are growing at this insane rate, 20 people. How was this possible? What has allowed you to grow this much, this fast, with such a small team? And this 300 visual is interesting, I imagine having these Spartans is a big part of it. Just what has allowed you to do this?

**Eric Simons** (00:34:40):
Yeah, I think a lot of it again, I mean if you kind of look at where a lot of the other folks, like the Cogen types of app space, have really been struggling, a lot of it has been scaling their servers with stuff. And it was kind of like both this overnight success, seven years in the making. All of this stuff, there's no way, if you rewind to year two, there's no way we could have, we would not be at the growth on DAUs, and revenue, or whatever. There's just no way. And so a lot of it is the technology we made, and most importantly, it's been the people.

**Eric Simons** (00:35:19):
The people... It's rare to find startups where you have the core group of five, six, seven people, that have been there for five years, plus. That's a pretty rare thing to see in Silicon Valley. It is usually, folks are at a startup for a year or two, they kind of go to another one. You know what I mean? And the problem with the turnover like that is that you can't take really long bets like the one we did. And so we've had, kind of from the get-go, again, this comes back from bootstrapping the previous company. Just having less people, and more context per head. That's just been how we do it, and we feel very strongly about it.

**Eric Simons** (00:36:00):
And the reason for that is, one, that you can have high levels of trust with anyone you're talking to, because you know that they have a lot of context. It's not like this person's completely in the dark, in some corner of the company that doesn't... You know what I mean? The second thing, everyone has agency to actually get stuff done, front to back. And there's no political community to get stuff approved by, there's no... So when you look at what happened with Bolt, I mean, we had engineers that were, front to back, were on a call with someone running into an issue, going and fixing it. Cooking up the UI on the spot, and landing this thing. Without involving anyone else on the team.

**Eric Simons** (00:36:41):
So I think it was the culmination of just high trust, and people, we all just have enjoyed working together in the past. Maybe that's why, that's the only reason that anyone would ever stay at a company for that long, or whatever. And so those sorts of stressful situations, I think, are make or break. Those are make or break for any team. And so, I think that what's happened is really, it is a direct reflection of the strength and the bonds of the people that are making this thing, and supporting the thing.

**Lenny Rachitsky** (00:37:22):
Yeah, I think that's such an important point, that you guys have been working together for many years. Most people won't have that benefit. When you're hiring people, when you hire this initial team, is there anything you look for that you think maybe people aren't looking for enough? Anything you prioritize when you're hiring new folks? Is it this idea that they can do a lot? They can do customer calls, they can do design, they can do engineering?

**Eric Simons** (00:37:42):
Yeah, for us, and even if the folks were hiring us, hiring people that don't care about the titles, and they don't care about... It's not like they're... People, of course it's people have a career trajectory, and that sort of thing, but they really are motivated by just working on cool things, and are chucking their ego at the door. And they're there to collectively build something great, not just kind of follow, and be the brilliant jerk. Most of the people that we've hired have been in Europe. We're a fully remote company. My co-founder and I are in the Bay Area. It's funny, back in 2018, we rented an office and stuff, and we were commuting into it. Because we thought we'd hire people here, and like a year into it we were like, "What are we doing? You and I are coming to an office for 10 people, we've hired, the people working for us are in Europe, or across the US." And we have one or two other people we've hired that are in the Bay Area at this point.

**Eric Simons** (00:38:51):
But yeah, I think we kind of look for folks that are intrinsically just trying to build great stuff, and are interested. And then the first people that we hired, the reason that we found them is that they were users of StackBlitz. A lot of people, the majority of people we've hired at the company have been people that actually came from our community, basically. So when we want to hire people, we put out a tweet and say, "Hey, we're hiring an engineer," and then we get DMs or whatever.

**Eric Simons** (00:39:22):
But yeah, those are the general kind of qualities we look for, though.

**Lenny Rachitsky** (00:39:28):
I'm excited to chat with Christina Gilbert, the founder of OneSchema, one of our long-time podcast sponsors. Hi, Christina.

**Christina Gilbert** (00:39:35):
Yes. Thank you for having me on, Lenny.

**Lenny Rachitsky** (00:39:37):
What is the latest with OneSchema? I know you now work with some of my favorite companies like Ramp, Vanta, Scale and Watershed. I heard that you just launched a new product to help product teams import CSVs from especially tricky systems like ERPs?

**Christina Gilbert** (00:39:53):
Yes, so we just launched OneSchema FileFeeds, which allows you to build an integration with any system in 15 minutes, as long as you can export a CSV to an SFTP folder. We see our customers all the time getting stuck with hacks, and workarounds, and the product teams that we work with don't have to turn down prospects because their systems are too hard to integrate with. We allow our customers to offer thousands of integrations without involving their engineering team at all.

**Lenny Rachitsky** (00:40:15):
I can tell you that if my team had to build integrations like this, how nice would it be to be able to take this off my roadmap, and instead use something like OneSchema. And not just to build it, but also to maintain it forever.

**Christina Gilbert** (00:40:26):
Absolutely, Lenny. We've heard so many horror stories of multi-day outages, from even just a handful of bad records. We are laser focused on integration reliability to help teams end all of those distractions that come up with integrations. We have a built-in validation layer that stops any bad data from entering your system, and OneSchema will notify your team immediately of any data that looks incorrect.

**Lenny Rachitsky** (00:40:46):
I know that importing incorrect data can cause all kinds of pain for your customers, and quickly lose their trust. Christina, thank you for joining us, and if you want to learn more head on over to oneschema.co, that's oneschema.co.

**Lenny Rachitsky** (00:41:01):
I want to ask a couple more questions about Bolt, and then I want to zoom out, and talk about where things are heading in the future. Let's talk about prioritization. I imagine you guys are just barraged with, as you described, after you launched, you're just barraged with requests. Like you said, there's a million monthly active users. I can't even imagine the feature requests you guys are getting, plus all the stuff you know want to build. Just how do you go about deciding what to prioritize, and what to actually build?

**Eric Simons** (00:41:28):
There's a lot of things that you just don't even know are possible to do, and so, people aren't going to be necessarily explicitly asking for them. And so there's been kind of a couple of these, where we use our gut instinct on like, "Hey, no one's asking for this, in meaningful numbers at least. But we think this is going to be a big deal."

**Eric Simons** (00:41:47):
Best example was last week with native mobile app support. That's, by reception, the biggest thing we've ever launched. And it was something that, even internally at the company, some folks were like, "This, I don't know. People are yelling about these other things." And it is, it's always this balance of how much are we just triaging various things, versus that new capabilities, but it was like, "This strikes me as an important one," where we put some chips into the middle of the table on. And had it dead right. It's just this mind-blowing experience, and now, there's just thousands of mobile apps being created a day, that weren't before. And how does that change things? I mean, now there's small businesses that, they would've never made an iPhone app before. It made no sense. It's super expensive. Now, that's not the case.

**Eric Simons** (00:42:43):
So there's kind of these things where it's like, "Hey, we should go and take bets here." But there's kind of this, I think the best analogy would be like, it's kind of like working at a restaurant, being like a chef. There's some amount of, there's feedback from the customers of, "This thing didn't taste good." And then there's like, "Hey, we've been cooking something interesting, and this tastes... I don't know. This, I think you're going to like this. I think this is a killer dish." And so you kind of have to balance those things.

**Eric Simons** (00:43:15):
And I think it's actually, largely, a function of just years of experience doing it. I think if you kind of rewound 10 years ago, I would have had really no, I wouldn't have had just the years of getting my butt kicked by the free market, to have cultivated a sense of this stuff. You kind of have to build your own gun and stick for it, I guess is the best way of putting it.

**Lenny Rachitsky** (00:43:39):
To unpack this a little bit further, do you have a cadence you guys work through to decide what to build, and ship? Do you have a weekly meeting every week? Because I know the answer is probably, really, "It's just chaos constantly, and fires we're putting out constantly." I know that's a lot of it, but is there some kind of process that you guys have for deciding what to build, and how to share it, and just work with the team?

**Eric Simons** (00:44:02):
We all meet every day. Pretty much the entire team gets on a call, and we just kind of front to back-

**Lenny Rachitsky** (00:44:07):
You meet, like a Zoom?

**Eric Simons** (00:44:08):
Yeah, every day at 8:00 AM Pacific, we're on a Zoom for at least an hour-

**Lenny Rachitsky** (00:44:12):
Every day? The whole company?

**Eric Simons** (00:44:13):
Pretty much the entire company. Yeah.

**Lenny Rachitsky** (00:44:14):
Wow. For an hour. Okay.

**Eric Simons** (00:44:16):
Yeah. And we just go over everything and I think we're going to probably start, as the team's kind of growing, we're going to start splintering off into different syncs, or whatever. But the thing about just having everyone in the same room every day is that, a lot of people will complain that it's... On Twitter, you'll see people say, "Oh, it's the most expensive use of everyone's time," but it's like, "Yep. But there's 0% fidelity loss in that. Everything, every day, is being audited front to back, and being discussed front to back."

**Eric Simons** (00:44:47):
So when you're in these times of just extreme growth, you want as close to 0% loss, on communications. And so that's how we've been doing it, especially since Bolt went online, and I think it was the week after Bolt went online we were like, "Every day, until we're through this or whatever, we're all getting on a phone call every day, and we're front to back doing this." And again, another reason why more context and less heads, every person at the company is aware of everything else going on at the company. So people can independently be making decisions that are generally, by default, more often correct than not.

**Lenny Rachitsky** (00:45:33):
That is so interesting. I've never heard that before. Especially for company growing, that is like yours. That is super interesting, that that's what you do.

**Eric Simons** (00:45:42):
I don't think we're going to do that forever.

**Lenny Rachitsky** (00:45:44):
Yeah, yeah. Of course.

**Eric Simons** (00:45:45):
But, yeah.

**Lenny Rachitsky** (00:45:47):
No, but I think that's a really cool thing to note, that that works. And has worked for you. Where do you, so say you talk about stuff, then where do you put stuff? Where do you put your roadmap? Where do you plan? Just, what tools are kind of in the stack of the company's toolset?

**Eric Simons** (00:46:01):
Yeah, on the engineering side, we use Linear heavily. On kind of product roadmapping, we're using Notion, and kind of making PRD type stuff in Notion. And we use Figma for design. No, actually, we use Bolt for a lot of design and prototyping at this point, as you can imagine. But yeah, I think that the tooling is nothing crazy. There's nothing crazy sophisticated. I think we'll be investing a lot more, and especially as you start splintering people out of being on the same call every day, so that's where this stuff really starts to matter. Because you don't have a time where you're able to dynamically catch things that weren't going to be brought up.

**Lenny Rachitsky** (00:46:45):
I love that you guys use PRDs. I love that you even used that term. There's a lot of talk of just like, "Oh, we got Bolt now we got all these tools, we don't need PRDs. We're just going to create a prototype immediately, and that's it." Talk about just why you still find that useful, and just what you put into your PRD, whatever that is for you.

**Eric Simons** (00:47:01):
Unless there's something that's very sophisticated that we're working on, we tend to keep them pretty light. I like to just have the minimal amount of context possible, that just ensures everyone's on the same page and that the key outcomes for whatever feature that we're working on, are going to be present when we get there. Because the things that, when these arguments get really beefy, you're looking at it at, "God, there's so much stuff to decipher." The problem is, a lot of people are going to gloss over it when it gets kicked to development, or design, or whatever. It's just going to start snowballing into a lot of stuff. It's just better to keep it as simple as you possibly can. At least, that's our approach to the thing. And often it's some of these things are like, "Here's a link to a Bolt."

**Lenny Rachitsky** (00:47:46):
"And here's what it might look like."

**Eric Simons** (00:47:51):
Yeah. And not just look like, "Here's kind of a working demo of what it will effectively feel like." And then, be. Because that just, if a picture is worth a thousand words, a live actual demo is worth millions. You can feel it. It's real. And that's what we're seeing, a lot of the businesses that are adopting Bolt now, that's the use case that they're using this for. Is high fidelity prototyping, because it's now faster to make real prototypes using Bolt. Before, it was too expensive. The idea of, "And let's prototype it, the engineers' code a proto..." It's like, that, it would take forever. It would be expensive. And now it's faster to do this with Bolt, in code, and have a real working software product, than dragging around frames and Figma to actually make a static version of it.

**Lenny Rachitsky** (00:48:43):
So let's actually talk about that. Just how far have companies gotten with Bolt? Prototypes is where everyone's kind of imagining these tools are at. I know that the goal isn't just to make prototypes, it's to build full scale. I imagine, long-term, Salesforce, Atlassian style companies, at scale. What are some examples of products people have built with Bolt that maybe would surprise people, in just how far they've gotten?

**Eric Simons** (00:49:08):
Yeah, I mean, especially... When you're starting greenfield stuff, you can use Bolt to build... you know. Like Salesforce, as an example. One of the first people that signed up for Bolt was this guy named Paul, and he's an entrepreneur, and doesn't know how to code. Built a CRM in three weeks, that has AI built into it and Stripe for billing, et cetera. He had gotten a quote from an agency for this, it was going to be 30 grand, and take six months. He had done in three weeks, and I think he spent 300 bucks on Bolt for the thing. So it's like, this is... And he's making money off of this. This is his start. Right?

**Lenny Rachitsky** (00:49:45):
Okay, so he built this, and he's selling it. People are paying to use it.

**Eric Simons** (00:49:48):
Yeah. Yeah.

**Lenny Rachitsky** (00:49:48):
Wow.

**Eric Simons** (00:49:50):
And there's many such cases of this. If you're looking at greenfield projects, 100%. Today with this current state of frontier models, you could absolutely build production grade software. You're not going to get a zero shot, but you're going to spend a couple of days, weeks, whatever. But the cost reduction there, 30 grand versus $300. It's 99% cheaper. Six months versus three weeks. I mean, it's like order of magnitude sort of fashioned delivery on the thing. And those numbers have helped, for the people that we talk to, that are building these full stack apps. People, they go to Upwork, work, they get a quote for five grand. They have it within 50 bucks. It's just nuts, what you're able to do with this thing.

**Eric Simons** (00:50:35):
And so I think on the flip side, a lot of the existing companies, there are very legitimate use cases where things are greenfield, spun up. A good example is public websites. Marketing pages, landing pages, whatever have you. Folks are adopting Bolt to just power those instead of using Webflow, for example. Because it's like, this is simpler to use than Webflow. And it integrates with the existing design system of the company, and et cetera. And the marketers can update it without knowing how to code, whatever. But then for product development teams, this is most commonly for, again, existing software businesses. They're using this to just accelerate the product development process, and in a way where it's not just like a Greenfield wholesale, "Hey, we're building the entire thing in Bolt," or whatever.

**Lenny Rachitsky** (00:51:20):
Can Bolt integrate with your existing code base, or not yet?

**Eric Simons** (00:51:23):
So yeah, we can actually open up repos in Bolt. You can go and use Bolt on your code base. It kind of depends on your setup. And we do have companies, again, that have marketing sites they're using this on, or their admin panel or whatever. And I think it's going to be a use case that we see a lot more people orienting towards. These LLMs are not great, depending on how big your application is, though. These things are not quite there, where if you have something that's a thousand files or something, or more, where you're going to be able to have a really reliable, super reliable experience per se. Within a year, we'll chat a year from now, I suspect the answer is going to be different. So it kind of depends on the size of the app, the scale of the app, and if it's too big, you're looking at the prototyping, just pure acceleration of product development. And if it's not, then you can just do it entirely from Bolt.

**Lenny Rachitsky** (00:52:24):
So this is useful. So what would you say are the major limitations of Bolt today, where people should just know, "Okay, it's not going to get you here yet. Maybe in the future it will." So it sounds like, if you have a really large existing code base, probably not the best tool yet. What else should people know?

**Eric Simons** (00:52:39):
I would say that's probably the main one, because I think if you have a large existing code base, you're going to need something like Cursor. And you're going to need to be a developer, meaningfully, to be editing that stuff. I think outside of that, there's a, just like using any other productivity tool, like Photoshop or Figma, or like a DSLR or whatever. There's some level of education, and using the tool, and learning how to use it, that's required to really unlock a lot of the maximum capabilities of the thing.

**Eric Simons** (00:53:10):
And the people that we see that are most successful with Bolt, outside of developers, the people we see that are most successful are people that are amazing PMs, for example. Because these are people that understand enough about how the technology works, typically, and their job is to direct developers on how to go and improve the product. And go and look into how to actually spec this thing out in a way that's executable, without lossiness in the communication. And when you think about, "Okay, how would you best interact with an AI developer agent?" It's basically that. You really want to be good at defining scope, and helping it go and debug various things, or whatever have you. There's a huge overlap of the skill set of being a rock star PM, and being really good at using, frankly, any of these text to apps or Cogen tools.

**Lenny Rachitsky** (00:54:08):
I love that you made that point. That's exactly the point I've been trying to make, I have a newsletter post about this. Because when all these tools came out, there was so many people saying, "Okay, PMs are dead. We don't need them anymore. We can just build things so quickly and easily, what's the point?" But I completely see the world the way you see it. The hard part now is, now it's easy to build the thing. Now it's, "What the hell should we build? Can we clearly articulate what it is we want to build?" And then, "Can we just have the taste to know, is this right, is this correct? Is this good? Is this going to solve the problem?" And then it's like, grow it, which is something also PMs think about. So, I completely agree. Basically it feels like PMs are, and a lot of PMs listen to this, so they'll love hearing this. To me, it feels like PMs are the best positioned role to thrive in this world.

**Eric Simons** (00:54:54):
Zero question. I mean that was, as Bolt was growing and we were like... Because we were a developer product before this, and so we expected the audience to be 100% developers that were using this. And we just kept seeing more and more and more people that were not developers using it, to the point where it's like, 67% of our users are not developers, at this point. And when I started talking to these folks, at first I was just weird, or whatever. It was like, "Well, what's going on here?" But then it just kind of clicked as like, "Oh, well, this is going to change everything. The entire software world order is going to get rewritten, here."

**Eric Simons** (00:55:30):
Because the way that companies are organized to build software today, totally going to change. The idea that again, PMs are the people that really understand, to the pixel level, what matters into making a great product experience. And often they're having... And listen, I'm a developer, myself. They have to go and harangue the developers to get things to be how they really ought to be, to the smallest levels. And now, how this is going to work, if you fast-forward one, two, five years, whatever. PMs, they're going to be "writing code", quote, unquote, instead of just writing a JIRA ticket and waiting for a developer to do it. The developers are going to be able to work on intellectually challenging tasks that LLMs are not well suited for, and still being augmented by LLMs to do it. But PMs are going to be able to go in and just make the changes themselves.

**Eric Simons** (00:56:24):
And what blew my mind is, it's not priced in, to any of these companies out there. And it's not reflected in the org charts of all the software companies in the world right now. That is going to completely change. The winners, at least, their org charts are going to completely change, and how they approach building products and shipping products. Completely. And it's starting, this is the beginning.

**Lenny Rachitsky** (00:56:50):
I want to follow that thread, but first of all, I want to also add, and correct me if you disagree with this. I think when we talk about PMs, that also applies to founders, like product thinking founders-

**Eric Simons** (00:56:58):
One hundred percent.

**Lenny Rachitsky** (00:56:59):
... very similar. And then, I think it's also important to note, if you also have engineering skills and design skills, you will be at an advantage. That only helps you. But if you're looking at this triangle of the triad of product, engineering, PM, it feels like the PME skills are the ones that will be most important and valuable. Although it'd be great if you can also be in code, and if you could also design really well.

**Eric Simons** (00:57:23):
Absolutely. And to me, it's the most exciting mix. I think PMs, designers and entrepreneurs that are non-technical, that's the most exciting thing to me, just because it's a brand new market that's being unlocked here. For the first time ever, these folks can directly code and build the product, themselves. Their vision, directly into the software itself. That's going to change everything. That is changing everything.

**Lenny Rachitsky** (00:57:48):
So when you talk about how org charts are going to change, what are you imagining there? Is it just fewer engineers, mostly, or what do the future org charts look like?

**Eric Simons** (00:57:57):
Good question. And I bet you there's going to be some Gartner analysis someday, years from now or whatever that's like, "Here's how the best," some term is applied to how the best companies are organizing. But yeah, I think that we're going to see developers probably being pulled off of a lot of the, generally speaking, pulled off of a lot of user interface type work. I would imagine. Except for the most complicated of those things. And you're going to see designers and PMs really, really leading the charge, and being responsible for crafting those experiences. And perhaps having a developer attached to be reviewing the code and making sure the guiding, the code that they're writing, reviewing those pull requests and et cetera. And I think maybe even the engineers are... Like you pointed out, having engineering skills is not going to hurt you. It's going to make you way more effective.

**Eric Simons** (00:58:58):
But I do think there's going to be, the leverage that the front engineer is going to have is, it is now insane, it's only going to get more so. And so I could see there just being fewer front engineers attached to, I'm seeing more product and design folks, with one or two engineers or something. And really having a larger matching of pods like that. Something like that strikes me as probably how this is going to start trending towards.

**Lenny Rachitsky** (00:59:34):
This touches on, we had a researcher from OpenAI in the podcast. She actually started her career, she worked at Anthropic first, as a front-end engineer. And said that once she saw what Clyde could do, for front-end engineering, she's like, "I need to move to a different function." And so she moved into research, because she saw that role disappearing, potentially. And that's exactly what you're saying.

**Eric Simons** (00:59:57):
Yep.

**Lenny Rachitsky** (00:59:57):
So let me ask you this, I don't know if you have a clear thesis on this yet, but say, you just had a kid. Say your kid is, in the future, starting school. Let's say your kid was starting college soon. Do you have thoughts on just what skills slash areas you think they should go into, versus avoid, that maybe are popular now and are going to be less popular?

**Eric Simons** (01:00:19):
Understanding how to leverage these AI tools is key. I wouldn't necessarily, I think maybe getting a basic understanding of how programming works, et cetera, is great.

**Lenny Rachitsky** (01:00:31):
It's like technical foundations, just understanding how systems work, how coding works.

**Eric Simons** (01:00:41):
Exactly. But it doesn't have to be, because I think back to if Bolt existed, like Albert and I say this to each other all the time. Since the get-go, StackBlitz, we've been building the thing that we wish we had when we were 13. And heck, for everything we built since then. And especially with Bolt. I mean, I don't know if I would've gone as deep as I did on learning how to code, and being an engineer, if that had been around then. The whole reason we got into it is we had ideas for products, and businesses, that we wanted to build. And coding was just a necessary requirement in order to do that.

**Eric Simons** (01:01:21):
And that said, I think people need to follow their intrinsic interests. If folks are really interested in really getting in the nitty-gritty of how computers work, and program leaders, or spark and compilers, or whatever, go for it. I think that stuff is still going to be relevant. I don't know if we're going to really have, we'll see, but to the degree that there's AGI where it's like, we don't have to think about anything ever again.

**Lenny Rachitsky** (01:01:47):
Yeah, that's always the answer here. Do we sell anything?

**Eric Simons** (01:01:50):
Yeah, it's like if we're at that point, it's kind of... I don't know, I'm not sure. But I think from, at least what I feel like seems like the next at least five years of what we're looking at, I think people are still going to, there's still going to be places to specialize, and really go deep. But I think you want to go into it with the idea, not like, "I'm going to go and learn computer science because I'm going to get a job for sure out of it."

**Eric Simons** (01:02:18):
I just think that's generally not a good... This is like, my co-founder and I, we didn't go to college. My co-founder dropped out of college after a semester or something, but I didn't go, because I was like, "We're coding," we were doing contracting at the time, making money. And it was like, "This is a lot of," you know, it would've been like a hundred grand of debt by the end of the thing, just for four years of in-state tuition, at U of I. 120 grand, I think, at that time.

**Eric Simons** (01:02:49):
And lo and behold, I mean there's a huge issue with this. Where people are kind of... There was a prevailing thought by society that going to college in the early 2010s or late two 2000s, that you're going to get a job on the other side. That's going to be high paying. And that just has not been the case for a lot of people. And I think that's just going to continue to be the case. But again, not to deter people from doing it, but you have to go into it being like, "I for sure, this is what I want, and I want to go and be the best that I can possibly be at this thing." You know what I mean?

**Lenny Rachitsky** (01:03:30):
I like the transfer to your kid is going to be like, "Don't even go to college," potentially.

**Eric Simons** (01:03:35):
Only if they want to. I think at 18, it's a huge ask. I mean, it's a huge ask, not even at 18. It's like at 17, because you could go apply for colleges. It's just such a huge... Like, a six figure debt commitment to someone who's making $0, or negative dollars, and that young. Unless you really have conviction, it costs nothing to go and explore and learn for free, online.

**Lenny Rachitsky** (01:04:03):
I want to come back to the skills that you think are going to be most important, and let me try to mirror back a few things you said, that I very much agree with.

**Lenny Rachitsky** (01:04:11):
So it feels like if you want to be successful in the world where AI can build things for you, more and more, what I'm hearing is get good at figuring out what people need and want, what problems they need solved. Get good at articulating it really well to the AI tools. And there's this talk, "You don't need to be a great prompt engineer. You don't need to work on prompting," but it feels like it's more and more important, because you tell it something and it goes off and builds a thing. If you're clear about it, it'll save you a lot of time.

**Lenny Rachitsky** (01:04:42):
So it's be good at figuring out problems people need solved, figure out how to articulate that problem well, and ask for a clear solution. Figure out how to grow the thing, feels like that's still going to be a need, because Bolt's not going to go and find every... I could see that still running paid ads, and stuff like that. But it feels like that's going to be ongoing need.

**Lenny Rachitsky** (01:05:03):
And then I feel like there's this kind of unstuck step, like helping AI get unstuck, and it feels like that's where maybe engineering skills will come in more and more. Thoughts on just that skill?

**Eric Simons** (01:05:13):
Oh, totally. Yeah. So we actually, two weeks ago I think, we announced this program called Bolt Builders. And it's basically the genius bar at the Apple Store, where as folks are building on Bolt that are not developers, they'll run into some nook or cranny where the AI just cannot figure it out, or whatever. And I think that's just going to continue to be the case for the time to come. That's our position, and that's why we spun up this program.

**Lenny Rachitsky** (01:05:38):
And these are humans, that help you out.

**Eric Simons** (01:05:41):
These are humans. And people that we're certifying. And so, in Bolt in the coming weeks or whatever, there's going to be a button where you can just say, "Hey, connect me with a certified expert." And you can chat with them live, and they'll help you get unstuck, and you pay I don't know, 50 bucks an hour. Whatever it is. And then you get unjammed and you keep prompting. And again, I just think this stuff is, it just all seems like gravy to me. Engineers get to focus on difficult challenges, not like cookie cutter, "Let's make another CRUD app," stuff. They get to, debugging is challenging, and fun. And going and working on intellectually stimulating tasks, and all this stuff that's just copy pasta, over and over, all this, error app. It's just like, let the AI do all that kind of crap.

**Lenny Rachitsky** (01:06:29):
This is a potentially new job, for now at least, just unstuck the AI. Which I think over time, it'll get better and better, and we maybe won't need these people. But I love that it's now AI first, and person second. Versus person building the thing, and then AI. Like when Copilot launched, it was just like, "Cool, here's a little suggestion for this function," and now it's flipped. "Here's everything." And then, "Oh, I don't know what to do here. Help us here." And then, it's like a human suggestion. Isn't that interesting? It's like, human Copilot is flipping it.

**Eric Simons** (01:06:58):
Totally. Yeah. That's what's wild is, I think Sonnet was really the first model that flipped the equation, because that was really us, and old Cursor, and all these other things. The rapid growth started the second Sonnet went online. We actually tried building Bolt almost exactly a year ago, with the frontier models at the time. Spent a week or two building it. It just didn't work. The output, the code output was not reliable enough. It would constantly, it would be a broken app, or it would look ugly, or whatever. And then we got a sneak peek of the Sonnet stuff in May and we were like, "Oh. Okay, we should take that project back off the shelf and green line it, because this might be it."

**Eric Simons** (01:07:36):
And lo and behold, that's exactly what has happened. But yeah, that's the big deal that is, kind of under the hood, this is... What's going on here is, a very critical threshold has been passed with LLM's ability to write production grade code and apps that actually look beautiful, and actually function well. It's not perfect, but there's kind of this zero to one moment that's happened where it's like, "Okay, so now, yeah. Now the AI is the first thing," and then you're kind of popping a developer in every now and then, versus the other way around.

**Lenny Rachitsky** (01:08:11):
I did not know that. I didn't realize that so much of this was unlocked with, like it's sitting on top of Anthropix work, and specifically Sonnet. That was the first model, you're saying, that could code well enough.

**Eric Simons** (01:08:22):
Yeah, zero question.

**Lenny Rachitsky** (01:08:24):
Wow.

**Eric Simons** (01:08:25):
Zero question. Absolutely.

**Lenny Rachitsky** (01:08:25):
That is fascinating. Just the amount of, I don't know, revenue and business and ecommerce that that one model has unlocked, is insane. I did not realize that.

**Eric Simons** (01:08:39):
It is. And in retrospect, I'd mentioned, we'd never done an AI product at StackBlitz, and it's tempting. Like when of ChatGPT went online, everyone started adding AI to their products. We just didn't see a clear place for a really added value. So I was not super bullish on, you know, a lot of people were like, "AGI is going to be here in 2023." You know what I mean? There's all this stuff that was being said, and I was like, "I just don't know if I necessarily buy how fast people say that it's going to move." And to a certain degree, that was the correct view.

**Eric Simons** (01:09:13):
What I didn't really think about though, is if AI, if LLMs are going to get better at a specific vertical, which are going to be the things that it would be. And if you look at law, for example, you want to make the best LLM for looking at case law. The problem with that stuff is, it's not deterministic. The judge's ruling is dependent on society's view of things at the time, political stuff going on, the jury. There's a ton of things, that it's not deterministic. And so you can't really create a lot of training data that's going to be super reliable, and produce really good results. And you can't just make up cases and say, "Theoretically, the judge would say this." Because, I mean, it just doesn't work.

**Eric Simons** (01:09:57):
Software is deterministic. When you write code and you hit run, it either runs or it doesn't. And that's the key insight Anthropic really had. They just went deep. And then, this is what they're doing, is just reinforcement learning on basically permutating every type of app you could ever build, and just spinning up tens of thousands of cores or whatever to do that. Just building tons of training data, and doing reinforcement learning, and making their LLMs the best in the world at building beautiful, reliable applications. I'm extremely bullish. It makes technical sense why, of anything, LLMs are going to get insanely better at writing code than probably most other types of applications for LLMs. Simply because it's something that can be extremely deterministic, and permutated thousands and thousands and thousands of times per second.

**Eric Simons** (01:10:54):
And so I think the broader trend here is... And Sonnet has woken everyone up. Google, and Open AI, all the... Everyone is now gunning for coding because, how big is the market opportunity to rewrite the software world order? It's trillions of dollars, or something, right? The world runs on software. So I think that just in a macro, the highest macro view, and why we went and raised money for Bolt. This seems like an extremely clear shot, there's this, you can kind of separate the hype of what people say and blah, blah, blah. When you break it down technically, this makes sense. It makes sense what's going to happen here. And for us, we're like, we are so happy to be well positioned to go and enable people to kind of ride this wave of the innovation that is here in LLMs, and is going to just keep coming, and therefore enable more people to build even crazier amazing software. So that's our world view, at least, of what's going on here at the macro.

**Lenny Rachitsky** (01:12:05):
And when did Sonnet even come out? It's been a while, right?

**Eric Simons** (01:12:08):
I think they officially, I think it was in June, when they officially put it online.

**Lenny Rachitsky** (01:12:12):
So since June, this is the worst it will ever be, the state of AI coding. And it's already this good. And there hasn't been anything, like they haven't launched their new model, since last June. So this tells us just how quickly things are going to start moving once they launch their next model. And as you said, everyone's gunning now for this, because they realized "We're behind on the coding piece." So, wow. This is going to get crazy.

**Eric Simons** (01:12:40):
I agree. Yeah, it's been, again, there was no blog post that laid all of this out for us. It's just been kind of this-

**Lenny Rachitsky** (01:12:48):
You just noticed the code was really good, basically.

**Eric Simons** (01:12:50):
And from there, it's just, we've been piecing together all this other stuff. So it's been kind of the thrill of a murder mystery of, "What is going on here?"

**Lenny Rachitsky** (01:12:59):
Oh, really.

**Eric Simons** (01:13:00):
Yeah. You know what I mean?

**Lenny Rachitsky** (01:13:01):
Just piecing together what you've seen on Anthropic releasing, is that what you mean? What have you been noticing, murder mystery-wise?

**Eric Simons** (01:13:06):
Well, and then the impacts of Bolt, where we have people that are not technical using this. How are they using it? Why are they doing this? And then, all the stuff we've talked about in this podcast has been the result of nine months of just R&D and seeing the results of it, and then going, "What?" And then digging in, doing another thing and then going, "What," again. And then it just keeps happening, because there's no charted course for this.

**Lenny Rachitsky** (01:13:31):
It's like an anthropology, if that's the right term, of just watching. It's like an emergent discovery, it sounds like. Versus you had the strategy, "Here, we're going to do this," where a persona to launch this thing will happen?

**Eric Simons** (01:13:42):
Yeah, a hundred percent. Yeah. I think that that's the best way to put it. It's the best way to put it. It's exactly that. And so it's very interesting to just-

**Lenny Rachitsky** (01:13:54):
To watch, and be a part of it, I imagine.

**Eric Simons** (01:13:56):
Yeah, yeah.

**Lenny Rachitsky** (01:13:57):
And I think as people say, "Okay, these are just toys, they're prototypes, it's not going to work with your existing code, it's not going to scale." It's important to note just what we talked about. This is a model from last June that this is possible on, and everybody's working on the next cutting edge model that will make this even better, and that's going to come real soon. Okay. Amazing.

**Lenny Rachitsky** (01:14:20):
I just have a few more questions to close out this conversation. One is just, what is coming next for Bolt? What are some of the cool new features that'll be launching before this comes out? Maybe right after this comes out, maybe in mid-March?

**Eric Simons** (01:14:33):
Yeah. Okay. So by the time, and I'm going to go back and tell our engineers, "I said this on this podcast-"

**Lenny Rachitsky** (01:14:33):
"I've committed. Sorry, guys."

**Eric Simons** (01:14:40):
This is, I found actually being a leaky faucet on talking on podcasts and stuff, my engineer is like, "How could you tell them..." "You just have to ship it faster, now. You got to make it real," right?

**Lenny Rachitsky** (01:14:49):
Or we'll get AI on it.

**Eric Simons** (01:14:52):
No, no. Yeah. But I think for us it's like, again, we've seen a lot of PMs, designers, entrepreneurs, et cetera, using Bolt. And so, we're really looking at better fitting with the tools that folks are using to do those things today. Bolt's not going to replace them, or something. And if you're working at a company, how do you integrate this stuff with your existing business, or your existing product and code base? Because that's the question we often get from people is, "How do I open my," like we're talking to one of the fan companies the other day, and like, "How do we open our production code base in this, that's like 20 years old?" I'm like, "You don't. None of this stuff. That's not what you do. This is for rapid product development in your use case."

**Eric Simons** (01:15:41):
So the features that we're going to be shipping, I'm pretty stoked about this one, so we've been working on this for a while and we've partnered up with a company called Anima to do this. But basically, so on any Figma URL, when you're looking at a design that you've made, if you just put bolt.new in front of that URL and hit enter, it's going to suck that design into Bolt, and turn it into a full stack app or mobile app, just out of the box.

**Lenny Rachitsky** (01:16:08):
That is genius.

**Eric Simons** (01:16:09):
Yeah. It's-

**Lenny Rachitsky** (01:16:09):
Amazing.

**Eric Simons** (01:16:10):
It's going to be nuts. Yeah, I mean, it's really fun to use. Because whether you're a developer or designer, or whatever, going and taking that and turning it into an actual coded app. And the thing is, once it's in Bolt, you can just keep prompting from there. You're like, "Yeah, well, add another page here." So you can have things that, where you want pixel perfect design, you can have it, and it'll translate one to one. And it's splitting out the assets. Anima's been doing this since 2017, Figma to code. They've got the best agent in the world for, they're the number one Figma plug-in, or whatever. And so in Bolt, it's going to just work. It's just deeply integrated

**Lenny Rachitsky** (01:16:53):
So it's bolt.new slash, the Figma URL, to the design.

**Eric Simons** (01:16:56):
Yep. That's all you got to do.

**Lenny Rachitsky** (01:16:56):
Amazing.

**Eric Simons** (01:16:57):
And in Bolt, we're going to also have a little Figma icon in the chat thing. So if you go to Bolt itself, you can click it and then paste the URL, or whatever. But yeah, but it's like that. It's from Figma to full stack app in a click. Literally. That's crazy. So that one's pretty cool.

**Eric Simons** (01:17:15):
And then the other one that we're working on is an integration with Slack, because often when you're talking about Figmas, like Figma links or whatever, at a company, you're in Slack or whatever. So you're having conversations about, "Hey, we should really add a page to this that does da-da-da." And so we're actually creating a Bolt Slack bot whose job is to basically act like a developer on your team. And so you can be, in a thread, you can be like, "Hey, I think we should a homepage." "Yeah, okay. @Bolt, can you whip this up real quick?" And it'll go and just suck down the entire conversation history so far in your Slack, on the thread or whatever.

**Eric Simons** (01:17:54):
You'll be like, "Okay, cool. So you kind of want me to take this Figma URL, which I can convert automatically," thanks to the future I just mentioned right before this. "I can just go and convert that thing out of the box, and then you want me to add a page to it, and then do this thing. Got it. I'm going to go do that. Oh, here's the URL where you can open it up and keep prompting." You know what I mean? So it's like having a developer or somebody to kind of kick this thing out. And you're just like, "Go make this thing real quick."

**Eric Simons** (01:18:15):
So those two things, I think it's, again. You start think about how our company's going to change how they're currently doing product development. And even just like, "Hey, we need to spin up a marketing site. Here's the thing." We're like, "Can you do that, Bolt?" "Yep. Let me go do that." You know what I mean? And that's why I'm kind of excited about those two, in particular, because I think it's going to be well received, and folks are going to be stoked about it. I hope. Knock on wood.

**Lenny Rachitsky** (01:18:40):
Those are awesome features. I love the Slack piece, because when I think about agents, there's always talk about agents. To me the simplest way to understand it is just a Slack bot, just that AI can talk to you like they're a person in Slack. And I love that's exactly what you're doing. And this is this gigantic feature of just like, "Hey, you just have this engineer now that can go build stuff for you."

**Lenny Rachitsky** (01:18:59):
Let me actually ask you a question along these lines, that I was meaning to ask, but I forgot. Is just, your engineering team, what are they using to build Bolt? I imagine it's a lot of Cursor. How much is Bolt, at this point, involved in building Bolt? And is there any other tools that they found useful, find useful, that are worth highlighting?

**Eric Simons** (01:19:18):
Yeah, good question. Yeah, we definitely use Cursor. Our folks use Cursor a lot. We use Bolt a lot for the product development process, like a ton, we're using it. And we're doing basically the flow that I described, where if things that need to be Pixel perfect, we're going to Figma for. And often we're taking that and we're pulling that into Bolt, because we've got access to the integration today. So pull that into Bolt and we're saying, "Hey, go add these things or whatever." Or just saying, "Hey, here's a screenshot of our UI, go do da-da-da."

**Eric Simons** (01:19:47):
Other AI tools that the developers are using. I think those are the primary ones. I mean, I think we've got a subscription to Claude, and ChatGPT, and things like that. But I think for development, Cursor is the main thing,

**Lenny Rachitsky** (01:20:03):
Yeah, it's cool how few tools, like there's so many AI tools, and it's interesting how few people actually end up using. It's like Cursor, Claude, ChatGPT, and then maybe another tool. Like Bolt.

**Eric Simons** (01:20:13):
Yeah. Totally.

**Lenny Rachitsky** (01:20:14):
Okay. Final-ish question. Say somebody is opening up Bolt for the first time. What's something that, imagine you could sit next to every new user that's just trying Bolt for the first time, and you could whisper a tip in their ear to be successful with Bolt. What would that tip be?

**Eric Simons** (01:20:33):
And this is like, because we have a lot of different types of users. I imagine you're talking about PMs or designers, and that sort of-

**Lenny Rachitsky** (01:20:39):
Let's do PMs. That's a good one. That's a lot of the audience here.

**Eric Simons** (01:20:41):
I would say, talk to this thing like you do a Linear ticket, or a JIRA ticket. That would be my advice. And talk to this like you would, like you're talking to one of the developers on your team. And what that means is, be specific on things that matter. And on things where, also, you can let it be creative. You can go to and just say, "Hey, make it prettier." And it does a good job, it actually does a really good job, when you give it just, vibes. So anyway, I think for PMs it's like, you have the skillset. You know how to do this. This is, just think of this as your coworker, your developer coworker.

**Lenny Rachitsky** (01:21:24):
I love that. Because these tools are so easy, you just go in and it's like tell it a thing, and then, cool. You have a website. It's coded, boom, done. And what I'm hearing here is, take a little time to craft your ask. It may be tempting to just start, "Cool. Build me a serum," and then you're stuck with that first version. And then you're like, "Oh, well, okay. I didn't mean that." So it sounds like your advice is take some time to craft the ask, and be clear about what you want.

**Eric Simons** (01:21:49):
Yeah, totally. Especially if you have a clear vision of what you're trying to build. And something reasonably sophisticated. And what I recommend everyone to do, if it's your first time trying Bolt and you're like, "What should I have it do, and I don't have an idea"? Tell it to build you a personal website. There's something like magic. You take your LinkedIn copy, and paste your LinkedIn bio and work experience, just like select text, copy, paste it. "I need a website, my name is so-and-so, here's my LinkedIn history. My favorite color is blue, and I like dogs." And then hit paste, right?

**Lenny Rachitsky** (01:21:49):
"Make it prettier."

**Eric Simons** (01:22:24):
[inaudible 01:22:24] actor. Yeah. And then, you know what I mean? And then you can hit deploy. And if you don't have a .com yet, now you can. Right? I mean, now you have a real, personalized service. I think there's kind of a moment around that where it's like, "Oh, okay, wow. This," it's zero shot, zero shot 99.999% of the time. You're getting a beautiful personal website that you didn't have before, that would've taken you an hour, on Wix. If not more. And that gives you the taste of, "Oh, okay, cool. So if I really take the time to think this through, and make a PRD, and then put that in piece by piece into this thing, the sky's the limit."

**Lenny Rachitsky** (01:22:55):
Eric, this has been just insane on so many levels. I have so much to process, I think a lot of listeners do too. Maybe as an actual final question, I saw this story about how when you were starting StackBlitz, or maybe even before StackBlitz, you squatted in the AOL office. Because you had some badge that still worked. Maybe just tell that story.

**Eric Simons** (01:23:18):
Yeah, that was the thing I was most known for. That happened like 2012, and I was 19 years old, so it's been a very long time. I think I'm 33, now or something.

**Lenny Rachitsky** (01:23:27):
The statutes of limitations are expended.

**Eric Simons** (01:23:30):
Yeah, yeah, exactly. But yeah, for a while I was like, "I have to do something more notable than living at AOL. This can't be what I'm known for." But now people are like, "Oh, wait. You're the StackBlitz guy, and you're the AOL guy?"

**Eric Simons** (01:23:43):
So when I first came out to Silicon Valley, we got into a, I was building a K12 education startup at the time, and this is back during the days where Y Combinator, they only gave you like 20 grand, and so there was this offshoot of Y Combinator called Imagine K12, it was Jeff Ralston who, I think he was CEO of YC a couple of years back. And then Imagine K12 merged into Y Combinator a couple of years after Rob went through it, but anyways, they had an office space at AOL. At the time, AOL was trying to reinvigorate the company and they were like, "Hey, we should get young startup blood in here, so let's rent out office space to teenagers," basically.

**Eric Simons** (01:24:24):
So I was there, and we ended up running out of money. 20 grand doesn't go very far in the Valley, so three or four months in, we were like, "Oh God, what do we do?" And I was going to the AOL office multiple times a week, because we had access cards to get in to get to the investors' offices. And I realized, I was like, "You know, they have couches here, and they have food. There's ramen that you can microwave, and there's a gym where there's a shower, and even you can do laundry." And then, so I was like, "I don't know, maybe while I figure this out, I'll just live out of here." And so that's what I ended up doing for, I think, four or five months. I was living out of this headquarters over on Page Mill in El Camino, in Palo Alto.

And then, I got away with it for a while just because the guards, the security guards, they worked 12-hour shifts. And so the guys that, when I was there at night... And I was coding, all day every day, basically. So the guys at night just were like, "Dang, this guy works really hard." And then in the morning they'd be like, "Wow, this guy is working really hard." And I became friends with some of them, and then eventually, I think there were also a whole bunch of Stanford students that I think they put bunks in one of the aisles. It was just started getting out control, so I think they started cracking down. And then one morning, at like 4:00 in the morning, a guard came in and threw me out.

**Eric Simons** (01:25:52):
I'm from Chicago. I don't know anyone. At that point I'm like, "I know no one in the Bay Area." So I went to a Starbucks, which was not open. I slept on the table outside of the thing. And I think I hit up one of the other entrepreneurs that was in the program. I was like, "Do you have a couch? I think I kind of need it, at this point." Yeah, the press got wind of it, and it was this worldwide story. But I lived on a dollar a day. That was the crazy thing. My burn rate was a dollar a day, at that time.

**Lenny Rachitsky** (01:26:22):
What did you use that dollar for?

**Eric Simons** (01:26:25):
This is back when McDonald's had the Dollar Menu. Literally. So it was like, I occasionally would go and get a cheeseburger or whatever. Yeah, it was ultimate scrappiness.

**Lenny Rachitsky** (01:26:41):
You're technically homeless. From homeless, to one of the fastest growing startups in history. Eric, what a journey. This is such an interesting point in time of your life, and of just tech. No matter what happens, I'm sure you'll be extremely successful, but it's such an interesting just point in that journey. And I'm thankful that you made time to share it with us.

**Eric Simons** (01:27:01):
It's always good to just have the perspective of, you should start companies to keep the mindset that you're doing it to have fun. So, stoked to see where this goes, one way or the other. It's going to be interesting.

**Lenny Rachitsky** (01:27:16):
Eric, final questions. Where can folks find online if they want to reach out, maybe follow up on some stuff you shared, and how can listeners be useful to you?

**Eric Simons** (01:27:23):
Yeah, totally. I mean, yeah, so bolt.new is the website, over on Twitter, I'm @ericsimons40 on Twitter, and I think our Twitter account is @boltdotnew, not with a period. It's like, B-O-L-T-D-O-T-N-E-W. And yeah, I'm curious to hear what folks think. I mean, this is, again, we are learning so much from the people that are coming and trying this thing out and giving their feedback. And within the first meeting of it going online, we were not the experts on how use the tool anymore, and it's been that way ever since. And so, I love hearing from folks on what they want to see next, and how this is helping them. And where they run into problems, like where we need to go and fix things. So my email address is Eric@stackblitz.com. That's Eric with a C. So I'd love to hear from anyone, whether it's a DM on Twitter, or an email.

**Lenny Rachitsky** (01:28:18):
Amazing. Eric, thank you so much for being here.

**Eric Simons** (01:28:21):
Awesome. Thank you so much for having me. This is a blast.

**Lenny Rachitsky** (01:28:23):
Bye, everyone.

**Lenny Rachitsky** (01:28:26):
Thank you so much for listening. If you found this valuable, you can subscribe to the show on Apple Podcasts, Spotify, or your favorite podcast app. Also, please consider giving us a rating or leaving a review, as that really helps other listeners find the podcast. You can find all past episodes or learn more about the show at lennyspodcast.com.

**Lenny Rachitsky** (01:28:46):
See you in the next episode.

---

## [13/15] Superhuman's secret to success: Ignoring most customer feedback, manually onboarding every new user, obsessing over every detail, and positioning around a single attribute: speed | Rahul Vohra (CEO)

**Lenny Rachitsky** (00:00:00):
Let's talk about product market fit.

**Rahul Vohra** (00:00:01):
You have to deliberately not act on the feedback of many of your early users, and this is at the same time as listening to people intensely and building what people want. That's what we're here to do, is to make something that people want, but it can't be all people. And the question becomes, how do you listen to them? And then even of what they say, what do you pay attention to and what don't you? The trick here is-

**Lenny Rachitsky** (00:00:26):
You're not doing what a lot of CEOs think they need to be doing with their time. A lot of CEOs think they need to spend time on hiring or org building and you intentionally, "I will spend time on product and marketing design."

**Rahul Vohra** (00:00:36):
This is a technique that I call the switch lock. It's born out of the observation that your calendar says what you thought you were going to do, but it's really only your trail of work that describes what you actually did. How can we capture that? So I came up with the following idea. What if I just did whatever the heck I wanted?

**Lenny Rachitsky** (00:00:56):
What's the most pivotal moment in your career, in your life?

**Rahul Vohra** (00:00:58):
I learned the real secret behind virality. There is no such thing as a truly viral product. What then is the true secret? It is-

**Lenny Rachitsky** (00:01:12):
Today my guest is Rahul Vohra. Rahul is the founder and CEO of Superhuman and one of the most thoughtful and insightful and articulate founders that I've met. As you'll see in our conversation, it's hard not to be captivated by Rahul's storytelling skills and also his really insightful takes on how to build great products and teams.

**Lenny Rachitsky** (00:01:32):
This episode is for anyone who's looking to build their product taste, help their teams move faster, learn how to think better from first principles. And also learn about Superhuman's very unique approach to building their company, including why they manually onboarded every single new user for years and why they decided to stop. Why they ignored most of their customer feedback on their way to finding product market fit, and also how you can use his approach to finding product market fit for your own company. Also, the power of game design in building great products, a very contrarian take on pricing strategy, what Rahul has learned about building scaled products on top of AI and LLMs and so much more.

**Rahul Vohra** (00:05:07):
Hello, hello and thank you for having me Lenny.

**Lenny Rachitsky** (00:05:10):
I have so many questions for you. We're going to have so much to talk about. I actually want to start with your time before Superhuman. When I was preparing for this chat, I actually asked you, what's the most pivotal moment in your career in your life? And you told me that other than starting Superhuman, it was selling your previous company Rapportive to LinkedIn. So let me just start there. What was that experience like? What do people not know about this phase in your life and just why was it so pivotal?

**Rahul Vohra** (00:05:37):
So for folks that don't know, Rapportive was my last company. It was the first Gmail extension to scale to millions of users. Basically on the right-hand side of Gmail, we would show you what people look like, where they work, links to their recent tweets, their LinkedIn profile and everything else that they were doing online. So if you were hiring, marketing, selling in BD, super useful. It turns out we somehow attracted most of LinkedIn's daily active users onto this one free app, and I then ultimately ended up selling that to LinkedIn. That by far, as you said, was the most pivotal thing I'd done in my career, prior to starting Superhuman.

**Rahul Vohra** (00:06:22):
Now, had I known that we'd amassed most of LinkedIn's active users onto one app, I would have sold it for far more. But the actual pivotal moment was really who I got to work with. Because I reported to LinkedIn's head of Growth, Elliot Shmukler. He was responsible for scaling LinkedIn from 25 million members to when I joined, north of 250 million members. And during my first one-on-one, I learned the real secret behind virality and big hint, it's not about viral mechanics. Overall that acquisition experience gave me the time to figure out what was next and the resources to truly swing for the fences.

**Lenny Rachitsky** (00:07:02):
Okay. Well, I have to follow this thread that you put out there, what the secret is to virality. What did you learn there?

**Rahul Vohra** (00:07:08):
Well, in my first one-on-one, I sat down with Elliot and I said, "Hey, I'm here to learn. Please teach me everything that you know about virality." And he said, "Okay. Well, hate to burst your bubble, but there is no such thing as a truly viral product." I said, "What do you mean? How do you explain Facebook for that matter? How do you explain LinkedIn?" And he said, "What I mean is, no app has sustained a viral factor of greater than one for any real period of time." Even Facebook in its heyday had a viral factor of about 0.7. And he told me that lasted for perhaps a year, so one person was creating about 0.7 new users.

**Rahul Vohra** (00:07:57):
I double-clicked again and I said, "Well Elliot, what about the address book import?" This is one of the things that LinkedIn got famous or infamous for. You could import your address book and then it would spam slash invite everyone who happens to be members of LinkedIn in your address book, and then eventually it would just invite everyone to LinkedIn. And he said, "That's an amazing feature, but you have to remember not everyone is going to use it all the time." So even that feature had a lifetime viral factor of about 0.4, and that's considered good.

**Rahul Vohra** (00:08:28):
 0.4 is good for a viral feature, 0.6 is great, something like 0.7 is absolutely incredible. You're in the stratosphere up with Facebook at that time. So I said, "Well, okay, all of these things by definition are going to Peter out. There's going to be an asymptote. None of these viral mechanics keep on compounding. Which actually makes sense, it would be a little absurd if things just kept on growing. What then is the true secret behind virality?" And he said, "It is word of mouth. It is the virality you can't measure that isn't a mechanic that isn't in a feature. It is when one user spontaneously tells another user about your product." That really colored how I think about growth and virality. Since then, it has shaped so much of what we do at Superhuman and so much of how I think about growing brands.

**Lenny Rachitsky** (00:09:23):
Wow, you're such a great storyteller. I'm just listening here, just captivated, "What is he going to say next?" That was fascinating. I actually have a post that I'm going to link to that, that very much aligns with what you're talking about, which is titled, Virality is a Myth mostly. It's based, I forget, on this book where they do all this research on actual viruses. It turns out they're not actually spreading in this exponential way, there's one person that spreads it to a lot of people and it keeps happening.

**Lenny Rachitsky** (00:09:52):
That's actually apparently what the data shows. I'm curious if you found this same thing, which is, when people think of an app as going viral, it's one person with a massive platform sharing it and their audience adopts it and that's just one to many and then it just happens a couple of times and it looks like it's going viral, but it's a person to many people, not many people to many people. Thoughts on that?

**Rahul Vohra** (00:10:14):
Yeah, we've definitely found that there are whales, to use the gaming terminology, that one person is going to be responsible for inviting 25, 50, 100 people, and they may have various motivations for doing that. In Superhuman, as an individual subscriber, if you refer somebody else and they sign up, you both get a free month, which is a great incentive if you're paying out of pocket.

**Rahul Vohra** (00:10:39):
We have people who send many, many hundreds of invites and there are some people who essentially have free Superhuman for life now due to how many people they've invited. But of course that incentive doesn't necessarily work inside of a company or inside of a team where ultimately it's the company paying for the product, so you have to then come up with new motivations for those people. That's where there really isn't any substitute to having a genuinely multiplayer or a genuinely collaborative product. That's one of the huge evolutions we've taken Superhuman through over the last, probably about two years.

**Rahul Vohra** (00:11:15):
Early last year we launched what we call Superhuman 2.0. The basic idea is, we saw almost every single other app of note become collaborative by default, Figma, Notion, Loom. These are all multiplayer or collaborative by default. Yet email, the one tool that we all use more than anything else, even more than things like Slack, was still firmly stuck in its single-player origins.

**Lenny Rachitsky** (00:11:43):
I want to come back to something that you mentioned that I didn't come back to you that I think is really core to what you just shared, which is word of mouth being so important. People talk about all these viral features and sharing contact books and all these things. And your point is, that takes you to a place, but really what helps a consumer-ish product spread is word of mouth, people sharing with each other. Which, then the question is, how do you do that? We're going to talk about a lot of things that you did to make Superhuman something people want to share, but in the end it's just making something people want to share. That's the definition almost. Then it's like, what makes people want to share stuff? It's amazing, it's helping them, something that is remarkable.

**Rahul Vohra** (00:12:22):
Well, it turns out, because you mentioned remarkableness, that is one of our core company values. If you think about what a company has to do, it has to grow. How do things grow? Well, let's take Elliot's advice at face value, and I believe it's true, it's creating something that people share. You mentioned one way of doing it, which is something that people want to share. There's actually another way, which is simply creating something remarkable, and you used that word, and that is one of the core values of Superhuman.

**Rahul Vohra** (00:12:53):
We have, create delight, create something that is so joyful that really truly brings people delight. We have deliver remarkable quality, something that is so striking, so compelling and worthy of attention that people can't but help tell others about it. Then we have build the extraordinary, which is a measure of the efficacy or the innovativeness of what we want to build. That's another trick, which is literally baking these raw ingredients for growth into your company values.

**Lenny Rachitsky** (00:13:23):
I didn't know that was one of your values. That makes so much sense. Okay, we're going to come back to that, because I think that is... There's so much to learn about how you think about product and how you think about building the company that builds the product. But I want to actually start here with how this conversation came to be.

**Lenny Rachitsky** (00:13:41):
The CEO of Product Hunt, Rajiv, tweeted months ago, he tweeted this and we're going to show this if you're on YouTube, "Superhuman's product velocity feels like it's kicked into another gear as of late. Does anyone else notice this?" I saw that, I'm like, "I completely have noticed this. It feels like there's just feature shipping left and right, AI this, AI that. It feels like it's just a new company." And I tagged you on the tweet. I'm like, "Hey Rahul, what's changed?" And you answered with a few things and it just made it clear there's a lot to learn about what you did.

**Lenny Rachitsky** (00:14:12):
Because a lot of companies are in this phase of just, "Things aren't moving as fast as we want. We used to be so much faster, we used to ship all these features and now we don't." So I think this is a really cool real case study illustrative example that we can analyze. So let me ask you this, what did you notice that told you something needed to change at Superhuman? And then what did you change that actually had the most impact on your ability to ship and move faster?

**Rahul Vohra** (00:14:37):
I think what we noticed was this sentiment, and we felt it first ourselves, but we also started hearing it from the market, from our users, from our customers, that we'd slowed down. And as a founder, as a CEO, that's the absolute last thing you want to hear. It's our job after all to speed things up. When I ask people what do they mean by slowing down, they didn't mean the product, of course, the product wasn't working any slower, but that the pace of delivery seemed to have slowed down.

**Rahul Vohra** (00:15:09):
I think to break this down, it's important to start by defining what we mean by a slowdown. There's the kind of slowdown that is unavoidable in certain spaces, and then there is the kind of slowdown that is quite avoidable. We actually had both. So starting with unavoidable slowdown, you can classify anything that you build in a company into one of two categories, solution deepening and market widening. Now, solution deepening means making your product better for its existing users, but not making it available to more users. Whereas market widening means making your product available to more users, but not making the product itself any better.

**Rahul Vohra** (00:15:50):
There are some spaces, there are some markets, there are some platforms where market widening is really fast and really easy, and there are some spaces, email is one of them where market widening is really hard and really slow. But when we started we had a great deal of focus. We only supported Gmail, we were only on the web. In those early years, we could pour every ounce of R&D energy, every engineering dollar, into solution deepening, making the product better for existing users. And of course users loved it. It's how we got to product market fit. It's how most startups do.

**Rahul Vohra** (00:16:24):
But at a certain point, almost every company then has to start investing in widening the market. For example, the market of people who will use a new Gmail front end but without a mobile app, does exist, but it is relatively small. This is something that every new email startup is going to learn sooner or later. In order to keep on growing, you are going to have to need to add an iOS app and then a MacOS app, and then a Windows app, and then an Android app. Then you'll soon want to support Office 365. But that's not one thing, that's actually three things, because you have to support Office 365 on desktop and then on iOS and then on Android. That's all much easier said than done.

**Rahul Vohra** (00:17:04):
I think we at Superhuman now know things about these APIs that literally no other company knows, and I would not wish it upon my worst enemy. So fast-forward to today, and Superhuman now works wherever you do on every combination of Gmail, Outlook, Mac, Windows, Web, iOS, Android, and this actually turns out to be a really great technology moat. Almost no other email app can claim this. It's taken many years of intense investment. I think we'll touch on this later, but it's one of the main reasons why we can sell into the enterprise, because we now know everyone can use it.

**Rahul Vohra** (00:17:37):
But this is the hard part, when you're doing that market widening, you're not solution deepening, so your perceived product velocity may decrease. You can avoid some of these things with some smart technology decisions, but mostly you just have to grind through it, and it is worth it to get to the other side. Then there's the kind of slowdown that is avoidable. If I remember my answer to Rajiv's tweet, that was the kind I was talking about. In that case it was our management structure, or who does what.

**Rahul Vohra** (00:18:07):
When we hired our initial executive teams, I followed very conventional wisdom. I ended up with a set of VPs and eight, I think direct reports, maybe even nine. I thought that's what you were meant to do. That's how startups are meant to scale. But as anyone who's been there knows, eight direct reports is a lot. It's a lot of hiring, it's a lot of goal setting, it's a lot of OKRs, it's a lot of accountability conversations, and fortunately also it's a lot of firing. No CEO ever gets their executive team right on the first try. That time I had for the things that I think I can genuinely be world-class at things like product and design and technology and marketing, that all began to rapidly disappear, and as a result the organization began to slow down.

**Rahul Vohra** (00:18:54):
Unfortunately, I was also tracking my time very closely, I had this crazy way of tracking it. At one point I noticed I was spending six to 7% of my week on these areas, these areas where I can truly be world-class at. So I had two realizations. Number one, as CEO, once you get to a certain scale, and we were definitely at that scale, you can actually define what you want the role of a CEO to be at your company. And number two, the Superhuman opportunity deserves everyone who works at the company to spend as much time as possible in their zone of genius, so that includes me as well as everybody else. What I did is, I hired a really great president, I went from eight direct reports to two, and the amount of time that I spend on product design, technology and marketing went up from six to 7% to about 60% to 70% of my week.

**Lenny Rachitsky** (00:19:49):
Just to mirror back a few things. One is, people may feel like you are not shipping as much as you used to because you're actually building things they don't care about, which is support for office and all these things that they don't need, but the business needs to expand, integrations with Microsoft and Android and all these things. I think that's such a good point, that it looks like nothing's happening when there's a lot of good stuff happening for other users that aren't you.

**Lenny Rachitsky** (00:20:15):
Then there's this point about people delegate. Then a leader of delegates, hires all these execs and they're like, "This is not what I wanted. Why have I done this?" And you think it's going to speed up, but it slows down. A couple threads here that are really interesting. One is this time tracking thing, I need to know how do you do this? The fact that you knew seven to 8% or whatever the number is, to that granularity of your time you're spending on things you wanted was that low, how do you do time tracking? Let's not go super far, but just what's your approach?

**Rahul Vohra** (00:20:48):
This is a technique that I call the Switch log. It's born out of the observation that your calendar says what you thought you were going to do, but it's really only your trail of work that describes what you actually did. So how can we capture that? And actually, how can we create a system of work that isn't tethered to a calendar, where you aren't at the behest of what some timetable says you do or you don't have to do? So I came up with the following idea, what if I just did whatever the heck I wanted? What if every single time I change task I just Slack DM'd my EA, but this also works in Slackbot, it just has to go somewhere. I Slack DM'd my EA and I said, "TS:," and then a few words for the task I was doing.

**Rahul Vohra** (00:21:45):
Well, that would create certain changes. Instead of having to constantly look at the calendar and think, "Oh, should I stop this task, start that task, I can just do what I want." If what I feel right now is, "Oh boy, I really need to prepare for Lenny's podcast, I'll go ahead and do that." And if I get bored or distracted eight minutes in, which sometimes happens because something else just bubbles up to the top of my mind, well, there's a reason that my body is bubbling it up to the top of my mind. I also practice transcendental meditation, so I'm very keen on the idea of being aware and listening to what's bubbling up.

So it's okay for me to then go and attend to that thought as opposed to start to expend my focus points or my discipline or willpower on the thing that I thought I was meant to be doing. All I'd have to do is I'd go back to Slack, "TS: Dealing with this other thing." And by the way, you should obviously turn up for your meetings. I'm not saying just blow through your meetings and not turn up for your one-on-ones. Definitely do those things. What I'm saying is, do what feels right for as long as it feels right to do. Then at the end of the week you can see where your time is going.

**Rahul Vohra** (00:22:55):
I realized at one point that I was spending only in those days 5% of my time on recruiting, whereas perhaps I should be spending 20 or 30% or more of my time on recruiting. But the biggest thing was, I saw I was only spending six to 7% of my time on product, on design, on technology and marketing. These are things where I know I'm really good at them. I should either be teaching people how to do them or doing them or some combination of both. That's probably the best thing for me. It keeps me really happy, very joyful, it keeps me sharp, but it's also scaling the organization. So that's how we had that kind of an insight. Once you have this Slack Log, you can then graph it and chart it and see where your time is actually going.

**Lenny Rachitsky** (00:23:37):
How cool. Clearly this is an app opportunity or an agent opportunity where you're just telling this thing every time. It's essentially tracking context, which we're always hearing, try not to avoid context which switches.

**Rahul Vohra** (00:23:50):
I think context switches are fine. There's definitely this idea that, for every interruption you have, the brain does take roughly 21 minutes on average to recover, to get back to the efficacy before that you were disturbed. It's a big deal, of course, I'm building productivity software, we designed Superhuman to minimize the amount of distraction and disruption that's possible within the app. But if you are working on something and at the back of your mind something bubbles up, you have to attend to it in one way or the other. Sometimes I just write it down, actually, I don't have my notebook with me, but it's really big. I have a gigantic, whatever twice the size of A4 is, I guess A3 sketchbook and I always have a 4H pencil, so whenever one of those thoughts comes up, I just scribble it down. Or I actually stop what I'm doing and I attend to that task, because there's a reason it's bubbling up right now.

**Lenny Rachitsky** (00:24:44):
I love that you know exactly the type of paper and pencil, 4H pencil, A3 paper, [inaudible 00:24:51]. Okay, this is going to be a theme. You mentioned meditation, you said you do TM, so you do 20 minutes in the morning, 20 minutes... Do you do it that style or you do a longer session?

**Rahul Vohra** (00:25:01):
I do about half an hour in the morning, including rest time. The physical rest component of it is very important to me. So it's 20 minutes of the actual meditation, then 10 minutes of rest. I do that in the morning as well as in the afternoon at around 3:00 PM

**Lenny Rachitsky** (00:25:13):
And you just carve that out in your calendar. Everyone knows Rahul at three o'clock, he's going to be out.

**Rahul Vohra** (00:25:17):
Absolutely. My EA knows, they're the one who's organizing the calendar and making sure things happen when they need to happen. They also know that nothing can override this TM block. Without it I genuinely start to fall apart. But with it, I'm able to access some very deep competencies that I didn't have before. I've been doing this now for about four or five years, and initially I simply felt happier, occasionally even more euphoric coming out of a really great meditation session. But over time I found that my ability to focus was increasing. I could hold attention on something for much longer, but I also was able to become much more creative and much more expressive.

**Rahul Vohra** (00:26:02):
These are well-known side effects, as it were, or intended effects for some people of TM. And interestingly about TM, if you compare it to other forms of meditation, they don't have quite the same impact across quite as many executive functions. So there's something particularly interesting that's going on with transcendental meditation as opposed to other forms that folks are still trying to unravel and figure out.

**Lenny Rachitsky** (00:26:26):
If folks want to, if they're inspired and they want to check out this form of meditation, any advice on where they could go learn?

**Rahul Vohra** (00:26:32):
Absolutely, a lot. But in summary, have a coach teach you. I had many false starts myself with meditation, trying the various apps, learning from books. None of it really worked for me. What worked was having one-on-one teaching from someone themselves who had been taught one-on-one the Yogic or the Raja tradition of teaching. This person in particular had also been a venture-backed founder multiple times over, so they're very well aware of the kinds of stresses that I tend to be under. And all of his clients are mostly in technology as well. If you're in the Bay Area, this person's name is Laurent Valasek. They run an institution called the Peak Leadership Institute. And this is all about how we can live a more integrated and whole life. Integrating wellness practices like meditation, but for the purpose of unlocking peak performance in life and in business.

**Lenny Rachitsky** (00:27:31):
Thank you for sharing that. That is very actionable. We're going to link to that in the show notes.

**Lenny Rachitsky** (00:27:35):
Okay. I'm going to try to bring us back on course. The other thing you mentioned that I think is really interesting is hiring a president. A lot of founders and leaders might be hearing this and be like, "Going from eight reports and doing all these things I don't want to, spending most of my time on the product and design and marketing, amazing." What did this president take off your plate and what is their responsibility and that allowed you to do the stuff you wanted to do?

**Rahul Vohra** (00:27:58):
The biggest thing was taking off the operations and the management of the executive team and the rest of the company. Think of the president role in Superhuman as an operationally extremely challenging and a very growthful role. It is perfect for someone who wants to go on to be a CEO in their next role. Instead of hiring and firing that team, instead of managing and setting their goals, instead of the accountability conversations, someone else who's now doing that.

**Rahul Vohra** (00:28:35):
In addition, because that's not the only job, in addition, they're also a very strong thought partner when it comes to corporate strategy. When it comes to, where do we take act one, our email product? How far do we go down the multiplayer path? How aggressively should we lean into AI? What's a reasonable gross margin in a world with AI? Are we from a financial perspective okay dipping now and then coming back later? When should we start building our second product? How do we think about our R&D strategy? Should we keep on hiring in the Bay Area, or as we've done for many of our recent hires, should we continue hiring in Latin America? Should we consider other time zones as well? And so on and so on and so on. I'm just randomly coming up with questions, but the list is truly endless.

**Rahul Vohra** (00:29:27):
Another way to think about it is, it's almost like a grown-up co-founder. The two people I co-founded the company with, Comrade and Vivek, they've long since gone from Superhuman. We're now a 10-year-old organization and I'm one those rare founders that is persisting and thriving actually 10 years in. That said, the journey never gets easier, it gets different and you still need that co-founding energy around you. I have a handful of people in the organization who are in their roles providing that kind of energy, that kind of input, and who thrive off doing so. Then the president role is definitely one of them.

**Lenny Rachitsky** (00:30:07):
Incredibly interesting. There's so much there. One, just a couple of things I'll share and then I want to move on to a different topic. One is just, it's cool the solution to helping you move faster and do the work you want to do is org design. That feels like a really doable thing. If you're finding you're not spending time on things you want to spend time on and things aren't moving as fast as you want, it's essentially you can find people to take on things that you don't want and shift the way that the org is structured and that could solve a lot of problems. That's what it did for you. Then I think it's also really interesting, there's this lesson here of as a founder, if you're just feeling depleted or just don't have the partner you want, you could bring someone on that could be that person.

**Rahul Vohra** (00:30:50):
Absolutely.

**Lenny Rachitsky** (00:30:52):
Okay. There's so much there. That was much more of a rich area than I even expected. I want to zoom out a little bit, and there's a couple themes that came up again and again when I talked to folks that you've worked with, investors in Superhuman. The two themes are contrarian thinking, in terms of building the company, and strong attention to detail. Let's spend a little time on attention to detail. Like I said, this is one of the things that came up again and again when I was asking people about you. So I have this quote from Ed Sims, and maybe your first investor. Were they your first investor?

**Rahul Vohra** (00:31:27):
Yeah, that there's a bunch of people on Twitter who are going to fight for that. But to set the record straight, Ed Sim did actually write the first three checks into Superhuman.

**Lenny Rachitsky** (00:31:35):
First three checks? At subsequent rounds.

**Rahul Vohra** (00:31:38):
Well, yeah. Quick sidebar on that, he runs Boldstart Ventures alongside his partner Elliot Durbin. They have a particular interest in backing second-time founders, but they'll also back first-time founders, and they love application and infrastructure areas like Superhuman, so we were like the perfect investment. He also wrote a check from his previous fund into a Rapportive, and I think I'd made him five X that money. Nothing to write home about, but definitely, "I'm going to back this guy again." So I went to him and I said, "Hey listen, this is going to sound crazy. I want to take on Gmail." He said, "Do you have a deck?" I was like, "Yeah, here it is one slide, here it is." And there was a screenshot of Gmail with most of it scribbled out, "I want to build that and it's going to be amazing."

**Rahul Vohra** (00:32:25):
So he said, "Cool, we're in. Can I wire you the money?" And I said, "No, I don't even have a bank account yet." I come back two days later with a bank account and he's like, "Cool, I want to wire you 750 K." And I said, "I don't even know what I'm going to do with that money. I'm not paying myself, I won't for a while. We don't have any employees. I can't think of anything I want to spend it on. Tell you what, I'll just take 250 K." And he was like, "What?" I'm like, "Yeah, I'll just take 250 K." We start having the conversation around venture economics. I'm like, "Yeah, it's fine, we'll figure it out." Then a few months back I took another 250 K and a few months back I took another 250 K as I began inventing ways and finding channels to deploy capital properly.

**Lenny Rachitsky** (00:33:11):
I love this story. I love all these stories you're sharing I've never heard before. And by the way, it is awesome. We're talking about him coming on the podcast, maybe breaking our VC rule. So specifically the story he shared with me that is maybe an example of you and your attention to detail is, he said that you created your own font because existing fonts weren't good enough. Is that true?

**Rahul Vohra** (00:33:31):
Kind of. Okay. The font that we use today is a modified version of Adelle Sans. The story there is, I looked at all of the major font families, and honestly none of them was what I would call truly excellent. That may sound like an odd thing to say. So let's, if you will permit me to talk about typography and email-

**Lenny Rachitsky** (00:33:55):
Please.

**Rahul Vohra** (00:33:56):
The first thing we did was, we took our UI and we laid it out in about 15 different styles using examples of the major font families. We actually printed these out and we left them on a desk in the middle of our office. Sometimes with design, you want to tune in to your immediate most visceral response, but sometimes you want to truly let a design marinate. And this was the latter. So we let these designs marinate, we let these font choices percolate. Like I said, none of them was truly excellent.

**Rahul Vohra** (00:34:31):
Number one, I was looking for a font that was in and of itself gorgeous. Number two, I was looking for a font that could also convey a message of any kind, without overpowering the sentiment of that message. For example, does the font work when this is inviting you to a party? Many fonts, including almost all serif fonts, are actually too somber or too sober for that. Or to pick another extreme, does the font work if it is informing you of somebody's passing, many fonts are just too jaunty for that. You wouldn't want that kind of message in Comic Sans, for example. And number three, I was optimizing for a font that made reading speed and comprehension really fast. And number four, I was looking for a font that made email addresses themselves look great. So I discarded all the 15 because they weren't good enough, and after searching high and low, I came across a font called Adelle Sans, which is designed by a foundry called Type Together, type-together.com. They have a whole bunch of lovely fonts, go check them out.

**Rahul Vohra** (00:35:36):
And if you go through my list, number one, Adelle Sans is gorgeous. I think each character is a work of art. It's beautifully formed. Number two, Adelle Sans is, I would say upbeat, it's optimistic, yet it's serious enough to convey any kind of message. It has just the right amount of personality, yet not too much personality. Number three, Adelle Sans is also unusually narrow, and that actually fits email particularly well. One of my pet peeves with Gmail, which by default uses Ariel, is that the lines are as wide as your window. So if you're in a wide screen, then the lines get really arbitrarily long. The problem with really wide and really long lines, is that they decrease reading speed. Because by the time you've reached the end of one line, your eyes have lost track of the start of the next line. And Ariel itself has fairly wide characters, which further exacerbates that.

**Rahul Vohra** (00:36:30):
So at Superhuman we, if you've used the product, you know this, we fix the line length or the typographical measure to the optimal length for reading speed, which depending on the font is around 90 to 120 characters. And Adelle Sans is quite narrow, so it actually lets us do this on quite small windows with fairly dense line. So we get a lot of information on fairly small windows without getting a very long typographical measure, optimizing for reading speed and for comprehension. Then number four, finally Adelle Sans has very unusual treatment of the at symbol in an email address. It actually puts the base of the A in the at on the same baseline as the rest of the text.

**Rahul Vohra** (00:37:15):
So for example, if your name has an A, my name does Rahul at Vohra, three A's and or two A's and an at, they're all actually on the same baseline. It's a small thing, but it makes the email addresses look incredibly natural. If you look at that and then you actually look at email addresses laid out in other fonts, those other the fonts look really clunky and awkward because the A is kind of shifted around and it just looks a bit silly in my opinion. Now Adelle Sans isn't perfect. So we then worked with a type designer on some of the specific details that there are some of the glyphs, which get a little pinchy as it were, and what we use today is very close to retail Adelle Sans.

**Lenny Rachitsky** (00:37:55):
And this was pre-launch or this was after you'd already launched?

**Rahul Vohra** (00:37:58):
We'd probably had about 10, 15 users at the time.

**Lenny Rachitsky** (00:38:03):
So I think that's pretty contrarian unique to be this focused on the font and the typeface before you even launched. This was like, "Is this even going to be a thing? Will anyone even care?" And I think this says a lot about the way you think about product.

**Rahul Vohra** (00:38:17):
Oh yeah, that thought never crossed my mind. I think we'll probably come to it later, but the idea that, is this never going to be a thing? I think that's a dangerous thought. We can't start thinking that way, because at what point do you stop second-guessing yourself?

**Lenny Rachitsky** (00:38:35):
Interesting. So you were confident this was going to work, so because I am so confident it'll work, I need them to get this right. There's also this trap founders fall into of just spending too much time perfecting a thing that never works and there's always advice launch early, launch often. Thoughts there? How do you find that balance? What's your advice there?

**Rahul Vohra** (00:38:57):
How much to spend time ahead of launch really does depend on the markets and the structure, the nature of your business model. For example, let's say you are building a marketplace in a greenfield opportunity, so imagine the Lyft or Uber in their heyday. There's a strong network effect, because the more cars you have on your platform, the shorter waiting times are, therefore people are going to preferentially use your app versus the other person's app. That's when there's no time to spare, that's when you probably shouldn't even be sleeping. You're going to hire the most aggressive maniacal people possible. You're going to work 120-hour weeks, because every marginal minute actually does matter. Every marginal minute in the market, growing compounding is going to make your next year even better.

**Rahul Vohra** (00:39:51):
That's actually not true of all startups and it certainly isn't true of something like Superhuman. Yes, working harder is always better and we work tremendously hard at Superhuman, but not to the point where it made sense to release something that didn't work. I'm reminded of a story of a founder that was in Y Combinator, told me about their demo day experience. They used Mailbox, which some folks may remember was also a startup, and Dropbox famously acquired them for about a hundred million dollars. The reason that they were well known, apart from the acquisition, is they were the first to popularize, swipe to archive or swipe to mark down, which of course is now standard in Superhuman and every other app.

**Rahul Vohra** (00:40:43):
This founder was using Mailbox and was having an amazing demo day. They're working the room, they're meeting investors, they're pitching their photography app in this case. He went home that night and went to his laptop, fired up mailbox and sent off a bunch of follow-up emails. He waited the day, didn't hear back, he waited two days, didn't hear back. On the third morning he figured something was up, so he fired up Gmail, went to his sent mail, and you guessed it, there were no sent mails there. So something had broken with mailbox. So he's cursing to himself trying to remind himself everything's going to be okay. Sent all the same emails from Gmail manually and they all came through.

**Rahul Vohra** (00:41:37):
But then one of the investors said, "Hey, by the way, you might want to check your email clients, because I've been getting some of your emails twice." Now he goes back into his Gmail, he sees that yes, actually the original emails that were queued up in mailbox have now indeed been sent, and some of the investors, and unfortunately most of the investors he actually pitched twice. Now, is this the end of the world? No, an investor can overlook that. Probably a good thing that you're trying new apps. But was it horrifying and was it really scary? Absolutely.

**Rahul Vohra** (00:42:08):
Imagine this wasn't investors, imagine this was a customer, someone who you were trying to convince to buy your thing and that you knew what you were doing and you had attention to detail and you had everything just buttoned up and under control. Well, now you've lost face, now you look foolish. That's why when you have mission-critical products like email where you are interfacing with customers, with candidates, with investors, it turns out to really matter. Email is mission-critical. It's not something where you can simply launch with a half-baked product.

**Lenny Rachitsky** (00:42:40):
This is such an important nuance take on, there's always this debate, how much to focus on craft and user experience, how much to focus on time to launch and get it out and speed. What I'm hearing here, which I completely agree with is, it depends on the market you're in and the criticality essentially of your product. So if it's email, it just needs to work and you need to get that right, you need to spend all the time, you need to get that right.

**Lenny Rachitsky** (00:43:03):
This reminds me of something else that when your early investors shared with me, Bill Trenchard from First-Round Capital. He talked about how speed was the thing that you just dialed up as a lever to 11. That's where you just, "We will make this the focus. Speed, speed, speed." I think maybe the lesson there is, you pick the thing that you think will most differentiate you, make you significantly better than what's out there. So just thoughts on how you decided speed was the thing you were going to obsess with, and advice for folks that are trying to decide where to dial up things to 11?

**Rahul Vohra** (00:43:37):
Bill is right and I agree with him, you have to pick something. Knowing what to pick is the trick. In the early days of Superhuman, I read a book on positioning that really influenced my thinking. It is, I believe called Positioning the Battle for Your Mind. It struck me how the most well-known brands have stood for one clear thing, they have a clear position. So in order for Superhuman to be memorable, I believed that we needed to occupy a clear position that was unique and which was available and which reinforced our product strategy.

**Rahul Vohra** (00:44:12):
In the first year of Superhuman, therefore, I interviewed hundreds of potential customers about their experience with Gmail and with Outlook. And predictably, almost everybody says that email takes way too much time. But interestingly, many people also said that Gmail and Outlook were way too slow. That was how I first thought that speed could be an interesting position for us. I then asked myself, "Is the position of speed unique and is it available?" And the answer was overwhelmingly yes, because almost no software was being sold or has ever been sold on the value proposition of speed. The last time I could remember anyone trying to do this, was when Google launched Chrome, and obviously that went incredibly well for them. You may remember they had slow-motion videos where they were comparing Chrome render webpages and showing that was faster than an actual strike of lightning. No one had done it since then.

**Rahul Vohra** (00:45:15):
I then asked, "Well, does speed reinforce our product strategy?" And again, the answer was overwhelmingly yes. I knew that our competition was not going to be startups, it was incumbents. And I also knew that incumbents generally struggle with speed, because by definition they have massive scale and usually entrenched architecture. Then finally I did what I call the cocktail party test, which is to look at the cocktail parties and to watch how people pitch your product to other people. In our case the pitches were simple. People would say, "Dude, you have to use it, it's really fucking fast." And that's it. That was the pitch. That's how I knew that speed would be a really great position for us to start with.

**Lenny Rachitsky** (00:46:01):
I'm excited to chat with Christina Gilbert, the founder of OneSchema, one of our longtime podcast sponsors. Hi Christina.

**Christina Gilbert** (00:46:08):
Yes, thank you for having me on, Lenny.

**Lenny Rachitsky** (00:46:10):
What is the latest with OneSchema? I know you now work with some of my favorite companies like Ramp, Vanta, Scale and Watershed. I heard that you just launched a new product to help product teams import CSVs from especially tricky systems like ERPs.

**Christina Gilbert** (00:46:24):
Yes, so we just launched OneSchema of FileFeeds, which allows you to build an integration with any system in 15 minutes, as long as you can export a CSV to an SFTP folder. We see our customers all the time getting stuck with hacks and workarounds, and the product teams that we work with don't have to turn down prospects because their systems are too hard to integrate with. We allow our customers to offer thousands of integrations without involving their engineering team at all.

**Lenny Rachitsky** (00:46:47):
I can tell you that if my team had to build integrations like this, how nice would it be to be able to take this off my roadmap and instead use something like OneSchema. Not just to build it but also to maintain it forever.

**Christina Gilbert** (00:46:59):
Absolutely, Lenny. We've heard so many horror stories of multi-day outages from even just a handful of bad records. We are laser focused on integration reliability to help teams end all of those distractions that come up with integrations. We have a built-in validation layer that stops any bad data from entering your system and OneSchema will notify your team immediately of any data that looks incorrect.

**Lenny Rachitsky** (00:47:19):
I know that importing incorrect data can cause all kinds of pain for your customers and quickly lose their trust. Christina, thank you for joining us and if you want to learn more, head on over to OneSchema.co. That's one OneSchema.co.

**Lenny Rachitsky** (00:47:33):
The next area I want to spend time on and I imagine we'll have much insight is, some of the contrarian ways you approach building Superhuman that a lot of companies never thought about doing that you did that worked out for you. So the first is manually onboarding every single new user. Sure, startups have done this, founders bring on some folks and then cool, show it to them and then they stop doing that and then it's self-service or sales teams. How far did you scale this manual onboarding phase of your company? How many people did you have onboarding people, how many people did you manually onboard?

**Rahul Vohra** (00:48:12):
So for folks that don't know, in those early days we insisted on one-to-one concierge onboarding, and it was absolutely the right thing to do. You couldn't use Superhuman unless you went through the onboarding experience. Now it's almost the reverse. Almost every new Superhuman customer goes through self-service. The onboarding experience is still there, but again it is absolutely the right thing to do. To answer your question, at peak we had about 20 people doing manual onboarding.

**Lenny Rachitsky** (00:48:40):
Okay, so it's not that many people. That's really interesting. Because I always imagined it was like a massive team, but 20 people can handle a lot, is the takeaway there. What was the scale where you stopped manual onboarding, just for folks that are thinking about doing this and then when to stop?

**Rahul Vohra** (00:48:55):
I think the reason to stop is that there will always be certain personality types who do not want to go through a one-on-one onboarding. At a certain point those people will become very important, and you'll need to be ready with a world-class self-service option. When we started building self-service, it seemed nearly impossible. In fact, it was terrifying, because it's difficult to overstate how much the entire DNA of the company was built around this idea that we would onboard users manually. After all, we did so much in our one-to-one onboardings and there's only so much that software can do. Now, we did after a lot of grind and persistence eventually figure it out and we have a world-class self-service experience today, but we did not at the time.

**Rahul Vohra** (00:49:49):
So the flip side is, why would you even do this to begin with? What we found is two things. Number one, the user metrics are excellent for things like engagement, retention, product market fit score, MPS, virality, for all of those metrics. I think you you'll significantly beat your industry benchmarks if you go to the effort of one-on-one onboarding your early customers. It becomes so powerful to have that early cohort of super fans when it comes to things like building a brand. If folks remember that conversation from way up at the top, what is it that creates true virality? It's not viral mechanics, it's word of mouth. It is brand. This is how you can kickstart a brand.

**Rahul Vohra** (00:50:32):
And number two, in a world where you can easily and quickly raise funding, like for example the zero interest rate phenomenon era, you can actually use dollars to avoid building a first-time user experience and all of the normal growth loops that you would then have to build. You would then instead focus all of your engineers on finding product market fit or in solution deepening or in market widening, but not for example on a first-time user experience, not for example on activation, because you have humans doing activation for you. By contrast I saw other companies often competing spend almost half their engineering dollars on those things, on self-service flows for products that ultimately did not find products market fit. So makes sense to do if you really want to create that brand, which I think all consumer-ish companies need to do. And if there is money falling off trees, for whatever reason, which we did have for a period of time, arguably AI companies have that again today. So if you can weave this into your strategy, I think you should, but you should also know when to stop.

**Lenny Rachitsky** (00:51:40):
Super interesting. I guess some factors to think about, because I wanted to ask you when should people consider doing this? If they're hearing this and they're like, "This is awesome, so many problems solved if I just have somebody onboarding every new user, everyone's activated. Amazing." So some of the variables you're sharing is, do you have like cheap cash to invest in say, it doesn't have to be 20 people, it could be a few people to start. Then if there's an LTV, ACV element of just are you going to make enough from a new customer? Imagine that's a variable. Is there anything else you think founders should think about?

**Rahul Vohra** (00:52:14):
Absolutely. You don't want to lose money doing this. We always made money doing onboarding to be clear, it's just that at a certain point the mass market, whether it for us it's enterprise or all of the prosumers in the world, you hit a top of funnel width, it needs to be wide enough where manually onboarding no longer makes sense.

**Lenny Rachitsky** (00:52:37):
Awesome. Okay, let's talk about product market fit. I know that everyone, when they think of Rahul, they think product market fit. You wrote this epic First Round post that described the way you guys approach product market fit. We're not going to spend a lot of time on describing it, because people can look it up. So let me just ask you this, what are a couple of things that you think people still don't understand about finding product market fit, getting to product market fit? Considering it's the most important thing you got to figure out as a founder. If you don't find something people want, nothing else matters. Anything there you want to share.

**Rahul Vohra** (00:53:12):
The core ideas are still weird enough that I'll start there. Which is number one, you can measure product market fit. Number two, you can optimize product market fit. Number three, you can systematically, even numerically increase product market fit. And number four, you can even have an algorithm write your roadmap for you, and that is a roadmap that is guaranteed to increase product market fit. Now, if that sounds crazy, I would be the first to admit it doesn't seem like that should be true, but go check out that post. I think it is still the most widely shared post on First Round Review, it's called How Superhuman Built an Engine to Find Product Market Fit, or just Google the Superhuman Product Market Fit Engine. And you'll see the algorithm laid out there fully explained and why it works.

**Rahul Vohra** (00:54:07):
I'd say the second thing is to get to product market fit, you have to deliberately not act on the feedback of many of your early users. This is at the same time as listening to people intensely and building what people want. That's what we're here to do, is to make something that people want. But it can't be all people. It can't be everybody. The question becomes, how do you listen to them? And then even of what they say, what do you pay attention to and what don't you? All of that's covered in the Product Market Fit Engine.

**Lenny Rachitsky** (00:54:45):
Okay, I got to follow this thought on algorithmically building your roadmap to increase product market fit. Talk about how one would do that.

**Rahul Vohra** (00:54:55):
Well, that's really the meat of the engine. Let's see if I can condense it here in a very easy to grok fashion. Let's assume for the sake of argument, that you can put a number on product market fit, and it turns out you can. Very simply, you're going to ask people, "How would you feel if you can no longer use this product?" You give them three responses. One of them is very disappointed, the other is somewhat disappointed, and the other is not disappointed. Very disappointed means, "I'd be devastated. I love this product. I need this product."

**Rahul Vohra** (00:55:30):
What Sean Ellis found, Sean Ellis, if you don't know him, is the guy who coined the term growth hacker, and he instrumented, benchmarked this initial question. What he found, is that the companies that struggled to grow almost always had less than 40%, very disappointed. Whereas the companies that grew the fastest almost always had more than 40%, very disappointed. And this question, this metric is way more predictive of success than something, for example, like net promoter score.

**Rahul Vohra** (00:56:03):
Okay, so far so easy. How do we make this number go up? Well, you want more people to be very disappointed without your product. The trick here is not to act too much on the feedback that the very disappointed people are giving you, because they already love your product. Also, not to act at all really on the feedback that the not disappointed people are giving, you because they're so far from loving your product that they're essentially a lost cause. But to focus on the segment of the somewhat disappointed people, they kind of love your product, but something, and I would wager something small, is holding them back.

**Rahul Vohra** (00:56:43):
You then divide them into two camps, the camp for whom the main benefit of your product resonates and the camp for whom it doesn't. What do I mean by that? Well, you go back to the people who really love your product and you basically ask them why? What is it about my products that you really love? In the early days of Superhuman, it would have been speed and keyboard shortcuts and the overall design aesthetic as well as the time that we were saving you. You then go back to the somewhat disappointed users, and in the Superhuman example, I would simply ask, "Wait, do you like Superhuman because of its speed or for something else?" And if it's something else, well, and this is hard to do, but politely disregard those people and their feedback. Because even if you built everything that they asked for, they're still pulling you in a different direction. And the thing that they like the most from your product isn't actually what the people who en mass love it the most for, is.

**Rahul Vohra** (00:57:38):
You have then articulated the subsegment of the subsegment that it makes sense to pay attention to, and there's another question in the engine to figure out what they don't like about the product. Now you have a list of things people love, you have a list of things people don't love, and you can work down that list to make the product market fit score go up. And basically at the start of every planning cycle, I advise spending half your time doubling down on what people really love and half your time systematically overcoming the objections of the somewhat disappointed users, but specifically those for whom the main benefit resonates.

**Lenny Rachitsky** (00:58:14):
That was an excellent summary. I know I said we wouldn't spend a ton of time here, but I'm really glad we did. That was really helpful. Let me ask you this, I know you used this initially in the early days, are you still operating in this way in some form?

**Rahul Vohra** (00:58:24):
We don't run the engine as is for Superhuman as a whole. There are enough subcomponents of Superhuman now that are almost individual products. For example, Superhuman for Sales, our multiplayer and collaboration features, how we think about the enterprise, AI is its whole thing, but we do sometimes run it on those individual pieces. For example, we'll ask a salesperson, the Product Market Fit Engine, as it relates to Superhuman for sales. As we think about starting new products, we would absolutely deploy the product market fit engine.

**Lenny Rachitsky** (00:58:59):
Awesome. The way you ask this question is an in-product interstitial sort of survey pop-up thing?

**Rahul Vohra** (00:59:04):
You can do it however you want. The way Sean initially benchmarked the number was via email surveys. I think email surveys work just fine. The key thing is, and this applies to any survey methodology, if you're going to change the method of surveying, all of your old numbers are invalidated. So it's just a new baseline going forwards.

**Lenny Rachitsky** (00:59:25):
Got it. We had Sean on the podcast and he describes this method in detail. So if folks want to explore the Sean Ellis test, listen to that podcast. We'll link to it.

**Lenny Rachitsky** (00:59:33):
Okay, next topic that I'm excited to get your take on, is game design versus gamification. This is one of the more unique ways you think about designing product. When people hear you talk about this, they think it's like, "Oh, gamification making things like games. Oh, it's Zynga, Farmville, I don't want to do that." But you actually have a really different perspective on why you need to think about game design as you design products. Talk about your insights there.

**Rahul Vohra** (00:59:58):
Well, I strongly believe that we should make business software like we make games, because when we make products like we make games, people find them fun. They tell their friends, they fall in love with them. It's another way actually of backing into where we open this conversation, which is you're making a brand, you are giving reason for word of mouth. It's actually an altogether different kind of product development. So how do we do this? Well, as you've said, it's not gamification, that doesn't work. Game design works, but game design is not gamification. It's not, for example, simply taking your product and adding points, levels, trophies or badges.

**Rahul Vohra** (01:00:40):
To understand why gamification does not work, we actually have to start with human motivation. There's a very interesting study from Stanford that demonstrates the difference perfectly. In the 1970s, these Stanford researchers recruited children who were aged three to four years old, and all of these kids were generally pre-interested in drawing. Some kids were told they would get a reward, a certificate with a gold seal and a ribbon. And some kids were not told about any reward and they did not even expect one or didn't know of one. Now each child was then invited into a separate room to draw for six minutes and afterwards they would either get the reward or not.

**Rahul Vohra** (01:01:21):
Over the next few days, the children were observed to see how much they would continue to draw by themselves. So the children with no reward, they spent 17% of their time drawing, but the children who expected a reward, sadly they only spent 8% of their time drawing. The very presence of a reward halved their motivation. So what's happening? What's happening here, is researchers differentiate intrinsic motivation and extrinsic motivation. With intrinsic motivation we do things because they are inherently interesting and satisfying, and with extrinsic motivation, we do things to earn rewards and to achieve external goals. That's the problem with rewards, is they just massively undermine intrinsic motivation. That's why gamification doesn't work. And when gamification does work, it's because the underlying experience was already designed like a game.

**Lenny Rachitsky** (01:02:19):
What makes something like a game? I know Superhuman is really good at this, of just your inbox zero quest that you're on. Just to make that a little more real, what is game design? What does that mean to you? What makes it feel like a game?

**Rahul Vohra** (01:02:32):
Well, maybe folks don't know this, but before I was a founder, you can probably tell, I was actually professionally a game designer. And as it turns out, there is no unifying theory of game design. To create games, what we need to do is draw upon the arts and the science of psychology, mathematics, storytelling, interaction design. And at Superhuman we've identified five key areas that we really care about, goals, emotions, toys, controls and flow. And across these we've identified many principles of game design. One example principle would be, make fun toys and then combine those into games.

**Rahul Vohra** (01:03:12):
A question I like to ask is, are toys the same as games? They do seem different. For example, we play with toys, but we play games. A ball is a toy, but football is a game. As it turns out, the best games are constructed out of toys. Why? Because then they are fun on both levels, the toy and the game itself. So for example, in Superhuman, one of our favorite toys is the time auto-completer. If you use Superhuman, this is the thing that appears when you hit H, when you snooze or set reminders on emails. You can type whatever you want, it can be gibberish and it does its best to understand you. For example, if you type in 2D, that becomes two days, 3H is three hours, one MO is one month. The time auto-completer is fun because it indulges your playful exploration.

In onboardings, it wasn't long before I saw people asking, "What can it do? Where does it break? How does it work? What happens if I keep on typing in a series of tens? Well, it turns out that's October the 10th at 10:10 PM. Well, how about a series of twos? Well, that's February the second, 2022 at 2:00 PM." Then you start trying more complex inputs like in a fortnight and a day, and that works, which is a pleasant surprise. And it's not long before you find more pleasant surprises like time zone math happens without you thinking about it. You can just type in 8:00 AM in Tokyo and it turns out that's 8:00 PM Eastern Time and you no longer have to do the time zone math.

**Rahul Vohra** (01:04:45):
Then most people were really delighted to find out that if you really want, you can snooze emails until never, i.e. you can literally type in never, and the email will never come back. It had like a little shrug emoji at the same time. Is this toy going to win awards? Nope. But is it fun actually, surprisingly yes. So what I would encourage people to do is, think about the features of their product. Do those features indulge, playful, exploration? Are they fun even without a goal? And do they elicit moments of pleasant surprise? If so, you have a toy and you can combine that with other toys and actually start to build a game.

**Lenny Rachitsky** (01:05:28):
If people were to listen to this segment of the podcast, they would never guess we're talking about B2B software and email, which I love. Let's talk about pricing strategy and your approach to pricing. Another very contrarian approach that you guys took where you charge $30 a month for email that was free, that people don't need to pay for anywhere. And it's worked and now a lot of companies are thinking of it this way. You've even raised your prices recently. What have you learned about pricing strategy that you think might be helpful to folks?

**Rahul Vohra** (01:05:58):
I always say the same thing when it comes to pricing, which is before you figure out pricing, you must first figure out positioning. Superhuman is the best email tool on the market. We fortunately have the metrics to show this. One of the cool things about selling an email tool, is you can compare the 30 days prior to using Superhuman to the 30 days after, or the year before to the year after. We do that obviously. We're able to show that people get through their email twice as fast with Superhuman, that they respond one to two days faster, and that they save four hours or more every single week. Because of that, we're very confident in saying that Superhuman is the best email tool on the market and that we're building it for high performing teams and high performing individuals. In other words, we serve the high end of the market.

Once you understand your positioning, you can then move on to pricing. And one of the best books on this is a book called Monetizing Innovation by Madhavan Ramanujam. And Madhavan covers a lot of ways to develop pricing. We used one of the easiest methods, which is the Van Westendorp Price Sensitivity [inaudible 01:07:08]. In the early years, we asked, I think it was around a hundred of our earliest users, the following four questions. Number one, at what price would you consider Superhuman to be so expensive that you would not consider buying it? Number two, at what price would you consider Superhuman to be priced so low that you'd be worried about its quality and you wouldn't buy it? At number three, what price would you consider Superhuman to be starting to get expensive, so that it's not out of the question, but you'd have to give some thought to buying it? And number four, at what price would you consider Superhuman to be a bargain? A great buy for the money?,

**Rahul Vohra** (01:07:45):
Now most startups orient around price point number four. This is especially true for greenfield opportunities, marketplaces, you've got to set the transaction value around price 0.4. Basically when you want as many people to sign up as is humanly possible, at the top of the funnel. But the price point that supports our best in class, best in category position, is actually the third one. It starts to feel expensive, but then you sit down and you think about the time that you spend in email, the ROI, and you still buy it anyway. It turns out that the median answer for the third question was $30 per month, and that's how we picked our price.

**Rahul Vohra** (01:08:27):
And once we picked our price, we then do a quick gut check on market size. For example, we're a venture scale company, but at the time the question that we had to ask is, "Could we grow into a billion dollar valuation?" Well, let's assume that at that point our valuation is 10 times our ARR, so our ARR would have to be a hundred million dollars. Well, that would be 300,000 subscribers at $30 per month. That is conservatively assuming no other ways to increase ARP. You mentioned price increase, you can also go up market, you can sell new products and so on. We asked ourselves, without those tricks, do we think we can get to hundreds of thousands of subscribers? And we answered emphatically, yes, so we went ahead with that price.

**Lenny Rachitsky** (01:09:13):
Okay, there's a couple more things I want to chat about in the time that we have and then I know you have to run. One is around AI and the work you guys are doing there. I know that's been a big unlock. And then two, the stuff you're doing in the enterprise. Then if we have time, there's a question I want to ask that I think is a really interesting way you guys operate.

**Lenny Rachitsky** (01:09:29):
Let's talk about AI first. It feels like there's this being in the right place at the right time. It feels like you guys have been building this for a while, and then AI just unlocked another stage in what you're able to do with email. Just talk about what you've done and what how you think about AI integrating into what you're doing, how it's enabled you to kind of take off again?

**Rahul Vohra** (01:09:52):
It's true that sometimes startups boil down to being in the right place at the right time. We actually had a massive AI launch recently about two weeks ago, but even before then we had multiple flagship AI features. Our first AI feature was write with AI, jot down a few words and we'll turn them into a fully written email. We actually match the voice and tone in the emails you've already sent. So unlike Co-pilot, unlike Gemini, unlike basically every other email app, the email sounds like you. This AI feature is way more popular than I expected it to be. On average today, users are using it 37 times per week.

**Rahul Vohra** (01:10:33):
Number two, our next AI feature was auto summarize, which shows a one line summary above every conversation. And as new emails arrive, it updates instantly. Again, unlike Co-pilot and Gemini, it's pre-computed. One of the things we do is, we go above and beyond to make these features really premium and feel amazing. The next AI feature after that was instant reply. Imagine waking up to an inbox where every email already has a draft reply. You would simply edit and then send, and sometimes you wouldn't even need to edit. I can share because we just finished this analysis, that over 2024, the percentage of emails that are AI written and sent with Superhuman has grown four times just in one year.

**Rahul Vohra** (01:11:20):
Then if I remember correctly, the feature after that was Ask AI. Email of course, is this treasure trove of critical information, things like project statuses, customer communication, meeting updates, deal execution, and so much more. And for over 40 years we've had to rely on what we hilariously call, search. You have to remember senders, guess keywords, scan subject lines, and now you can just ask, "Where is the queue one offsite?" or, "What are my flight details?" Or, "What is the top five most positive customer responses to the Ask AI launch?" A task by the way, which previously used to take me 20 or 30 minutes to read through all the emails and then create that report now happening in less than five seconds.

**Rahul Vohra** (01:12:09):
Recently we, like I said, announced our biggest evolution yet. Superhuman AI is constantly helping you. It's organizing your inbox. It's also making sure you never drop the ball. We have things that we call Auto Labels. You can now write a short prompt like job applications or requests to review work, and you can then immediately see when emails match that prompt, when people apply for a job or they ask you to review work. With Auto Reminders, if your email needs a response, Superhuman will now automatically set a reminder. You don't have to remember to do that and you'll never drop the ball again. All you need to do is hit send. With Auto Drafts, Superhuman will now automatically draft your follow-up emails for you and will soon be drafting replies to basically every email that needs a response.

**Rahul Vohra** (01:12:59):
And finally, with what we call Workflows, you can now turn email into repeatable automated workflows. For example, I often get emails from people who are interested in working at Superhuman, and I would normally reply to that candidate and I would let them know that the team will take a look. I'll then forward to the original message, including any resume or any letter to our head of people and operations and ask her to reach out, if interested. With Workflows, I can now automate this entire process. It's, you can imagine, creating a little flowchart of what has to happen. Not only does that save a huge amount of time, with Workflows you don't even have to be in your inbox. In fact, you don't even have to be working. You could be on vacation while Superhuman AI is working for you.

**Lenny Rachitsky** (01:13:53):
This sounds like product market fit to me. This all sounds wonderful. It just makes sense. This is the stuff we've been promised, our underwater cities and flying cars and then just email that just works magically and replies for us and all these things. I love all these things you're doing.

**Lenny Rachitsky** (01:14:08):
For folks that are building with AI. I'm curious, what's maybe been the biggest surprise, either good or bad, building so deeply on top of AI models that you think might be helpful for folks to just, "Watch out for this," or, "Hey, check this out."?

**Rahul Vohra** (01:14:25):
I think for me the biggest surprise has been how unpredictable the user love has been in terms of what they love and what they don't love. For example, write with AI. This sounds like a commodity feature and on all surface level it is. Every email app, every writing surface has a write with AI feature in. I would wager ours is the best at emails and surprisingly that's what we do. But the surprising thing was just how much people love it and how often it gets used. 37 times per user per week is still mind-blowing to me. I had not expected that, so that's the most surprising thing.

**Rahul Vohra** (01:15:11):
And on the flip side, there were certain AI features where I did expect a ton of usage, but we didn't quite get the usage that we were perhaps hoping for. Hopefully I'm not AI Kramer, but basically everything I thought would work out well, people use it less than they thought they did. And everything where I was like, "I don't know, but let's build the thing," people love that.

**Lenny Rachitsky** (01:15:33):
Interesting.

**Rahul Vohra** (01:15:34):
Maybe I should just create an anti-me to do AI road-mapping.

**Lenny Rachitsky** (01:15:38):
That's in a simple agent right there, whatever Rahul says, do the opposite.

**Rahul Vohra** (01:15:41):
Yeah.

**Lenny Rachitsky** (01:15:42):
Okay. Another maybe a last topic. I know that you guys are starting to move into the enterprise. When people think of Superhuman, they think of it's consumer-y, it's for people, and you guys are doing a lot of work to make it a B2B enterprise product. For founders maybe that are starting to think about this transitioning from PLG to sales led and B2B enterprising, what have you learned about just what it takes to get to that point and what does that sales motion look like for you guys?

**Rahul Vohra** (01:16:10):
In some ways, it's very like selling to prosumers, except these users are not coming from Gmail where prosumers would normally come from, they're coming from Outlook. And Outlook users have very different expectations to Gmail users. For example, Outlook users expect their email app to also be a fully featured calendar app, whereas Gmail users are happy with those two things being entirely different. As a result, we've invested in calendar very heavily and we continue to do so. There's only so much I can say, but it's pretty exciting.

**Lenny Rachitsky** (01:16:49):
[inaudible 01:16:49].

**Rahul Vohra** (01:16:48):
Outlook users are also used to certain safeguards, like if you've used Outlook in an enterprise, warnings when a recipient is external to your domain or what Outlook users might know as sensitivity labels. And as a result we've built support for external recipient indicators and sensitivity labels. But in some ways it's very different to selling to prosumers because there are other stakeholders involved. For example, we've built support for enterprise mobile management by implementing Microsoft Intune.

**Rahul Vohra** (01:17:22):
We recently sold one of the big three strategy consulting firms, which is super exciting. I can't say which one, but they love Superhuman and they have thousands of people internally using Superhuman. This is after a year... They've been piloting for a year and then accelerating over the last few months. We only just got them the mobile app, believe it or not. Because, at an enterprise like that, there are significant controls on what a allowed compliant mobile app can and cannot do. For example, IT needs to be able to control which apps can save attachments or which apps you can copy and paste text into from email. And for many enterprises, those controls are super important.

**Lenny Rachitsky** (01:18:14):
Wow, okay. So it sounds like essentially just building all these features that large companies need, is kind of the road you're on right now.

**Rahul Vohra** (01:18:21):
Exactly. And there's two stakeholders. There's the users, which are actually quite different because they're Outlook users and Calendar is one of the main ways that manifests. There's a whole bunch of other stakeholders, IT is one of them, but there are others as well. For example, companies this large have workplace management groups who want to see analytics of how people are working, how they can make their teams more efficient, so it truly is a multi-threaded sale with multiple stakeholders.

**Lenny Rachitsky** (01:18:49):
They had a product from Linear on the podcast [inaudible 01:18:51], and he actually, I don't know if you heard that episode, but he talks about how they decide what to prioritize, the thing they never build is middle managers needing to track how their reports are doing and things like that. That's an interesting opportunity for you guys maybe to cut stuff. I don't know.

**Lenny Rachitsky** (01:19:07):
Anyway, I want to end on one more nugget. Okay, I'm glad we have time for this. You shared that you have this system internally at Superhuman for making decisions. You call it Single Decisive Reason, SDR. What is that?

**Rahul Vohra** (01:19:22):
SDR is a thinking tool that I picked up from Reid Hoffman during my time at LinkedIn. The idea here is that for important decisions, you should be able to identify one, one reason that on its own supports the decision. It's based on the observation that all too often we rely on a collection of weak reasons to justify decisions. It's very, very easy to do this. Imagine you are contemplating a decision, you write a list of the pros and the cons. There are three pros, but let's say there are 10 or 15 cons. The sheer number of cons, the effort of thinking them through, the time it took to write them down, is going to affect you, consciously or worse subconsciously. This is especially true, I've seen in group settings, which just in general are a little bit more risk averse and a little bit more consensus driven.

**Rahul Vohra** (01:20:17):
So whenever anyone is making a decision and they bring that decision to me and they say, "Well, we want to do this because of X, Y, Z, and there are multiple reasons." I ask them, "What's the SDR? What's the single decisive reason?" If they can't yet isolate it, that tells me they haven't yet figured out why they want to make the decision. It doesn't mean the decision is wrong, it just means that they haven't figured out the singular reason why we should do the thing. They can then go through their list of reasons and ask, "Is this alone enough to support this decision?" Meaning if this was true and all the other things were not true, would I still do it? And sometimes we still do, but actually sometimes we don't. We realize that a collection of weak reasons alone means that, for example, the outcome is less likely than we thought it was, or it was hiding a really strong reason on the other side of the decision.

**Lenny Rachitsky** (01:21:11):
That is very cool. This is just when someone comes to you with a decision, the way you use this idea is, you ask them what's the single decisive reason?

**Rahul Vohra** (01:21:20):
Pretty much. Yeah. And what they can't do, obviously this happens, people are human and natural, they'll usually start mentioning three or four things, and that's fine. And then I will say, "Okay, but if only one of those was true and you're still advocating for this decision, what is it?" I think that's just a bar for a good decision.

**Lenny Rachitsky** (01:21:40):
Why is that so important? Because you found that a bunch of low quality reasons just don't add up to a good reason to do something?

**Rahul Vohra** (01:21:50):
Multiple reasons, which is ironic. But that's my SDR for why SDRs work. Which is yes, multiple low quality reasons rarely add up to a high quality reason to do something. But there are also other things as well, which is, any decision you take has an opportunity cost. Any feature you build is another feature that you didn't build. If we're going to build this for a collection of weak reasons, whereas we could build that for one strong reason, I'd much rather build that for one strong reason. Now this is all other things being equal, and these things often end up being quite complicated, but you can apply SDR all the way down. You just did that to me, what's my SDR for SDR?

**Lenny Rachitsky** (01:22:30):
There we go. Rahul, is there anything that we haven't covered that you wanted to cover? Is there any last piece of wisdom you want to leave listeners with before we let you go?

**Rahul Vohra** (01:22:43):
I feel good. I think we covered a lot. Thank you for asking amazing questions. This was really fun.

**Lenny Rachitsky** (01:22:51):
This was incredible. Okay, so let me just ask you this then. Where can folks find you online? Where can they check out Superhuman? What should they know before they try it out? And then just how can listeners be useful to you?

**Rahul Vohra** (01:23:02):
If you want to find me online, I am generally on X. That is x.com/rahulvohra, R-A-H-U-L V-O-H-R-A. My DMs are open, so feel free to ping me. If you're going to do that, I would suggest also emailing me, that's rahul@superhuman.com, and hopefully I'll see your message soon.

**Rahul Vohra** (01:23:22):
If you haven't tried Superhuman, then gosh, what are you doing? This is my call to you to do so, because your time is worth more than whatever you think it might be. So go download Superhuman and give it a shot. Invite your team. The metrics are real. I know they sound like the kind of metrics that startups make up, but getting through your email twice as fast, responding one to two days sooner, saving four hours or more every single week, they're all real.

**Rahul Vohra** (01:23:51):
Actually, speaking of which, the consulting firm I mentioned earlier, because they're so into data and into analysis, they wanted to corroborate those numbers for themselves, and so they did. They ran their own internal case study on Superhuman, and they were like, "Yeah, you're saving our partners 3.3 hours per person per week. And there's only one other tool that we've bought that does that, which is ChatGPT. So thank you. We love Superhuman. We're rolling it out." If that sounds interesting to you or your company, please do give it a shot.

**Lenny Rachitsky** (01:24:25):
That is super cool. Reflecting back on what I imagine this conversation would look like, a lot of contrarian thinking and attention to detail, I think that's exactly what it was. Rahul, you're awesome. Thank you so much for being here.

**Rahul Vohra** (01:24:39):
Thank you. Bye everyone.

**Lenny Rachitsky** (01:24:40):
Bye everyone.

**Lenny Rachitsky** (01:24:43):
Thank you so much for listening. If you found this valuable, you can subscribe to the show on Apple Podcasts, Spotify, or your favorite podcast app. Also, please consider giving us a rating or leaving a review, as that really helps other listeners find the podcast. You can find all past episodes or learn more about the show at lennyspodcast.com. See you in the next episode.

---

## [14/15] How to win in the AI era: Ship a feature every week, embrace technical debt, ruthlessly cut scope, and create magic your competitors can't copy | Gaurav Misra (CEO and co-founder of Captions)

**Gaurav Misra** (00:00:00):
There's rarely a time like this where so much is possible. Even like five, seven years ago, it's so hard to start a company. Everything feels like it's done, someone else is working on it. Suddenly, it's a time, right now, which I've never even experienced, where everything you try just works.

**Lenny Rachitsky** (00:00:14):
With people constantly hearing about all the things happening. Is there any tools or processes or approaches you've figured out to help stay focused?

**Gaurav Misra** (00:00:21):
Our engineering goal is every engineer should ship a marketable product every week.

**Lenny Rachitsky** (00:00:27):
I love just how wild that sounds. How do you maintain quality and make it all cohesive?

**Gaurav Misra** (00:00:31):
I actually think as a startup your job is to take on technical debt because that is how you operate faster than a bigger company. Bigger companies don't take contact technical debt, they pay it usually right away, or they're paying back technical debt from the days when they were a startup.

**Lenny Rachitsky** (00:00:45):
Is there anything else that in how you operate and the way you build product that you think is really unique and interesting?

**Gaurav Misra** (00:00:50):
We have what we think of as the public roadmap. This is basically what people have asked us for. There's all these surface areas where we receive user feedback, but these are all features that every competitor knows about. If a user is asking us for it, they're asking everybody for it.

**Gaurav Misra** (00:01:04):
It's not going to be a game changer in terms of winning against your competition. So we have a second roadmap which we think of as a secret roadmap.

**Lenny Rachitsky** (00:01:15):
Today my guest is Gaurav Misra. Gaurav was an early employee at Snap where he led the design engineering team, which he explains in the conversation. He's also an engineer at Microsoft and a couple other companies. Most recently, he's the co-founder and CEO of Captions, one of the most successful and cutting-edge consumer AI products, which lets you generate and edit talking videos with AI. They have over 10 million users and have raised over a hundred million dollars.

**Gaurav Misra** (00:04:53):
Thank you. Thanks for having me. Excited.

**Lenny Rachitsky** (00:04:55):
I very rarely have early stage founders on the podcast, but I wanted to chat with you because you're at the center of so much of what is top of mind for a lot of builders these days, AI and video, and just consumer and social apps.

**Lenny Rachitsky** (00:05:11):
Also going viral and finding new marketing channels. So I think there's a lot that people can learn from the way you approach product, the way you've built product and the way you just think about where things are going. So again, thank you for being here.

**Gaurav Misra** (00:05:26):
Appreciate it. Honestly, it's an exciting time. I got to say like there's rarely a time like this where so much is possible. In normal times, if you think about even five, seven years ago, it's so hard to start a company. It's so hard to come up with an idea. It's just like everything feels like it's done, someone else is working on it. Or it's like, oh it's been tried three times and failed three times. And suddenly, it's a time, right now, which I've never even experienced honestly in my career, where everything you try just works.

**Gaurav Misra** (00:05:56):
There's so many possibilities. There's not enough people in the world to work on them. Honestly. There's more things that can be done than there's people available to do them. It is just such a rare thing. And honestly, it's not going to last forever. We are going to catch up to this, but just feels lucky to be part of that movement. It's awesome.

**Lenny Rachitsky** (00:06:17):
When you said everything is working, I think what's an important distinction there is the building of the tool works. The tech is now there to build all these things that have not been possible before. The thing that is increasingly, difficult, and I want to get your take on this, is getting anyone to pay attention and stick with your thing because it's so easy to build stuff and everything is just awesome and interesting. It's harder to get people to pay attention and stick with your product.

**Lenny Rachitsky** (00:06:42):
So I guess is there anything there, you've learned that you've built a number of successful products. We'll talk about Snap and what you're doing now, about just, I don't know, what you need to think about these days to get anyone to pay attention and then stick around.

**Gaurav Misra** (00:06:54):
Yeah, I mean honestly it's a great point and I think there is a lot of hype obviously, and part of it, that's what's driving a lot of this growth for a lot of companies. And I think from a user acquisition/marketing perspective, in a world five or seven years ago, if you were making something novel and you went to users and it was like, "Oh, we got something better." People are going to be like, "Well, whatever. Everybody says they got something better. I don't care." But today, and this is not probably the way you should do it, but you can go and just say, "We've rethought this thing with AI." And a bunch of people will just be like, "Well, how?" Or "Maybe I should check this out."

**Gaurav Misra** (00:07:33):
They'll just try it. Obviously, you have to deliver on the promises. If you don't deliver, people will come in, they'll play around a bunch and then just leave. But if you can truly deliver on the promises, there's great opportunities to require users at scale. So I think that's slightly different. And I don't know how long that lasts, but it is definitely a different time from that perspective. I do think also at the core of building products is solving problems. I think a lot of people sort of get caught up in this, well, it's cool and people will come for the cool right now. People will come in and be like, "Well, let me check it out. It's cool."

**Gaurav Misra** (00:08:11):
But at the end of the day, if you're just building a playground and people play around in the playground and then they leave after playing around, it's not a business. So I think that is still key. You have to be solving real problems.

**Lenny Rachitsky** (00:08:25):
As we were talking, I'm thinking about every day there's something that would maybe a few years ago be news for a year. Holy shit, this is now possible. Now it's like every day something like that happens and then we're like, all right, so what I think about is like, we'll have AGI one of these days or super intelligence and everyone's going to be, "Oh, amazing." And then, "Okay, what's for dinner?"

**Gaurav Misra** (00:08:46):
Isn't that already happening? Think about, in a way, I self-reflect on this sometimes if like, you've seen Iron Man and stuff, they have the J.A.R.V.I.S thing and you've seen Interstellar and they have the TARS machine. They're talking back and forth with these things like bouncing ideas. That is science fiction. That's literally science fiction. Okay, it's not perfect, but it exists in a way that nobody could have imagined. That's science fiction has become reality and I feel like nobody cares.

**Gaurav Misra** (00:09:18):
In a way, you would've expected the world to be turned upside down, but it feels like almost in a way so slow and people are like, yes, adoption is happening, but I feel like it's almost a shocking development in a way.

**Lenny Rachitsky** (00:09:31):
It feels like you guys have done a good job staying top of mind and continuing to get people excited because to your point, there's so much happening. How do you get people to continue to be like, "Oh, okay, wow, what their building is actually is interesting and continues to be interesting."

**Lenny Rachitsky** (00:09:45):
Anything you've learned about just what it takes to stay top of mind and continue to pull people back and get people re-excited over and over?

**Gaurav Misra** (00:09:51):
Hundred percent. I mean, I think honestly it just comes down to not just AI for the sake of AI or AI for the sake of excitement or hype or novelty or whatever that is, it's actually effective AI like AI that solves real problems, practical problems. And the fundamentals haven't changed. In a way, there's three steps to building products. You identify a user problem, you apply some technology to solve that problem, but then finally you have some mechanism to find people who have that problem. If you can do all three of those things, then in any environment you can create great products. But I think right now what's different is so much is changing on the technology side that you can create products that could not have been created before and solve problems that could not have been solved before. And that's creating the opportunity.

**Gaurav Misra** (00:10:45):
And for us, especially in the video space, it's truly endless. We've just begun, our goal specifically for video is not to build professional tools. We're not building for professionals at all. We're building for the person who could not have created video before. They didn't have the tools, the skills, the means to be able to create video and now they can because they're able to jump over that skill gap or that time gap. Maybe they're business owners, they don't have time, they want results, and honestly a lot to solve there just tons.

**Lenny Rachitsky** (00:11:20):
Solve people's problems. Easier said than done, but it's a good reminder. In the end that's all that matters. Something that I always think about with people in your shoes is just how do you not get overwhelmed and how do you know what to pay attention to? How do you stay focused?

**Lenny Rachitsky** (00:11:36):
Any tips there for folks that are just reading every day, a new announcement, and then just like I just, how do I? What do I do? There's too much.

**Gaurav Misra** (00:11:44):
It is the new problem of product development in a way. There's too many possible paths you can go down. There's too many ideas, there's too many things you could do. And I mean obviously, prioritization is always an important skill set and has always been, but it's become an even more important skill set right now because you have to figure out what not to pay attention to. Our general framework for it is to look for user demand, and actually the easiest way to check for user demand is to just see what has virality.

**Gaurav Misra** (00:12:10):
Usually, what has virality and what people want to share and talk about, there's something at the core of it that actually is interesting. Now, it may not always be interesting in a way that's like maybe it's a one-time use case. Maybe it's not something that people would do repeatedly. Maybe it's not something you could build like a subscription business off of, but oftentimes there's some things, some core element of it that has resonated with people. And if you can identify that core and then mold it into fitting into your business, it's actually a great way to identify what actually works. And we have these tools right now. We don't have to build anything. You can just kind talk about it and people will share it, share the idea.

**Gaurav Misra** (00:12:49):
And you can measure how well the product might be received even before you built anything. So it's a great tool we use for prioritization. We spend a lot of time on social media. Obviously, our app is often used for social media, so a lot of our employees will spend a lot of time on social media. We look at what the trends are, what's happening, and based on that we can get a pretty good read of what might resonate well with people.

**Lenny Rachitsky** (00:13:15):
So as a leader of a company with people constantly hearing about all the things happening, is there any tools or processes or approaches you've figured out to help people continue, stay focused, not get excited about every shiny new object and actually ship things? I

**Gaurav Misra** (00:13:30):
I mean honestly, it's all about incrementality in a way. I think we do aim to ship every week. Our engineering goal is every engineer should ship a marketable product every week. And so what's a marketable product is a product that you can show to users and the user might subscribe or pay for the app just for that or come to the app essentially just for that. And that's why table stakes features, let's say we're talking about word processor or something. If you had auto format or just table stakes stuff like justify alignment or something, no one's going to come to your word processor for justify alignment. You can market that because it's obvious, of course it exists, but if you did something unique that nobody else has done, you can go and show that to people and people will come to your app just for that.

**Gaurav Misra** (00:14:22):
And even if your app doesn't have a lot of the obvious stuff, maybe it doesn't have justify alignment, people will jump over that just to use these new tools and new abilities that you might be building and marketing. So we try to do every engineer one marketable feature per week, and a lot of that stuff may not work, but a lot of it does work and we can figure out obviously, where to put in more effort, things that start to work, we double down on those things, build more. People often complain because think about it, in one week where we're shipping, it's not complete, it's MVP, truly.

**Gaurav Misra** (00:14:57):
And we slice the hell out of it. We take the design and we cut, cut, cut until we can really say that it's going to be useless if we cut anymore. We get that out and people come in. And if things are going well, people will use it despite all the problems that it might have, and now people will complain and we'll have a list of problems and we know what to do next. That's a starting point essentially. As long as we're shipping one a week, we get a ton of volume of features and products and directions we're releasing, cut a lot of that. What remains expand from there. So it works really well, and it keeps people focused.

**Lenny Rachitsky** (00:15:37):
I love the simplicity of that. I love just how wild that sounds for a lot of companies I imagine. Every engineer ships a marketable feature or product every week.

**Gaurav Misra** (00:15:46):
Yes.

**Lenny Rachitsky** (00:15:48):
There's some people listening to this and are just completely stressed out by this idea and there's some people listening who are like, this is exactly how I want to work. This is how every company should build.

**Gaurav Misra** (00:15:56):
Yep.

**Lenny Rachitsky** (00:15:58):
How do you maintain quality and make it all cohesive? I imagine that's the big trade-off. Just, any tricks there for folks that want to maybe start operating this way.

**Gaurav Misra** (00:16:06):
Quality is not something you compromise on most of the time. I think yes, there's strategic compromises in quality, but most of the time what you want to do is have a bar for quality where people should come in and if they're using the feature, it should work, right? Of course. And the way to cut down on time, and I think this is a mistake people make a lot of the time, is when time is being pressured downward, a lot of times engineers, PMS, designers, they will cut on quality rather than cutting on scope. And actually you can cut on scope. It's actually, the method that we use is we look at every element that's going to take any time to build and we just say, what if we remove this? Is the product still useful?

**Gaurav Misra** (00:16:48):
And we keep repeating that until we remove whatever's left and we say it's going to be useless at this point. And that becomes the one-week project, right? It actually really works. It narrows down to the core of what you're really trying to ask. So for example, let's say we wanted to build something to add an image on your video or something like that, and this is a really basic idea. I just made it up right now. And you might imagine a design in which you import your image from your camera roll, but before it lands in your video you might want to remove the background. You might want to change the hue and saturation or something like that. And you might expect a designer to design all of those features and you let it design, but you really quickly realize that you can cut all of that stuff.

**Gaurav Misra** (00:17:38):
You can cut the background or you can cut the hue saturation. All you really need is pick. And then there might be a picker. We need a picker with a library, with a lot of different type. What if you want to pull from the cloud? What if you want to pull from the drive or something like that? Cut all of that, right? And essentially come down to the core, which is just native picker from the camera, lens, straight in the video, no UI. And that is already, that should be useful. If that's not useful, then anything else built on top of that is also useless. So that's how we might go about it.

**Lenny Rachitsky** (00:18:12):
That last sentence is so key to this. It's the core idea of ship small iterative features before you invest a lot in something to figure out is there anything there, is this worth spending weeks on?

**Gaurav Misra** (00:18:24):
Totally. And I think the coolest part of this method is the first thing that the users will come in, they'll use the thing, they'll import images and the first thing they'll complain about is what bothers them the most? Is it human saturation? Is it background removal? Is it picking from the cloud? You'll just get the most complaints about that thing.

**Gaurav Misra** (00:18:43):
People will be like, and people will be honest about it or they'll be like, "This sucks. It doesn't even have background removal. What kind of image thing is this?" And you have to take that feedback and just next week you can ship in a single week all the things that the user's complaining about.

**Lenny Rachitsky** (00:18:59):
And then they're like, wow, this team is shipping like crazy.

**Gaurav Misra** (00:18:59):
Yes. Exactly.

**Lenny Rachitsky** (00:19:01):
Solve all my problems. So responsive. This connects a common sign of product market fit, which is when people are complaining about the thing that means they actually care enough to complain and that's a really good sign if they're complaining about something.

**Gaurav Misra** (00:19:13):
It's very true, very true. If nobody complains, it's almost red flag.

**Lenny Rachitsky** (00:19:18):
A lot of this is turning into an archeology of a modern product team and startup. So I want to keep digging. This is not where I was planning to go, but this is awesome. I love that this approach of every engineer shipping something every week that's marketable connects directly to where I started this conversation, which is how do you stay above the noise?

**Lenny Rachitsky** (00:19:36):
And part of the answer is just ship stuff constantly, and just continue to impress people. Like, "Here's a new amazing video feature." "Look at this thing."

**Gaurav Misra** (00:19:43):
Exactly. Yep. I think it's definitely key, right? And there's enough area and enough scope for that to happen. I think truly in normal times it may not be possible to create that much roadmap that quickly, but I think because there's so much innovation underlying all this, there is that scope available. The roadmap almost seems unlimited, just truly.

**Lenny Rachitsky** (00:20:07):
Okay. The other question I imagine people would be wondering is how do you work on longer-term projects that take many weeks? There's also infrastructure, I guess, back-end stuff. So maybe answer those questions.

**Lenny Rachitsky** (00:20:17):
How do you think about long-term stuff and then how do you deal with back-end stuff that isn't a feature that anyone would care for?

**Gaurav Misra** (00:20:22):
Yep. Usually, we'll dedicate time to that separately. For example, usually Q4 for us is infrastructure quarter. We just go and build all the infrastructure. Q4 is generally, we've already delivered a ton of products and stuff. We're feeling pretty good about the rest of the year. Things are winding down. Obviously, holidays and stuff coming up. And so we spend all that time paying the technical debt.

**Gaurav Misra** (00:20:48):
I actually think there's a unique thing to think here about technical debt in general. And as a startup, your job is to take on technical debt because that is how you operate faster than a bigger company. Bigger companies don't take on technical debt, they pay it usually right away. Or they're paying back technical debt from the days when they were a startup, and they took on a lot of it. I mean Snaps, I used to work at Snap and there was a lot of examples of that over there, and I'm sure it happens at every other company.

**Gaurav Misra** (00:21:19):
And we think about it as like, well, is this a problem we need to solve today or is this a problem that the 50th engineer or the hundredth engineer or the 500th engineer can solve? And if it is a problem that a future engineer can solve, we should use that future engineer now. Essentially, that's what we're doing. And we're saying we're going to push this to somebody in the future. And by the way, if the company fails, that engineer will never be hired and all this won't matter anyways. So it's like financial debt in many ways. Financial debt is taken on to create leverage. It can be a good thing like if you're buying a house, you take on debt and you can buy something probably more than you can afford without taking on debt.

**Gaurav Misra** (00:22:04):
And it's the same thing. You can create products that you wouldn't be able to build with a small team that you have by taking on strategic technical debt. It's very positive actually.

**Lenny Rachitsky** (00:22:13):
Wow, this is such a cool idea. And where my mind goes is that future engineer may be an AI agent engineer.

**Gaurav Misra** (00:22:19):
Exactly, yeah.

**Lenny Rachitsky** (00:22:21):
Just solving problems, just on technical debt in you.

**Gaurav Misra** (00:22:24):
Exactly. Some engineer in the future Five-hundred engineer many years from now will get a promotion because they solve this big problem that those really bad early engineers created.

**Lenny Rachitsky** (00:22:36):
So obviously, there's a line to this. There's only so much debt you can take on before you become a big problem.

**Lenny Rachitsky** (00:22:43):
Is there any thoughts on just that balance of just how much is too much and how if it's enough for a net feature engineer or just-

**Gaurav Misra** (00:22:50):
Yeah, I mean I think generally the rule of thumb is every piece of debt that you take on you have to pay interest on. So if there is debt that you've taken on, there's 1% or 2% of your time that is going to be taken away every day in maintaining bugs and issues and restarts and crashes and things that are happening with that. Because you did it the fast way, something's going to go wrong with it. Every day. 1% of your time will be taken away. If you take on enough debt, you'll be paying 80 or 90% interest and you'll not have any time to do anything new. You'll just be paying interest. That's all.

**Gaurav Misra** (00:23:23):
And that's when you get into the mode of like, oh, we're just keeping the lights on. We don't have any engineers to do anything. We're just keeping the lights on. That's the failure case for a startup. So in a way, you have a technical debt runway. Once you run out, once you've taken on too much debt. And if you haven't delivered value in that time, enough value to hire the engineers to pay the interest or just pay off the debt, you'll get in trouble.

**Lenny Rachitsky** (00:23:46):
I love that. That's such a nice heuristic of how to think about when to invest in something. I don't want to go down this too far, but just a thought I have is ... because sometimes there's big technical decisions you got to make that impact the way everything builds or is built in the future. I imagine those you spend more time on and take really seriously.

**Gaurav Misra** (00:24:02):
Definitely. Yeah, I mean I think as long as it's possible for wherever it's like a two-way door, you can do whatever you want. I mean this is a classic methodology. If it's A one-way door, it's worth thinking about and sort of doing correctly at least as much as the one-way door would matter to you in the future.

**Lenny Rachitsky** (00:24:22):
How much do your engineers use Cursor and tools like that to build? How much is AI helping your team move?

**Gaurav Misra** (00:24:28):
A hundred percent, yeah. I mean everybody's using it. It's super helpful. I mean even I'm using it honestly. Yeah, it's a huge multiplier for the team, no doubt.

**Lenny Rachitsky** (00:24:40):
And is a Cursor specifically. Is there anything else that you guys found useful?

**Gaurav Misra** (00:24:43):
Yeah, we are using Cursor. Yep. We've tried all the different tools. We were using Devin as well, which is another, you know? That's more advanced, I guess. It's solving bugs for you.

**Lenny Rachitsky** (00:24:52):
Yeah, Devin's basically, I think it's 500 bucks a month and it's like an AI engineer that you just chat within Slack.

**Gaurav Misra** (00:24:58):
Exactly, yeah. In a way, these are the types of things that us as a startup can do that bigger companies can't just, you know, they can't just pull in Devin. They have to get 30 lawyers in the room first before that happens.

**Lenny Rachitsky** (00:25:11):
And they're all called Devin, these are like agents. Everyone's going to have hundreds of Devins working at their company.

**Gaurav Misra** (00:25:15):
Exactly. You can have multiple Devins. I actually heard you can have a manager of Devins who's managing Devins.

**Lenny Rachitsky** (00:25:21):
I love that managers are all getting layered, like unlayered and then they're going to have AI managers. That's the ultimate bait and switch.

**Gaurav Misra** (00:25:30):
Yep.

**Lenny Rachitsky** (00:25:32):
Okay. Is there anything else that in how you operate and build the way you build product or set up the way you build product that you think is really unique and interesting that other people might be able to learn from?

**Gaurav Misra** (00:25:42):
Our process is a bit interesting in that way. We have a design team, we have a PM team. We're very early on those teams right now. And obviously, we have engineering. And we have all the different surface areas. So iOS, Android, web. There's backend team, machine learning team, research team. So generally, when we're developing products, we may start off with a PM first approach where we're finding some sort of overall issue that we want to take on some new area or pillar we want to take on and then creating sort of product specs from there.

**Gaurav Misra** (00:26:17):
But a lot of times we'll also start the opposite way. We'll first design something without even having any idea of what or why we're doing it, but we'll design a bunch of different things and then we'll sit down with the PMs and look at the designs and just go over one and the next and the next until we find interesting things and ideas that pop out of that.

**Gaurav Misra** (00:26:36):
And a lot of times that leads to us discovering things that we wouldn't have discovered if we were just too focused on the metrics and the numbers and things like that. So it's almost reversing the process a little bit and starting with design first, but it can often result in finding unique ideas basically. I also think that we have a unique setup in how we create our roadmap. So normally you have a single roadmap and we actually divide a roadmap into two different roadmaps. So we have what we think of as the public roadmap. This is basically what people have asked us for. So there's all these surface areas where we receive user feedback and we look at all that feedback and people will ask for features. They'll ask for, I want background removal, I want to undo and redo, I want to upload longer videos, whatever it is, a bunch of different features.

**Gaurav Misra** (00:27:26):
And we'll just make a list of that. And just like anything else, we'll prioritize it and we'll look at how many people it affects and what the possible markets are and just get it done basically one at a time.

**Gaurav Misra** (00:27:37):
But these are all features that every competitor knows about. These are public. If a user's asking us for it, they're asking everybody for it. And every team has essentially more or less the same list and everybody's prioritizing it. And yeah, sure you can win a little by extra nicely prioritizing it or winning a little in prioritization or execution or something, but it's not going to be a game changer in terms of winning against your competition. So we have a second roadmap which we think of as a secret roadmap. So this is a roadmap that nobody asked for anything on this like literally, nobody has ever asked for it.

**Gaurav Misra** (00:28:13):
And if a user were shown something on it, they might be like, "I don't need this. I don't know what this is." But given our unique vantage point, our unique understanding of the problem set, the user space and the technology, we've come up with some special ideas that we think will completely revolutionize how something is used where we can truly change the behavior of the user. I think that's what at is at the core of. It's like people are doing things one way if we're able to show them another way. And once they try it, they never go back. That's what a product is, that's success. And those are the types of ideas that we put on the secret roadmap. These are things we never talk about publicly, never tell anybody about, and we announce them and just give them to users and see the effects.

**Gaurav Misra** (00:29:00):
A lot of this we come up with through brainstorming. So we do actually do quarterly brainstorming, company-wide, everybody's included like everybody from. It's not just a product team thing, it's like engineering, recruiting, everybody's included in. And we all come up with marketing, obviously, everybody comes up with ideas, we vote on the ideas, rank the ideas, and then the product team takes over from there and thinks about like feasibility and technology and what the different things could be. So this is a way where we can take all that noise that people are getting, everybody's browsing social media, seeing all these different things that are blowing up, these models and advancements and we can get all that information together and provide a unique internal roadmap where how are we going to approach and create value out of all of these different advances that are happening.

**Gaurav Misra** (00:29:49):
So that's our general methodology. And a lot of times the biggest wins will come from the secret roadmap. That's the game-changing stuff. It's not going to be the user requests usually that are going to do that.

**Lenny Rachitsky** (00:30:02):
I love just how calling it the secret roadmap makes it extra interesting. [inaudible 00:30:07]

**Gaurav Misra** (00:30:06):
Exactly, yeah. It's a secret.

**Lenny Rachitsky** (00:30:09):
It's a secret. I'm not even going to ask you what's on that secret roadmap. You can't tell me.

**Lenny Rachitsky** (00:30:15):
What's an example of feature that came out of that secret roadmap that's been a big deal for you guys?

**Gaurav Misra** (00:30:18):
Tons. I mean, I'll give you an example from a long time ago. One of the first AI features we added after the app initially took off was this feature called eye contact. So this was a feature where if you're recording something, oftentimes people who are new to recording a video might read from a script or a teleprompter or something like that and they might have that off-screen. So it looks like you're reading and it's not great from the perspective of the video itself or the viewer of the video. So we had this feature where it basically shifts your eyes to look at the camera.

**Gaurav Misra** (00:30:52):
And we were actually the first company to build this. We worked with Nvidia on this. It's actually really interesting because when we originally reached out to Nvidia about this. They were not sure why we needed this. And they actually gave it to us pretty openly and were excited about some sort of partnership of how can we get this technology into something that could be useful.

**Gaurav Misra** (00:31:19):
But we saw this creator use case which was unique, and it was one of the ideas that came out of the brainstorm and we threw it on there, we launched it. It was a huge success. I mean, I'll be honest, the video, the ad that we made, a social media post that demonstrates this was so viral, it was made in basically every language around the world. It still till today gets millions of views. We find reposts and reposts of that thing that other people have created that get millions and millions and millions of views because people are like, "Wow, this is a great idea."

**Gaurav Misra** (00:31:59):
And now it's been copied the hell out of, I think it's available basically on every app you can imagine. For good reason of course. But that's one of the ideas that came out of it.

**Lenny Rachitsky** (00:32:10):
You talked about how you come up with these secret roadmap ideas. I'm just intrigued by this. I'm going to spend a little more time here.

**Lenny Rachitsky** (00:32:14):
Does your team ever work with an AI LLM to help brainstorm? I imagine that's where things will go, where you're actually jamming. The AI agent is brainstorming along with you.

**Gaurav Misra** (00:32:25):
Honestly, I would like for it to go there. It hasn't gone there yet. We haven't done that exactly, because the problem is context. And I think just the context of understanding the user, the use case, it's so abstract. Even right now, I feel like I understand our users obviously, but I can't exactly verbalize why that is or how that is, a little bit abstract. And I spend a lot of time with RPMs and designers imparting anything that I understand and I've learned over the many years I've been working on this, how do I impart this to them? But then it's a challenge because I can't even verbalize it myself. And so it's an extra hard challenge to figure out how do I put this context? How do I make it available to an LLM when I can't even put it into words exactly. And honestly, this is probably my own feeling but, and I need to work on this, but there is something to it.

**Gaurav Misra** (00:33:26):
I do remember at Snap for example, I think one of the most unique things about Snap and the CEO Evan Spiegel was that he had an unmatched understanding of the user. I think years and years and years of the company's existence past, almost a decade. And nobody understood the user like he did. He would come up with ideas that everybody would disagree with and we would launch them and there would be hits, just hits after hits. And nobody would understand why. Everyone would line up and be like, "Great." Round of applause for everyone, but no one knew why.

**Gaurav Misra** (00:34:07):
A great example of that is a lot of this was figured out in retrospect too. I think there was a point at which Snap declared that they're a camera company and a lot of people laugh at them and said, "Camera. What are we making digital cameras or something?" Or, "Why is it a camera company?"

**Gaurav Misra** (00:34:22):
But I think at the core of it was this idea that Snapchat opens to the camera and that was actually the differentiator. That actually that small decision was holding the entire company against all competition because when the moment passes where your friend is doing something funny and you need to capture it, you're not going to open Instagram or anything else because it doesn't open to the camera. You're going to open Snapchat because you can capture it right away. And Instagram can never copy that because all their metrics are going to go down as soon as they do that. So that is a fundamental understanding. And I figured this out much later, actually, but it's such a powerful idea.

**Lenny Rachitsky** (00:35:08):
I'm glad you talked about Snap. That's where I definitely wanted to go. This is where I was going to start. So I'm glad we circled back to your experience at Snap. So the reason I am interested in this is if you think about social networks like Snap is basically the last social network to have launched and stuck around other than TikTok, which I don't think is a social network. I think it's just this content platform. I don't think you're really interacting with people really. And that was 2011 when it launched. So it's been like 15 years since the last social network launch that has worked.

**Lenny Rachitsky** (00:35:40):
And I think it's interesting also because there's rarely been a lot of insight into just how Snap operates. You were there really early. You're a big deal at Snap. You built a lot of really important features. So I wanted to spend a little time here, and it feels like a lot of things you learn from Snap you're bringing to your company now. So let me just ask, I think you may have answered this, but I'm curious if there's something else here just broadly maybe other than Ev's brain, what do you think was core to Snap being a successful consumer social product?

**Gaurav Misra** (00:36:12):
There were a couple of different things that went well. I do think for a company like Snapchat or Social Network, the core product market fit can be extremely strong. Essentially, the reason that people are downloading it, the way that it's spreading, the way that it's distributing, the way that it's inviting friends or sending Snaps or whatever it is, that product market fit can be so strong sometimes that it can be hard to actually build something because you actually can't tell if what you're building is what's responsible for growing the thing or if it's actually hurting it and it's growing despite what you're doing basically.

**Gaurav Misra** (00:36:53):
And I think because of that, it actually sometimes teaches people the wrong things. It teaches people that the contrarian thing that they were doing was right when it was actually just wrong and the company just grew despite it. And I think some of the things that Snap did well and it needed to do really was to continue innovating, right?

**Gaurav Misra** (00:37:17):
Because for a company like Snap, it has a ton of competition. Social networks are monopolies by nature and there's a lot of reasons for Facebook or any other social network to stop the growth of Snapchat. And they tried, they tried really, really hard. And the way that Snap was avoiding that was by innovating. I think the core of it was the setup that they had, which was very unique. I've never seen anything like it. I've worked at a bunch of different companies, but obviously there's a CEO and the CEO was very product-led, his designer himself, but he surrounded himself with the design team. That was sort of the central team in the company. And the design team was like 10, 12 people. Basically, pretty small, even at 5, 6,000 employees it was that small still.

**Lenny Rachitsky** (00:38:02):
Oh wow. At 6 or 6,000 employees. The design team was, you said how many, five or six people?

**Gaurav Misra** (00:38:07):
10, 12 people.

**Lenny Rachitsky** (00:38:08):
10, 12. And to add to that, there's no PMs really for a long time. That was before.

**Gaurav Misra** (00:38:12):
For a long time, yeah.

**Lenny Rachitsky** (00:38:14):
Big difference.

**Gaurav Misra** (00:38:14):
Initially, there were no PMs at all. PMs were introduced with monetization. Once monetization was a big sort of element, that's where PMs came in. Today, I think there's a ton of, or there's an adequate number of PMs across the company, but there was a long period of time, especially when the innovation was happening, when there were a much, much smaller number of PMs and it was very designer led. But at the same time, I think that's slightly misleading in the way that these weren't your sort of average designers.

**Gaurav Misra** (00:38:43):
These were designers who were actually PMs as well. That's what the secret sauce was. They were able to not just design but also do the PM part which is a big responsibility. It's a lot of work, especially for that many employees, but it gave the CEO a way to have granular control over what exactly was being launched in which part of the app at all times.

**Gaurav Misra** (00:39:05):
Because he could meet with a set of 10 or 12 people and know every change that was happening that was user impacting. A lot of changes were being worked on that were infrastructure and types of things that keep going on in the back end where you're improving ranking and whatever that might be, performance and things like that. And those were not usually his concern. He was concerned with what UI are we adding where? And if you needed to add UI to the app, you needed it designed. And if there's no designers in the company, except for a handful who talk directly to the CEO, you create a very granular control over what's being launched in the company. So everything needed to be approved by Evan. If you hadn't approved it, it's not going out. So the design team actually held a lot of power in that.

**Lenny Rachitsky** (00:39:50):
This is awesome. So what I'm hearing partly is, I don't know if this is true, but it feels true that to make a consumer app that is successful and breaks through, you almost need a singular mind that continues to stay in the weeds on everything. And the way Evan did that is very close to the design team who basically ran product.

**Gaurav Misra** (00:40:10):
That's very true. Yeah, it's very true. And he was able to keep the context of the entire app in his head at the same time. He knew the interdependencies and what we're doing and why we're doing it. And so that gave him just very granular control over the company's product roadmap.

**Lenny Rachitsky** (00:40:26):
It makes me think about Brian Chesky and Airbnb is a consumer product, it's not a social network, but I wonder if that's just an interesting insight just for consumer products. They will generally do better if there's one person with a really ... the right sort of combination of experiences, insights, and just they continue to run and own every detail.

**Gaurav Misra** (00:40:46):
Definitely. And also the ability to bring about change, the ability to truly energize an entire organization to do something that's not just incremental but fundamental.

**Gaurav Misra** (00:40:58):
<< Founder mode >>

**Lenny Rachitsky** (00:41:00):
Exactly.

**Gaurav Misra** (00:41:01):
Founder mode. That's what we're getting to, basically.

**Lenny Rachitsky** (00:41:04):
Yeah, ever heard of it. Okay. And then you said that these designers, so I know it's famous that Snap had no PMs for a long time. Designers were PMs. This point you made about the designers where PME is really important. I think a lot of people look at this, they're like, "Amazing. We're just going to hire just designers. We don't need all these PMs. Slow everything down. Just tell us what not to build." Can you just talk about the level of these designers? What allowed them to be as successful as they were without any PMs?

**Gaurav Misra** (00:41:32):
Yeah, I mean I think what was expected from the designers now was not just the ability to design, the skill set of designing, which all of them were IC designers by the way. And there were no reports, so they weren't allowed to have reports actually. And so they were designing everything themselves, but they also had to have the leadership skills to go figure out the roadmap, write all the documents, work with the different teams, figure out shipping schedules and just know everything, not just the technical and the engineering part, but the UX and the UI and the product needs and why are we doing this.

**Gaurav Misra** (00:42:15):
The roadmap, there's just a ton to keep in mind. And that means that it was a job that was just highly ... it was very high workload. No doubt, very high workload. These people work really hard and they were paid highly too. For what it's worth, they were paid way higher than you would expect designers or PMs or engineers to be paid with quarterly bonuses and all kinds of things.

**Lenny Rachitsky** (00:42:43):
That's interesting. And it reminds, people always say, "Why do you need PMs?" There's like someone has to do the work that a PM does. They're not sitting around doing nothing. And it's important to note the person that will take on the PME work, they have to be good at it and enjoy it. And a lot of designers don't want to be doing writing docs and organizing stakeholders and getting alignment and ...

**Gaurav Misra** (00:43:03):
100%, 100%. That's why it was so hard to find those people who were able to do two things. I actually think there's an insight in there is innovation between when you're merging craft right between two different functions. And I do think there's something special about one person doing two different functions or at least being able to do. And I think a lot of unique insight and innovation can come from that.

**Gaurav Misra** (00:43:31):
I actually think on my personal side, I eventually joined the design team. I started at Snap on the engineering team. I eventually joined the design team over the last two years that I was at Snap. And a big part of what I did there was create this function called design engineering. And that was actually a different combination. It wasn't the designer PM. It was the designer engineer. The person who can think of the UX design it and also build it and launch it, all of those things.

**Gaurav Misra** (00:44:03):
And we saw both the ability to take designers and teach them engineering and take engineers and teach them design as part of that. Obviously, the reason that we created that function was very different. It was actually to continue innovating as the company got bigger. One of the problems that we identified was that as the company got bigger and bigger and there's like 500 engineers, 1,000 engineers, 2,000 engineers, 3,000, suddenly it just becomes very difficult to do everything.

**Gaurav Misra** (00:44:32):
Everything is a six-month project or a one-year project. Every product is a massive investment of 500 engineers and a lot of time. And so you really have to pick your bets. If you get it wrong, if you are innovating and trying to create new products and you spend 500 engineers for a year and it doesn't work, it's a big problem. You're going to be in trouble, especially if we're coming like Snap where everybody was copying what they're doing so they had to constantly innovate, create new stuff and push the bounds.

**Gaurav Misra** (00:44:59):
I think Evan's philosophy was always he didn't fight the things that were getting copied, right? Stories got copied pretty much straight up. A lot of things that Snap created got copied, but he was more of the mindset of like, "Let's expand the pie, do something new and push the boundaries." We'll keep innovating basically. And so to do that with that scale of a company becomes really hard. And so we had this idea of let's create a small team where we can go and pretest a lot of these ideas because we had a lot of ideas and we can't go and build all of these things. So the idea was create a small team of these design engineers, people who are able to do the entire product design engineering process in their head and can put together early versions of the product, which we would actually bake into the Snapchat app itself.

**Gaurav Misra** (00:45:46):
And we were able to even test, for example, run a test in Australia, see how it's performing. Run a test in a couple of high schools, just a couple of high schools, see how people behave. And that way we already have data on how this might perform in a real world environment, but we haven't built it to production level. It's a prototype, essentially. It's how a startup might build something.

**Gaurav Misra** (00:46:08):
The same idea of what we're doing at our company now, build fast, get it out there, get feedback, understand whether it works or not, and then work with the engineering team to build it at a scale. Once we understand the product and the dynamics, then it makes sense to put on 500 engineers for six months to build it.

**Gaurav Misra** (00:46:25):
So that was a big part of it. I think the nice thing that came out of it that was completely unexpected but actually transformational for me in a way was obviously in big organizations, alignment is a big issue. How do you get everybody on the same page? And a big part of a PM's job is actually to create alignment and it can be a lot of work because you go talk to all these stakeholders and get them on the same page.

**Gaurav Misra** (00:46:48):
But one of the insights that we had, which was unique was as the company gets bigger, you can actually create alignment by causing internal virality. If there's enough people in the company, it actually starts acting like a consumer base might. If you share something interesting with someone, they will share it with somebody else because they think it's interesting and you can actually create virality inside a company.

**Gaurav Misra** (00:47:16):
So one thing that we would do is we would create these prototype products. We would just go into an area, redo a bunch of stuff, create these prototype products that didn't exist in Snapchat normally, and then we would just share the build and it would explode. It would just go viral inside the company. Day after day we would hear from engineers, then managers, then VPs, then eventually from Evan being like, "Oh my God, everyone's talking about this. Why am I the last one to hear about it?"

**Gaurav Misra** (00:47:47):
So it would create instant alignment across the company of this is exciting, this is something that we want to get behind. And everyone would be asking, "When are we doing this? When is this happening? I see someone's already working on it." So it was a great way to do that. And once we really understood that the product actually had good dynamics and we had tested it, it was a great way to get it out in front of everybody and create this idea of, "Hey, we're all working on this. This is the future."

**Lenny Rachitsky** (00:48:17):
**Gaurav Misra** (00:50:00):
100%. Getting things in people's hands, trying it out. Oftentimes, unless you truly try it out, in design, it can, in theory, look good with all the perfect conditions, but when you actually use it, you realize it's actually not that useful, for example. Or when you give it to users. And some of this is intuition, honestly, just like anything else, but there's nothing like getting something in the hands of users at the end of the day.

**Lenny Rachitsky** (00:50:23):
I love how many of these things you brought over to your current company, and I'm trying to think about, one is this idea of just constantly innovating feels like that's informed and tell me what I'm missing, but that feels like that's informed. The ship market will feature every single week. This idea of getting design, starting almost with design versus PM a lot of times. I'm curious why you don't even go straight to prototype in those cases. Is it just the tools aren't there yet or?

**Gaurav Misra** (00:50:47):
I mean, I think our shipping process is fast enough that within a week we can get it out anyways. So that way we just get user feedback, which is even better.

**Lenny Rachitsky** (00:50:55):
And then the other really interesting thing, I'm trying to visualize that triangle of a product team, the triad of PM, engineer, design. Feels like you guys at Snap took the corners, not the corners, the line of that triangle. And you have design engineers. You have design PMs. I imagine engineers were PME already. They're very product oriented PMs. Did you have a function called design PMs? Probably not.

**Gaurav Misra** (00:51:20):
I mean honestly, it's interesting.

**Lenny Rachitsky** (00:51:21):
Sorry, engineer PMs.

**Gaurav Misra** (00:51:23):
Yeah, engineer PMs should be a thing, I feel like, or every engineer should strive to understand the product, right?

**Lenny Rachitsky** (00:51:29):
Yeah. A lot of companies operate that way. Like Stripe, I think they had hundreds of engineers before they hired the first PM because I think the engineers were doing what they did at Snap to do the PM work. So it feels like at your company you don't operate that way. It feels like you have PMs, engineers, designers. Talk about why you decided not to approach things that way.

**Gaurav Misra** (00:51:48):
I do think PM is a very valuable function. I think it may be actually, and maybe I'll get roasted for this, but I think at the end of the day, not hiring PMs at Snap might've been one of those decisions where it actually succeeded despite that and because someone needs to do that work. If you don't have enough people to do it, then nobody truly owns it and then it doesn't really happen. Or if it doesn't happen, no one's responsible, which is not the right structure you want in an organization. So I think though, that being said, there was something unique to be said about what if a designer had the PM mindset.

**Gaurav Misra** (00:52:29):
It's actually the same idea as what if an engineer had the PM mindset and then you get even crazier. What if the PM had a design and engineering mindset? I think all we're talking about is everybody truly understanding all the functions that they're working with. Having a fundamental, broad understanding of the functions they're working with.

**Gaurav Misra** (00:52:48):
At Captions, we're actually going even one step further than that. Why shouldn't the PM understand marketing? I think that's actually the biggest opportunity for PMs to understand is how do we actually find the users who have this problem? I think that's a big part of solving the problem. I have a unique take on this in terms of I actually think PMs should own all the way to marketing in a way. And the reason is that if you think about marketing, it's expanding the surface area of the product, right? It's like search marketing is just placing a button to your product in Google.

**Gaurav Misra** (00:53:28):
Facebook ads is just placing a button to your app in Facebook. It's almost like you work at Facebook. You work at Facebook, you have a button in the app somewhere, you make a specific thing and people show up. The funnel begins there and you have all the metrics all the way from the beginning, all the way from when the user tapped on the button in Facebook and then they went down all the steps and then they landed on some onboarding screen and they did the thing, they used the application.

**Gaurav Misra** (00:53:56):
That's where the journey begins. And all of that is, in a way, it's a product. It's the same skillset. Understanding users from that point on is I think that's fundamental. How do we not do that today? We should be. So that's how we think about stuff. But I think the core idea is that every function should understand every other function deeply as much as possible and maybe even to the level where they can operate in that function. And that just increases the likelihood that all decisions being made in the company at the micro level will be optimized for all possible parts of the funnel that different people are essentially looking after. That's something we think about quite a bit.

**Lenny Rachitsky** (00:54:44):
I completely agree with that take. It's interesting that at Airbnb, Brian was famous for changing the titles of all product managers to product marketing manager for exactly this point because he's like-

**Gaurav Misra** (00:54:56):
Makes sense.

**Lenny Rachitsky** (00:54:57):
... you should be doing the marketing, you shouldn't just be building the thing. And to me, I've always assumed as a PM, your job is for this thing to grow and to get adopted and be loved.

**Gaurav Misra** (00:55:07):
Of course.

**Lenny Rachitsky** (00:55:07):
So it's interesting people don't already think of it that way.

**Gaurav Misra** (00:55:10):
I agree.

**Lenny Rachitsky** (00:55:11):
But obviously, it's hard to learn the skills of being awesome and paid growth and SEO and product marketing, messaging, positioning, but I completely agree. That's such an important element of building a product. You're not just building a thing, hope it works. Goodbye. So I love that that's how you think about it. And so I guess when you hire PMs, it sounds like you look for marketing instinct and some experiences.

**Gaurav Misra** (00:55:31):
100%. And at least the ability and instinct to be able to learn it.

**Lenny Rachitsky** (00:55:38):
Yeah. Okay, so I'm going to share one other thing that I thought as you were talking that I think is really interesting and it comes up a bunch on this podcast and this connects back to Ev and what we can learn from his success. So Patrick Olson once tweeted this tweet that has really stuck with me, which is it was around user research and the way he described it is user research isn't go to user research that informs what you build and then you build that. It's instead you do user research, it informs the mental model you have as a leader, a product builder of what your customers need and what pains they have, and you adjust that model in your head and then that's how you decide what to build. And it feels like Ev is very much that. His head was learning what people need, teens in particular, and it just worked.

**Gaurav Misra** (00:56:28):
Yeah, I think it's very spot on. I would say though Snap didn't like user research as a function for the longest time. I think there was one user researcher in the company until, again, 5,000 employees, the post IPO basically. But I think the people that were making a lot of the product decisions and the CEO himself, of course, were very steeped in how the user behaves and how they operate. They understood that.

**Gaurav Misra** (00:56:56):
I do think Snap also had a unique way of thinking about how to determine if a product is within scope or out of scope of what their mission was. And I think a lot of companies use the cyber framework and we try to as well, but essentially, the idea at the core was that they want to enable private sharing in a safe way. So I think that makes it clear that certain things just are out of scope for Snap.

**Gaurav Misra** (00:57:28):
It's actually one of the reasons why Snap wasn't the company to discover "short form video," TikTok style stuff because it was just against the nature of the company to even try something. It was against the mission of the company. Public sharing means possibly bullying and bad behaviors, which is exactly what Snap was trying to avoid. We don't want those behaviors to develop on the app. So for example, on Instagram stories, you can share somebody else's stories to your followers. I can take your story and share it to my followers. You can't do that on Snap.

**Gaurav Misra** (00:58:02):
And there was a discussion about should we do this? No, because it can enable bullying. Essentially, you're not consenting to your thing being shared to my followers and that's essentially bad. So a lot of it was done based on this type of pillar-based thinking of this is our mission, this is what we're trying to do, does it fit within or is it outside? If it's outside, we don't do it no matter what the cost of it is, no matter how exciting it is.

**Gaurav Misra** (00:58:30):
And even on Spotlight, the big challenge was like how do you take something like that and put that inside the Snap mission? So that was something we worked on quite a bit. Yeah, I mean I think there's tons of stories about earlier versions. I mean Snap almost had essentially what is TikTok earlier than TikTok existed and it died out because it didn't align with the mission essentially, but happy to get into it.

**Lenny Rachitsky** (00:58:56):
Yeah, that actually would be really interesting, because interesting that these things are important. It's important to have these clear values in the mission of the company and to not focus on things that are outside that. And then you hear these stories of they had TikTok potentially. So yeah, whatever you can share there, that'd be awesome.

**Gaurav Misra** (00:59:12):
Yeah, I mean, I don't know if you remember this, but there was this product called Our Stories, and essentially it was MyStory, but it was a public story. And it started off with this idea of campus stories where you can post to your campus and other people can see it. And that actually started creating a lot of virality because essentially people would post. There was viral moments truly where people would post stuff like, "Oh, I think two people fell in love on it or something like that." Those types of things really went viral and it had really good engagement.

**Gaurav Misra** (00:59:47):
But at the end of the day, the problem was that we were against algorithmic essentially ranking of those types of things. So there was a curation team that was looking through every single one so that there's no negative behaviors happening essentially on the app. That was just not scalable. Even though it had really high engagement and was doing well, it just wasn't feasible to have a person looking at every single thing posted to determine whether it's appropriate or not.

**Gaurav Misra** (01:00:15):
It ended up dying out, but it looked like what was an early version of TikTok before it had launched. So I think in a way though it was a good thing because I think Snap does have a mission and I think it is solving a problem. I do think there is a bifurcation of social media at this point. There is what you traditionally think of as social networking where you share things with your friends. And by the way, remember the days where that used to be the way that apps would go viral. You would share things with your friends and then they would share with their friends and everybody was worried about friend sharing and how do you send to a friend and can I text message my friend or whatever.

**Gaurav Misra** (01:00:59):
That time is over. Virality now happens through a completely different mechanism. It happens through essentially algorithms that are deciding whether your piece of content is worth showing to an arbitrary number of people, and this is the new age of social media. It's TikTok, it's YouTube Shorts and Instagram Reels and so on. And I think actually it's changing the fundamental nature of how people interact, fundamental nature of how things go viral. And I actually think from a regulatory perspective, we should be thinking these as differently.

**Gaurav Misra** (01:01:37):
On one side, you have something where you're deciding who sees something and then on the other side, you have something where the company is deciding which means that it's semi curated, right? It's actually the company's voice. So yeah, I don't know, should Section 230 apply to that? I have no idea. Or maybe not. Maybe we're thinking about this the wrong way, so it should be interesting.

**Lenny Rachitsky** (01:02:03):
Wow. All right. Well, I'm out of my depth on the legality decision, so I'm going to not follow that thread, but I imagine there's something really interesting there actually. So you've been talking about just how much things are changing and I just wanted to follow that thread and specifically, you guys are at the cutting edge of what is possible with AI video.

**Gaurav Misra** (01:02:25):
Yes.

**Lenny Rachitsky** (01:02:25):
It feels like we're approaching and maybe we're there. This world where you have no idea if it's real or AI. I'm curious, first of all, just how far you think we are from that and second of all, the implications on the world where you can just generate any video you want.

**Gaurav Misra** (01:02:40):
It's fundamental. At the end of the day, a time where video images, audio can't be trusted actually hasn't existed for a while. If you think about ... I mean there was a world in the 1800s where there was no video or audio or images and everything was proven by he said, she said for the most part. And it's possible that if everything can be generated and anything can be created and it looks just as real as if it were real and there's no way to tell, then we might actually return to that world where there's no way to prove anything besides physical evidence or he said, she said.

**Gaurav Misra** (01:03:20):
And I think that's scary, but also possibly opens a bunch of new opportunity for someone to figure out how to solve this problem. I think it's going to be a big problem. I do think today, we are almost there in terms of creating absolutely photorealistic video. I mean the very recent models, a very cutting edge is just about ... It feels like a few centimeters away from achieving it, but I do think to fully get there to the point where it cannot be differentiated at all, it's still a couple of years away.

**Gaurav Misra** (01:03:52):
I also think that it is use case driven in a way. I think thinking about Captions for a second, we take a unique view on what type of video we want to focus on. Video generation and text to video generation. If you look at it today, it's all silent video. There's no audio and it's often what you think of as stop video or B-roll, right? You can actually make a movie with B-roll. And a lot of a movie or a TV show or a social media post or an ad actually is dialogue or monologue. That's actually what it is is people talking to each other, to the camera, interacting. That's actually what makes true story.

**Gaurav Misra** (01:04:35):
B-roll is supportive elements that are showing up to set the scene or something like maybe before the scene opens, you see a few shots of New York City or LA or something, and then you jump into the room and now two people are talking. So our goal is to solve the talking video problem. How do we create video where people are delivering dialogue or monologue or things like that? And that's what we focus on purely. And there actually isn't a lot of work happening in that area today and it's not a solved problem. We're getting there, we're getting closer and closer, but today's models actually bifurcate a little bit.

**Gaurav Misra** (01:05:15):
So there's a set of companies today that are able to create these types of what we're talking about is avatar videos. They're using this technology called neural rendering. It's actually not a technology that's affected by the transformer and diffusion model revolution or the large model revolution, essentially. This is a technology that existed separately and it doesn't have anything to do with the AI growth happening right now. It just happens to produce semi-realistic outputs, but it actually stops at some point because it's not clear how it becomes generalizable in every situation.

**Gaurav Misra** (01:05:55):
It has to be trained on people individually. So you might ingest a little bit of video of you and then you can generate you. And so it's a different technology and a different outcome, essentially. And a bunch of companies using this type of model, a bunch of companies are doing general text to video with no audio today. These are large generative models and they have the capability to do more, but that frontier just hasn't been reached yet. I think there's no doubt in anybody's mind on the research side that it is 100% solvable. It's just like somebody has to go do it and we haven't gotten there yet. Nobody has had the time to go and do that yet. So that's where we're at, essentially.

**Gaurav Misra** (01:06:35):
We're working purely on large generative models for talking videos. So that's our core focus. I do think though, from a safety perspective, we have a unique framework or how we think about it. So generally videos divide into two categories. So for us, we think on one side of what is documentation, so this is the type of video that it could be a personal video where you're taking a video with your friends and you're hanging out, you're at a restaurant. It's documenting what happened. You had fun, whatever it was, it's for your memories. And there's a non-personal version of this which is like, oh, it's like a reporter documenting a crime or something that happened or whatever it is and who was involved, where was it? Maybe it was a natural disaster or something, and this is for history. We want to see what happened.

**Gaurav Misra** (01:07:27):
And there's actually no benefit to AI-generated video in any of this. Actually, all of this, it's just negative. It's all negative. If we are generating fake versions of reality to fool people, there's just nothing good about that. And we want to stay away from that, essentially. We want to design products and build products that make it difficult to use for that particular use case, for anything that falls within that. And on the other side, you have what we think of as storytelling.

**Gaurav Misra** (01:07:56):
Now this could be ads, it could be social media posts, it could be TV, movies. All of these things are storytelling. They're designed for entertainment, they're designed for fun. And nobody believes if you watch a Geico commercial, you're not thinking that the gecko is real selling insurance somewhere out there. You know that this is fabricated and it's for entertainment. And same with reality TV even, right? It's called reality TV. It's definitely not reality and social media, ads, all this stuff falls in the category.

**Gaurav Misra** (01:08:28):
And if we can enable more people to tell stories and entertain other people and get their message out there, that is pure positive. This is where we want to focus. And a lot of our effort in the product and design process goes into how do we design products and build products that specifically make it really hard to use on one side and really easy to use on the other side. And that's the real challenge.

**Lenny Rachitsky** (01:08:53):
That's really helpful. Something that I'm really curious about as you're chatting is ByteDance just released a really amazing model. I was actually just looking at it where you put a photo in, I think, and it just creates a video of this person talking in all these different ways. Where does that fall amongst the buckets you just described?

**Gaurav Misra** (01:09:09):
I think that falls exactly in the area that we're in, which is talking people and that's what they're going after as well there. So that's actually one of the first examples of a large model that a larger company has released where it's able to do these dialogue or monologue videos. And I mean you yourself, you've seen it, so I'm not going to describe it too much, but as you know, it's highly expressive. It doesn't look like an avatar video. It looks like ...

**Lenny Rachitsky** (01:09:37):
Yeah, it's wild.

**Gaurav Misra** (01:09:38):
And that's because of the technology that's used is fundamentally different. It's just like this is using a true large diffusion model is what they use. Whereas most companies that are working on avatar technology are actually using something pretty basic in comparison.

**Lenny Rachitsky** (01:09:53):
How long has it been since that Will Smith spaghetti video? Just to give us a reference of how fast things are moving?

**Gaurav Misra** (01:09:58):
Oh my god, it's been so fast, right?

**Lenny Rachitsky** (01:09:59):
I think it's a year.

**Gaurav Misra** (01:09:59):
Amazing.

**Lenny Rachitsky** (01:09:59):
Or is it like two years?

**Gaurav Misra** (01:10:03):
I think it's probably about a year and a half, two years. Right?

**Lenny Rachitsky** (01:10:04):
Wow. We'll link to that video and then you could tell basically that video is the state of the art of AI video one to two years ago. And then we'll link to this other Omni something. I forget what it's called. I'm just showing what it's like today.

**Lenny Rachitsky** (01:10:18):
Geez, Louise. Okay, final question and this is around something that I know you have a really interesting insight on, which is that you see marketing using AI video basically as the final frontier of how people will experience AI is marketing, is seeing it in marketing channels. Talk about why you think that's the case and just what that looks like.

**Gaurav Misra** (01:10:42):
It comes back to what we were talking about before where the reality is that no matter how interesting, advanced and amazing a technology is, science fiction has become reality. We were talking about this. What was literally science fiction on TV is real now and most people still don't even know about it, to be honest.

**Gaurav Misra** (01:11:01):
My parents live in India and they are the only ones in the neighborhood that know about ChatGPT and they write these amazing notes to the community just with all these words. And people are just like, "How did you get so good at writing?" And they're not telling anybody, but there's still a ton of people who don't even know that these advancements have happened. And so adaption is actually much slower, even for the most exciting things. Of course, in tech circles, everybody's talking about it, but the reality is it takes a while to get out there.

**Gaurav Misra** (01:11:33):
And I think for companies that are going to succeed, they're going to have to figure out how to market these products so that they can be the ones to reach all these people that have the problems that they're now able to solve. And we think about that every day. So on that note, as a consumer product, we spend a bunch of time and money on marketing our products, and we often use performance channels and all kinds of things, but about a year ago, we would run AI video in ads and things like that, and we would get all these comments of people being like, "Oh my God, this is so fake. Don't show me this."

**Gaurav Misra** (01:12:08):
And around that time, the technology got just about good enough that suddenly, those comments stopped happening and suddenly, you could get performance that was even better than actually recording with a person because you could just try more things. You could just generate 30, 40 possibilities and one of them would win and it would win more than the one creative you can get from a person. And more interestingly, when you think about localization, you're going to go do that in every language. Once you discover winning creative, now you have to go localize that in every market and rebuild it from scratch.

**Gaurav Misra** (01:12:51):
It's just a ton. And oftentimes it doesn't perform as well because it's been rethought essentially. But we found that just translating it with AI was able to get performance almost as good as the original, in the original language. So this is going to fly to the entire market. I think wherever there's dollars to be made, saved, it is inevitable. It will be consumed and it will very quickly be a lot of social media.

**Gaurav Misra** (01:13:21):
I mean, you could imagine a social network of the future where, and this is dystopian by the way, so watch out. You could imagine a social network of the future where all content is generated. None of the people are real. I mean, the algorithm isn't tailoring whose content to show you, but it's purely generating content that is completely catered to you, with people and everything completely catered to you. I don't think it's out of the question. It almost seems inevitable in a way, but that's not too far away, I think. That's actually very possibly real in five years or something like that.

**Lenny Rachitsky** (01:14:00):
What I'm imagining, because it's hard to imagine a social network where it's people because usually we want to know who these people are. I don't care random sharing status updates, but I can see a TikTok that is all AI.

**Gaurav Misra** (01:14:11):
Exactly, exactly.

**Lenny Rachitsky** (01:14:13):
Wow, just content tuned to your loves and interests.

**Gaurav Misra** (01:14:17):
Exactly.

**Lenny Rachitsky** (01:14:17):
And just random videos. Wow.

**Gaurav Misra** (01:14:21):
Yep. Because do you know, you see a TikTok feed, you don't even know who's real or not today, right? It's not like we-

**Lenny Rachitsky** (01:14:27):
Right. That's how I would approach it. I would just join TikTok and start uploading videos that are AI generated.

**Gaurav Misra** (01:14:32):
Exactly.

**Lenny Rachitsky** (01:14:33):
And then build a whole network of that. Oh my god, the future is wild.

**Lenny Rachitsky** (01:14:37):
Let's go to failure corner. Something that I try to do with this podcast is share moments where things didn't go well. There's all these stories of everything's going great all the time. All this foundries killing it, building a billion-dollar company. Oh, so awesome. But they don't know all the things that go wrong. So let me ask you, is there a story you can share of when things didn't work out? When you failed?

**Gaurav Misra** (01:14:59):
At the beginning of the company, we actually had a bunch of time where we spent figuring out what we wanted to do, and I think it's an unconventional story almost in a way because we started off the company, the first thing we did was build the Captions app. We launched the app. That was the first thing we did. Took two days to build it. We put it out there and it immediately took off. It was absolutely shocking because I built it on a weekend. We put it out there, I called my co-founder on Monday. I'm like, "It's at the top of the app store. We're getting like 600 videos a day."

**Lenny Rachitsky** (01:15:31):
Top of the app store, holy shit.

**Gaurav Misra** (01:15:33):
And we didn't do anything to enable that. It just happened on its own. It was almost anti-climactic in a way because we thought it would be a lot more time spent figuring out the product before that would happen. And so it felt like, "Wait, this can't be it, right? It can't be this fast. How did this happen?" So we got distracted because of that, because we were like, "Oh, okay. Well, maybe ... This is cool. It'll work. That's great, but we got to figure out what the product is."

**Gaurav Misra** (01:16:07):
And so we spent at least a year, year and a half thinking about building social networks and all kinds of things when we should have been working on Captions because there was product market fit there. And how we figured that out is Captions was sitting on my personal account, so I wasn't checking that a lot. About a year and a half into the company, as we were working on other projects and stuff, I went back to my personal account, just opened it, and I saw that there was $500,000 in there.

**Gaurav Misra** (01:16:38):
I looked at a chart and it was just growing. The revenue was just growing completely on its own. No employees, no releases, no bug fixes, no customer support. There was like 2,000 open support tickets that were unanswered for a year and a half, and great reviews. It's just going completely on its own. And so that was a clear sign to me. It was like, "Oh my God, you should have been working on that. That product works."

**Gaurav Misra** (01:17:05):
And so we immediately had a meeting. I mean, it was tough to figure out what the right path was at that point because we'd invested so much time in other things as well, but reset, and we got back on the track with Captions, and literally as soon as we started releasing the first features into it, it blew up. What looked like a vertical line at that time became a horizontal line, and the new vertical line was so vertical that the old vertical line became a horizontal line, essentially. And it's continued since then, which is crazy. So we basically wasted about a year and a half.

**Lenny Rachitsky** (01:17:42):
I love that new way of thinking about a hockey stick moment where not only is it going vertical, but the rest of the chart is now just flat along the bottom of the axis.

**Gaurav Misra** (01:17:49):
Exactly. Yeah.

**Lenny Rachitsky** (01:17:51):
For people that may not know what Captions is, I try to describe it at the beginning and we'll link to it and stuff, but basically, the reason you thought it was nothing is it just adds captions to a video that you record.

**Gaurav Misra** (01:18:01):
It does.

**Lenny Rachitsky** (01:18:01):
Added captions.

**Gaurav Misra** (01:18:02):
Exactly, yep. So I think we wanted ... Our thought was we're going to build a social network, but first we got to build a creation tool for the social network. And we knew that we wanted to use AI to create video, and it seemed obvious that, "Oh, speech to text, a solved problem, we should start with that." So that's why we decided to start with Captions because it was a solved problem at the time. What was funny is that once GPT and stuff started coming out, a lot of the things that were unsolved became solved very quickly. So timing was almost perfect.

**Lenny Rachitsky** (01:18:37):
And that aligns to something you shared earlier. Just so many of these problems that were not yet solved are now possible, and the companies that are in the right place at the right time benefit greatly who've been just waiting for this part.

**Lenny Rachitsky** (01:18:50):
The other thing that I think is interesting about that story is you try to build a social network. I think it was around high schools and things like that. As we've seen, it's very difficult to build a new social network. So let me just get your sense. Do you think it's possible for somebody to come around and build a new, the next Facebook, the next Snap, the next whatever?

**Gaurav Misra** (01:19:09):
I think it's definitely possible. I do think ... Let me tell you something crazy, actually. The social network that we had at the time, we actually remove it from the app store, so it's not available anymore. But til today, there are people, there are thousands of people that are using it, posting on it, and all the different things, which actually speaks to the power of the social network in a way. It is hard to create and hard to kill. I mean, I think X is actually a great example of that too. A lot of movement happened there and it continues to work, I guess somehow. So testament to that.

**Lenny Rachitsky** (01:19:49):
The power of network effects, especially someone once described this so well, they're like Twitter/X. They changed the brand, they changed the team building it. They changed the URL. Like everything changed about it except the network effect of the people in it.

**Gaurav Misra** (01:20:07):
It's true. It's true.

**Lenny Rachitsky** (01:20:09):
I just saw a story that they're making billions of dollars. He's actually turned it around. It's actually becoming a really profitable company.

**Gaurav Misra** (01:20:16):
Wow.

**Lenny Rachitsky** (01:20:17):
Yeah. It just came out the other day. So Elon did it. Well, with that, we've reached our very exciting lightning round. Are you ready?

**Gaurav Misra** (01:20:25):
I'm ready. Let's do it.

**Lenny Rachitsky** (01:20:27):
What are two or three books that you have recommended most to other people?

**Gaurav Misra** (01:20:32):
I have to say here that I actually don't read books. It's actually something that I decided on purpose where I decided I don't want to build my skill in reading, and I want to build it in listening and watching instead, because I think that's the future.

**Lenny Rachitsky** (01:20:49):
I love how intentional that is, and I love how it's a really cool way of saying, I don't read books. The future isn't reading, but I love that you have books behind you, so [inaudible 01:20:58].

**Gaurav Misra** (01:20:58):
I do. Yeah.

**Lenny Rachitsky** (01:20:59):
[inaudible 01:20:59]

**Gaurav Misra** (01:21:00):
[inaudible 01:21:00] didn't read, they're back there.

**Lenny Rachitsky** (01:21:03):
That's funny. Okay, cool. I want to ask more questions, but I'm going to keep going. Lightning around. Speaking of watching and listening, do you have a favorite recent movie or TV show you've really enjoyed?

**Gaurav Misra** (01:21:12):
I like Silo and Severance. I mean, obviously, I think everyone's watching these. There's a book around Silo too.

**Lenny Rachitsky** (01:21:18):
I read that. I read all of them. There's three of them.

**Gaurav Misra** (01:21:20):
There are.

**Lenny Rachitsky** (01:21:20):
It sucks to watch the show because you know all the tricks that are about to happen, and I'm just like, "Why am I watching this? I know where this will go.

**Gaurav Misra** (01:21:27):
Yeah. I mean, for what it's worth, it does seem like the show is going on a slightly different path.

**Lenny Rachitsky** (01:21:31):
It is. That was also what annoyed me. Just like, "What the heck? This is made up. All is made up shit." I don't like that when I watch the show. So two reasons I'm not watching it but [inaudible 01:21:40].

**Gaurav Misra** (01:21:39):
Don't worry about it. I didn't actually read the book. My wife read the book and then she told me the story.

**Lenny Rachitsky** (01:21:44):
Okay, okay. I was worried. I was worried. Okay, cool, and Severance. Okay, great. I love Severance.

**Lenny Rachitsky** (01:21:51):
Next question. Do you have a favorite product you've recently discovered that you really like?

**Gaurav Misra** (01:21:55):
My favorite product, honestly, is Linear. I'm not going to lie, just because it's so well-designed and it's so easy to use. I also like Superhuman. I mean, these are obvious answers, but I do use these things every day and it's hard to create products that you use every day and don't hate. So props for them.

**Lenny Rachitsky** (01:22:13):
Cool. I haven't announced this on the podcast yet, but this is a good time, whoever's listening right now, is I just launched a bundle where if you become a paid subscriber to my newsletter, you get, listen to this, a year free of Linear and Superhuman and Notion and Granola, which is incredible AI app for note-taking and Perplexity, Perplexity Pro.

**Gaurav Misra** (01:22:35):
Ooh. Nice.

**Lenny Rachitsky** (01:22:36):
$2,000 in value for the price of my newsletter, 200 bucks, going to that.

**Gaurav Misra** (01:22:40):
Damn, that's real value.

**Lenny Rachitsky** (01:22:43):
It's an unbelievable deal, and it's a no-brainer at this point to buy a subscription, but this isn't an ad for my newsletter. I'll keep going.

**Lenny Rachitsky** (01:22:50):
Next question. Do you have a favorite life motto that you often find yourself coming back to, sharing with friends and family in work or in life?

**Gaurav Misra** (01:22:57):
I actually learned this because someone else told me that I keep repeating this thing, but I have this framework of how I want to operate at work, basically. Right? I think I love to compete and to win at the end of the day. And I think that to win, you have to be the best. But I also think the easiest way to be the best is to be the first, and that actually is key.

**Lenny Rachitsky** (01:23:25):
And so is the motto the easiest way? Is that the-

**Gaurav Misra** (01:23:27):
That's it. The easiest way to be the best is to be first.

**Lenny Rachitsky** (01:23:31):
Be the first. Interesting. Okay. I have to resist following threads here because I want to make this lightning round.

**Lenny Rachitsky** (01:23:37):
Okay. Final question, just for fun. What's the coolest, most wild AI video you've seen recently? Is there one that comes to mind of like, "Wow, that was something"?

**Gaurav Misra** (01:23:47):
I mean, honestly, I got to say the OmniHuman stuff was pretty cool.

**Lenny Rachitsky** (01:23:52):
The ByteDance video that we talked about.

**Gaurav Misra** (01:23:52):
Yeah, exactly. I mean, the broccoli talking. I don't know if you saw that one. There was a little broccoli delivering a little speech.

**Lenny Rachitsky** (01:24:01):
Interesting.

**Gaurav Misra** (01:24:03):
Yeah, it looked like it was animated by an animator.

**Lenny Rachitsky** (01:24:08):
Just imagine being a kid these days and just seeing stuff like that.

**Gaurav Misra** (01:24:11):
I think you're probably just used to it, right? You're just like, this is just normal.

**Lenny Rachitsky** (01:24:13):
Mm-hmm. It's just like we were saying, AGI is just going to come around.

**Gaurav Misra** (01:24:17):
Exactly.

**Lenny Rachitsky** (01:24:17):
All right. Cool. What's for dinner? Cool. That's great.

**Gaurav Misra** (01:24:20):
Yep.

**Lenny Rachitsky** (01:24:20):
Amazing. Gaurav, this was incredible. It was so insightful on so many levels. Two final questions. Where can folks find you and what you're building if they want to learn more? And then how can listeners be useful to you?

**Gaurav Misra** (01:24:30):
Awesome. Yeah, I mean, definitely find me on LinkedIn. That's where I live most of the time. My DMs are open, et cetera, et cetera. So feel free to send me a message. And I think what'll be useful, I mean, we're building out our early product and design team, so if AI video is interesting, if consumer apps are interesting, now's the time to join. We're really small, early, we work together across the team, so there's going to be no better time to join, basically.

**Lenny Rachitsky** (01:25:01):
And you get to ship a marketable feature every week.

**Gaurav Misra** (01:25:03):
Exactly. I mean, that's the PM's dream. Think about it. Right?

**Lenny Rachitsky** (01:25:06):
The PM's dream. Yeah, I like that that's a filter. The people that get excited about that, great fit. The people that are stressed out by that, not the place to be.

**Gaurav Misra** (01:25:15):
Exactly.

**Lenny Rachitsky** (01:25:17):
So awesome. All right. Gaurav, thank you so much for being here.

**Gaurav Misra** (01:25:20):
No, thank you. Appreciate it.

**Lenny Rachitsky** (01:25:22):
Bye everyone.

**Lenny Rachitsky** (01:25:25):
Thank you so much for listening. If you found this valuable, you can subscribe to the show on Apple Podcasts, Spotify, or your favorite podcast app. Also, please consider giving us a rating or leaving a review as that really helps other listeners find the podcast. You can find all past episodes or learn more about the show at lennyspodcast.com. See you in the next episode.

---

## [15/15] A better way to plan, build, and ship products | Ryan Singer (creator of “Shape Up,” early employee at 37signals)

**Ryan Singer** (00:00:00):
I often use this analogy of if you're doing a home renovation, you can have the most beautiful rendering of the new bedroom and we're going to have these lamps on the side of the bed that are coming out from the wall. But if you haven't checked if there's electricity in that wall there or not, it's going to drastically change the cost and the time and everything.

**Ryan Singer** (00:00:16):
What we need to do in a shaping session is we come out with some kind of diagram where engineers, product and design, they're saying, "We understand that." So the first thing is we are not going to start something unless we can see the end from the beginning. We're not going to take a big concept and then say, "What's the estimate for this thing?"

**Ryan Singer** (00:00:37):
We're going to go the other way around and we're going to say, what is the maximum amount of time we're willing to go before we actually finish something? How do we come up with a idea that's going to work in the amount of time that the business is interested in spending?

**Lenny Rachitsky** (00:00:54):
Today my guest is Ryan Singer. Ryan was one of the first few hires at 37signals, and through his experience of building Basecamp and 17 years of building product at 37signals, he wrote a book called Shape Up, which shares a very different approach to building software.

**Lenny Rachitsky** (00:01:10):
Appetites instead of deadlines. A big focus on bringing design engine product together into a room to shape the plan versus writing long PRDs or trying to finalize designs before you start building.

**Lenny Rachitsky** (00:01:22):
I've noticed more and more teams adopting the Shape Up method, and especially with AI starting to change how we work and build product, there's this shift coming in how product teams will operate. And so I thought this was the perfect time to do a deep dive into the Shape Up method.

**Lenny Rachitsky** (00:01:37):
This episode is basically going to give you everything you need to give Shape Up a shot on your team or at your company to see if it fixes the problems that you're having shipping great products.

**Lenny Rachitsky** (00:01:46):
A big thank you to Des Trainer, Bob Moesta and Chris Speck for suggesting questions and topics for this conversation. If you enjoy this podcast, don't forget to subscribe and follow it in your favorite podcasting app or YouTube.

**Ryan Singer** (00:04:46):
I am really happy to be here. Thanks a lot.

**Lenny Rachitsky** (00:04:48):
I think this is going to be a legendary episode. There's a lot of interest these days in different ways of working, especially ways that are Agile and SAFe and Scrum and all these ways that people hear about working. Especially in this world of AI where everything's just changing. It feels like there's just an increased interest in exploring different ways of working and specifically it feels like there's been a rise in interest in Shape Up the stuff that you talk about.

**Lenny Rachitsky** (00:05:14):
So, I am really excited to basically help people understand what is this way of working, is it right for them? What are ways to start implementing it? What are maybe some pitfalls you may run into? And as much as possible get into a lot of real talk about how things are actually going on product teams that people often don't like to hear.

**Lenny Rachitsky** (00:05:31):
So, first of all, have you also seen this increased interest in Shape Up?

**Ryan Singer** (00:05:38):
Yeah, I think it's interesting that we're talking now. I mean, the book came out in 2019 and it's, I've been hearing more and more like, "Oh, we know somebody who's trying it or we're hearing it when we go talk to other companies." So, I think, it's a wave that's slowly building.

**Ryan Singer** (00:05:57):
And it's funny, when it came out, I even tried to have an online forum to get everyone who's interested to talk together and what I started to learn pretty early on is that people don't like to talk about their struggles shipping.

**Ryan Singer** (00:06:13):
Especially CPOs and CTOs don't like to go on a public forum and say, "Our company isn't shipping or our engineering team is stuck, or our team is always lost in the weeds." That's not an easy community topic on an online forum.

**Ryan Singer** (00:06:28):
So, I think, there's also some reasons why it's been word of mouth slowly gathering steam.

**Lenny Rachitsky** (00:06:33):
That's something I struggle with on this podcast. As you said, it's a product and product teams don't want to be sharing when things aren't going great. That's why I introduced Failure Corner on the podcast where it's like, "Okay, but tell me a time things didn't go well."

**Ryan Singer** (00:06:46):
Oh yeah, that's great. Yeah, because it's so hard to get to that, right? And it's not all just this golden path, rosy, where we're all shipping beautiful, meaningful things all day.

**Ryan Singer** (00:06:55):
It's a hard business and there's no perfect school either that produces expert product managers and CPOs and CTOs and stuff like that. So we're all trying to figure this out and we don't have a lot of sources, so there is a lot of struggle.

**Lenny Rachitsky** (00:07:12):
And there aren't many options for how to build product. All people really read about is Scrum/Agile/SAFe as they scale and then there's Waterfall, which, "No, I never do Waterfall." Then there's the start up way of just ship and maybe one or two week cycles and then there's Shape Up, so it feels like it's one of the rare other options that exists. And so-

**Ryan Singer** (00:07:33):
That's one of the things I've been hearing. It's, I hear like, "Oh, we thought there was only Scrum or Kanban, and then we heard there was this Shape Up thing. What's that?

**Lenny Rachitsky** (00:07:42):
And I think it's always been connected to Basecamp. We're going to talk about that. Just, it works for companies that are nothing like Basecamp. Maybe just touch on that briefly.

**Ryan Singer** (00:07:53):
Well, I mean, that came as a surprise to me. I mean, when I wrote the book, I had been in Basecamp at that time, I think 15 years, and I actually didn't even know the outside world. I mean, it was Jason's idea to even write the book. Because he said, "Look, a lot of people are going to want to know about this. A lot of people are struggling." And I'm like, "Well, okay."

**Ryan Singer** (00:08:16):
I knew our inside story of we had some growing pains and we had to be able to formalize the way that we were working and shipping so that as we brought new people in that they could participate in that and we could stay fast. So, I knew our internal struggles, but I honestly didn't know anything about the outside world.

**Ryan Singer** (00:08:33):
And it was only after the book came out that it gave me this excuse to start talking to people from all kinds of different companies, and it's been really interesting. There are some really amazing cases of companies of very different characteristics from Basecamp, like VC funded, significantly bigger, very different pressures, different team structure, different skills who are doing it.

**Ryan Singer** (00:08:58):
At the same time, there is also a lot of questions that are coming in my way because honestly, there are so few teams that are structured just like Basecamp, that there are a lot of gaps in the book of like, "Well, what about this, and what about that, and how do we do that in our situation?"

**Ryan Singer** (00:09:11):
So, a lot of my focus today is actually closing those gaps and helping people figure out how can I make this work for me or for our team?

**Lenny Rachitsky** (00:09:19):
And you specifically told me, you just now call it instead of Shape Up it's Shape Up In Real Life or Shape Up For Real Life.

**Ryan Singer** (00:09:24):
Yeah. Well, my wife heard me saying the same thing over and over again on every phone call. And she's overhearing me and she's like, "You have to make a course, you have to do something. You always are saying the same thing."

**Ryan Singer** (00:09:37):
So then this led to this course that we made, which is called Shaping and Real Life. And well, yeah, the idea is the real life part, right? How do I make this work if my designers don't code? It's very contentious to get engineering time. You know what I mean? When there's all these different pressures that Basecamp didn't have.

**Lenny Rachitsky** (00:09:56):
We're going to get into the nitty gritty of how this actually works and the key elements, but can you just give a very short overarching summary of how Shape Up is different from how other product approaches are?

**Ryan Singer** (00:10:07):
I think the way it's different starts with how the way we were working was a little bit different. So, I started working with Jason and David on the first version of Basecamp, which was the flagship product of 37signals back in 2003. We were a team of three.

**Ryan Singer** (00:10:25):
And, I mean, I think, it's for any really small team, when you're just starting out, you don't need a process. You don't need a way of working. It just happens organically because you're together. You don't have to explain it to other people, it just happens on its own, right?

**Ryan Singer** (00:10:40):
But there was always this really intense urgency from both Jason and David, "We've got to get to something we can ship. We have to finish this and move on. We have to get to something that's done." There was just no tolerance for big things that got fuzzy and started to drag. There was always this sharpening to get to what is this thing really and when are we going to get to the end soon?

**Ryan Singer** (00:11:03):
And on top of that, even when we were building V1, David wasn't actually full-time as our only technical person. He was programming 10 hours a week. So, we had this really intense pressure of how can we really use David's time well?

**Ryan Singer** (00:11:20):
We don't want to ever give something that this is the thing we want to build, and then it turns out not to be what we want and we have to throw it away and then come back again or you know what I mean? Those bad cycles of waste.

**Lenny Rachitsky** (00:11:32):
Let me actually ask about this because this is really interesting. So this is DHH. He was working part-time when he started 37signal.

**Ryan Singer** (00:11:40):
10 hours a week. Yeah.

**Lenny Rachitsky** (00:11:41):
Was he working on Ruby basically and that whole thing?

**Ryan Singer** (00:11:45):
Well, Rails came out of the first... So, he told Jason, "I want to try building this in Ruby," because before they had done some collaboration and David had done things in PHP before that and he had this new idea, he wanted to try Ruby, this language he fell in love with.

**Ryan Singer** (00:12:04):
And then the framework, Ruby on Rails, he ended up releasing that after Basecamp was standing. Because it was extracted from the things that were necessary to give V1 of Basecamp to stand up.

**Lenny Rachitsky** (00:12:16):
So that's what he was doing the rest of his time instead of-

**Ryan Singer** (00:12:19):
I don't know. I don't know what he was doing the rest of his time.

**Lenny Rachitsky** (00:12:19):
Probably something great. Something-

**Ryan Singer** (00:12:23):
But he's always been like that. He's always doing something interesting. He's either racing or who knows what, you know what I mean? But all I knew was that we got 10 hours of that time.

**Lenny Rachitsky** (00:12:33):
Yeah, I love that that was a constraint to design a way of working that uses engineering time most efficiently.

**Ryan Singer** (00:12:39):
Yeah, I mean, put that together. So, David's constraint of 10 hours a week and then Jason has this, I mean, I think, many really successful founders and especially CEOs have this thing, it's like all they want to see is movement. You know what I mean? Forward, forward, forward.

**Ryan Singer** (00:12:55):
So when do we get to see it? When do we get to try it? When do I get to put it into somebody's hands? So that combination, there was just so much urgency even though there was no outside pressure. You know what I mean? It was completely just let's say cultural energy of how do we keep getting somewhere and getting to something that we can celebrate and get excited about.

**Lenny Rachitsky** (00:13:17):
I love that. That's an attribute, I think, of a lot of successful founders. So that makes sense to hear that.

**Ryan Singer** (00:13:22):
Totally, totally. And that's why where I come back to you, this is the part of the story where, I think, so many companies would say, "Yeah, I know that experience," right? Because, I think, that's probably the seedling of, as you said, of successful companies.

**Ryan Singer** (00:13:34):
Is that combination of urgency and also that those guys were so talented and that they had a clear vision of what they wanted to do and all of that. It's this amazing time actually, these early days.

**Lenny Rachitsky** (00:13:44):
Is there anything more to the backstory that's important to share or super interesting?

**Ryan Singer** (00:13:47):
Yeah, I mean, the other big piece of it was, so Jason and I were this product, what do you call it? The two thirds. So, I was doing UX at the time and I was doing hands-on coding as well. So we're very, very integrated. Everybody does a little bit of everything. All of us were coding.

**Ryan Singer** (00:14:09):
Jason was in the actual app templates as well doing HTML and CSS to do the views. He's doing hands-on design. We're all very much connected with why are we building this? What is this? David's doing the bulk of the programming, and Jason and I were having these little sessions.

**Ryan Singer** (00:14:28):
These little sessions where we would really figure out what the idea was and there would be this moment where you would have a few strokes of this Sharpie pen on a big pad of paper and all of a sudden you'd be like, "Oh yeah, that's the idea. That's the thing we want to go try to build."

**Ryan Singer** (00:14:46):
And for me, those sessions with Jason, they were these short, very, very intense sessions where you're trying to crack the nut together. Where's the idea? What's the concept? How can we go... What's this thing that we're going to go and 10 hours later, right, David's going to come back and we're going to be like, "This is awesome. This thing works and it does something we're excited about," right?

**Ryan Singer** (00:15:10):
That was really the seedling. I mean, actually that continued over years and years and years, those sessions. And that's the seedling of this word in the book shaping. What does it mean to do shaping? It wasn't sitting alone writing a document, it wasn't making a bunch of requirements.

**Ryan Singer** (00:15:31):
It wasn't making a beautiful Figma file to represent a concept that could maybe be a feature. It was this super intense, really exciting collaborative, "What about this? What about that? Oh, maybe this." So that was a really big part of how we worked also.

**Ryan Singer** (00:15:51):
Very intensely collaborative sessions to figure out what the idea is and getting it sharp enough and crispy enough that we could very confidently get a yes from David. That he would know exactly what it is and what it means and come back.

**Ryan Singer** (00:16:06):
It would be what we pictured and it would work the way that we hoped so that we would keep going and we wouldn't have to reverse or go back to the drawing board.

**Lenny Rachitsky** (00:16:15):
What it sounds like is essentially you're trying to maintain the startup way of working as the company grows. Everything you're describing is how it feels to be at a startup, and this is a method to keep that. Does that sound right?

**Ryan Singer** (00:16:28):
That's exactly what became Shape Up, was how do we hold onto that as much as possible? I mean, one big ingredient, we had an advantage also, which was that Jason and David hired so deliberately slowly, and this is a fortunate side effect of the fact that they didn't take investment money.

**Ryan Singer** (00:16:51):
So there was never that moment of now's the moment when we grow. It was always one person. And then the organism adapts. One more person. And so, this natural way of working, it was organically spreading. There were, I think, maybe 10 years before we had the first, "Wait a minute, what just happened?" That project didn't go well, that's not how things normally run.

**Ryan Singer** (00:17:26):
Of course there were always ups and downs, but it was about 10 years later when we had the first project. I mean, I remember the project. I remember being at the end of... It was at that time already, it had been maybe six weeks or seven weeks. We hadn't yet completely locked the six-week thing that went into Shape Up.

**Ryan Singer** (00:17:47):
And I remember we had a review session and there was a fairly new person who's doing half of the work on that team, and we had the review session and it was like instead of, "Oh, look, this is about ready to ship." It was like, "There are a lot of open questions here."

**Ryan Singer** (00:18:06):
"And not only are there a lot of open questions here, we're not getting quick answers as we're asking." And what we're starting to realize is like, "Oh, not only is this not going to ship, but we can't even see the end of this."

**Ryan Singer** (00:18:26):
That was one of those moments where you're like, "Oh, this isn't going to automatically, organically just keep it spreading as we hire forever." You know what I mean? We did reach a point where it's like, "Oh, we're going to have to figure out when this goes well, why does it go well and what do we do differently and how do we formalize that so it's reproducible as we keep onboarding more people?"

**Ryan Singer** (00:18:52):
That's actually when Shape Up as a framework started. That's when I really started to lean in and I took over that responsibility of, "Okay, how do I systematize this?"

**Lenny Rachitsky** (00:19:03):
That's a great segue to let's actually talk about how the Shape Up method works. Maybe just at a high level, what are the core ingredients to the Shape Up method?

**Ryan Singer** (00:19:10):
There's basically three maybe big things. So, the first thing is this notion of we are not going to start something unless we can see the end from the beginning. So, we're not going to take a big concept and then say, "What's the estimate for this thing?" We're not going to say, "Oh, we need to build a calendar and then do a whole bunch of Figma files or write a whole bunch of requirements and then ask for an estimate."

**Ryan Singer** (00:19:39):
We're going to go the other way around and we're going to say, "What is our appetite for this? What is the maximum amount of time we're willing to go before we actually finish something?" And we have that startup moment that we talked about. That moment of like, "Ah, you know what I mean? It works. We got somewhere." At least this, if not the whole project, this meaningful piece, we can literally walk away from.

**Ryan Singer** (00:20:03):
So then what we found was that there was a lot of experimentation. We found that six weeks is the maximum that we can see into the future where we could actually say, "How do we work backward and figure out something we could build in that six weeks and really land it?"

**Ryan Singer** (00:20:21):
That's the first piece. Is working backward from the amount of time we actually want to spend on something and say, "What can we do? What could we shape so that after that amount of time we've gotten to somewhere we want to be?"

**Ryan Singer** (00:20:36):
It's like if you're going to buy a car or you're going to buy a house or you're going to rent a new flat or whatever, you have to have a budget in mind, right? And the budget then is how you choose between all kinds of alternatives and make a lot of hard choices and trade-offs to figure out like, "Well, I want the faster engine, but I have to give this up, or I want it to be fun to drive, but we also need space for longer road trips." You're making all these trade-offs, right?

**Ryan Singer** (00:21:07):
So, this second piece is this work that we call shaping and the shaping work is, how do we actually take this fixed amount of time that we've given for ourselves and vary the scope? How do we come up with a idea, some version of this that's going to work in the amount of time that the business is interested in spending?

**Ryan Singer** (00:21:34):
So, this is those creative sessions that I was talking about where we're jumping all over the room in front of the whiteboard and getting to an idea. And there, the really key thing is that we're getting to an idea where we can see the idea. We understand why we're doing this.

**Ryan Singer** (00:21:52):
We're wrestling with the problem and we're wrestling with the solution until we have an idea that we can actually say, "This is what we want to go build." So it's not just calendar or dashboard or newsletter builder, but it's this idea of how we're going to tackle this problem about the calendar request, right? So that's the shaping.

**Ryan Singer** (00:22:15):
And then the third piece is when we've actually carved out a fixed amount of time, when we've shaped a solution that is from a experience standpoint, from a functionality standpoint, from a technical standpoint, doable and desirable, something that we can make happen in that amount of time, then we can give it to a team.

**Ryan Singer** (00:22:41):
We don't have to do the sometimes called scrum, the paper shredder. That's where you take an idea and then you split it into 100 tickets, right, and you hope that it all glues together still after you've done that step. We don't want to do that.

**Ryan Singer** (00:22:57):
Instead, we want to have a whole idea, give it to a team so they see the whole, they really understand it, right? And then they can come up with their tasks and they can figure out how to track that and break it into pieces so they can actually take more responsibility.

**Ryan Singer** (00:23:13):
And so what we see is way more engagement, especially from the technical team, right? Because instead of, "Here's your ticket," or "Here's your user story," it's like, "Here's the thing you understand, that makes sense, and now you're going to have freedom to figure out how to actually make this a reality."

**Ryan Singer** (00:23:32):
There's going to be a million things to solve in the implementation detail, and now you have a bunch of fun problems and you don't have to keep asking questions to other people to understand what this is or how do I make a trade off or that thing.

**Lenny Rachitsky** (00:23:44):
One of the core elements of this, and I want to confirm is that you can pick and choose these things into your team. You don't need to do this wholesale, correct?

**Ryan Singer** (00:23:54):
You don't need to do any of it. So, this is where it helps to look at what's going wrong and what are we trying to fix? And then what do I want to bring into this, right?

**Ryan Singer** (00:24:08):
And usually, what I notice is that people, they like to start sometimes with, "Oh, I want to give the team, let's say six weeks and I want to give the team more latitude or let's say more creative freedom, that they're going to be responsible in this six weeks to figure out how to make it happen."

**Ryan Singer** (00:24:24):
And usually a lot of the drive for that is, "I'm getting tired of having so many meetings and rituals and things that are not actually working on the problem and doing the work." I mean, especially scrum teams, they often complain about that.

**Ryan Singer** (00:24:40):
So, what they sometimes see in this is like, "Oh, I love this idea that the team is just going to be cooking for six weeks and they're not going to, we're going to meet as needed and we're going to workshop things, but we're not going to be busy with all these rituals all the time," right?

**Ryan Singer** (00:24:54):
Now, the thing that's tricky is that if you want that reality of the team happily buzzing and humming like some happy bee colony for this six weeks, they need to have a lot more clarity around what's the thing that we're solving, right?

**Ryan Singer** (00:25:12):
And so when we start working backward from that, then what we see is that, "Oh, well, if we don't shape better, then the team isn't going to have the clarity that they need to take over that responsibility, so they can make choices and make decisions and make trade offs so that they can get to the end of this thing." And the worst is that sometimes see cases where people are like, "Okay, we're doing Shape Up. So, you guys are going to build the newsletter builder, okay? But you only get six weeks to do it. So use your fixed time, vary your scope and enjoy your responsibility." You know what I mean? Which is just cruel, right?

**Ryan Singer** (00:25:50):
Because I think I'll quote Bob Moesta, who's been on your show a couple times, "You can't put 10 pounds of crap in a five pound bag." So, it's a high academic statement and we can't just take any project, no matter how giant it is, and then throw it at a team and say, "Figure it out and ship something meaningful in six weeks by cutting away scope," right?

**Ryan Singer** (00:26:13):
So, it starts to raise questions about how do we actually decide together what this project is? Do we actually have clarity around what the idea is and what we're going to build?

**Lenny Rachitsky** (00:26:30):
Let me follow up on a couple of the elements. So appetites, I think for any product manager, engineer, designer, anyone that has experienced, "Okay, we estimate this landing page is going to take a couple of weeks. Great, let's work on it," and then it ends up being six weeks can understand why this makes sense.

**Lenny Rachitsky** (00:26:47):
It's just like, "This landing page is not that important to us. Let us just say we will commit two weeks to it. We'll do as much as we can in two weeks and then we move on. And scope is not allowed to go beyond that." Makes total sense. This just makes so much sense as you listen to this, especially for people that have...

**Lenny Rachitsky** (00:27:00):
So much sense as you listen to this, especially for people that have just fallen into the problems of estimates not being accurate. Then there's a six-week element and the key there is your, and this is counter to maybe two-week sprints like Scrum, is that kind of the where this comes?

**Ryan Singer** (00:27:18):
So, actually, it turns out that the six-week is only a maximum and that's really where this number does some work for us. If we think of six weeks as a maximum, that's going to force us to ask some really good questions to ourselves about what piece of this do we really think we can land? Because if you try to say, in six months, we're going to ship this thing, you can't get your arms around all of the problems that have to get solved for an entire six-month chunk of work to actually happen. There are so many unknowns, there are so many ticking time bombs of things that we didn't understand or couldn't foresee, but if we set a ceiling at six weeks, we have a much better chance of, I think that's the size of something where we can actually shape it and surface enough unknowns and reveal that complexity before we're in the middle of it.

**Ryan Singer** (00:28:15):
It doesn't mean that we can't use this technique to do a two-week project, especially if you're on a growth team, you don't want to wait six weeks or, you know what I mean? You're going to have to artificially bundle things together to do six weeks. It's like, look, I've got something I want to ship in the next week and then I've got a thing that might take two weeks after that and then a week after that. So, it's more a question of what we're trying to take on. It's really that upper limit.

**Lenny Rachitsky** (00:28:39):
Okay. So, it could be a two-week cycle and the appetite is-

**Ryan Singer** (00:28:41):
It could be a two-week thing.

**Lenny Rachitsky** (00:28:43):
Cool. So, it's like we're going to build this new landing page, we're going to give it two weeks and then do a shaping session on that.

**Ryan Singer** (00:28:48):
Now, the other side of that is when it comes to feature development or building something that's going to be needy enough to sell, then there's very few things that are going to be a substantial value add to a product that you can do in two weeks. So, then you get into a point where, well now, we're just sprinting and we're just taking one bite after the other. And then that's where we can land in that situation where we feel like, "Ah, I can never see the end of this. I just keep going back and saying, one more sprint, one more sprint, one more sprint." But six weeks is this long enough chunk or sometimes, four weeks that the question is kind of, what's big enough that we can actually get somewhere with this amount of time?

**Lenny Rachitsky** (00:29:32):
And there's an implied element to this that I think is worth highlighting. The whole idea is you commit to the appetite and if you are not on track to hit that instead of extending the date you cut in order to still hit it.

**Ryan Singer** (00:29:48):
This is a tricky one.

**Ryan Singer** (00:29:51):
So, you're right that it's implied, but the thing is, in real life, if you make a commitment and you get alignment that we are going to spend six weeks of engineering time building this thing, if you get to that end of that six weeks and something is going wrong, it wasn't shaped, we can't see the end of it. It's more complicated than we thought. All these different things. And by the way, we can also talk about why those things happen, but when we get in that situation, if we're at the end of the six weeks and it's not looking good, we can't just cut off what we agreed to that made this thing valuable. We can't just cut the scope and say, "Oh, well now, we managed to ship inside of six weeks." That's going to kill everybody's morale. Everyone's going to feel disappointed. We're going to feel like this wasn't really worthwhile.

**Ryan Singer** (00:30:43):
And now, we go into the next cycle with this debt feeling that we didn't actually finish the thing we were supposed to finish, so now, we're like overtime. None of that is good. And if we also go the other extreme and just say, well, should we say in the book, we had this principle at base camp which was this notion of the circuit breaker. If a project is not on track to actually finish after the six weeks, we're just going to cancel it and rethink. Almost no teams have the stomach for that, but the version of that that's more stomachable is look, we can't just cancel the project and then say, "Let's see what comes next." But what we can do is say, "We're not going to keep reinvesting in something that we don't understand."

**Ryan Singer** (00:31:34):
So, let's take this out of build mode and bring this back into shaping mode, which might mean different people, a different conversation asking different questions, doing a different kind of work to suss out what is it that's fuzzy here? What is it that we couldn't see? What do we not understand? How do we get to the clarity that we need, so that we can actually say this thing is going to happen if we give it another whack.

**Lenny Rachitsky** (00:32:02):
I love just how real this approach is not this theoretical. Okay, cool. After six weeks, use just the scope and it's all that's cool.

**Ryan Singer** (00:32:10):
Yeah, you just cut the scope.

**Lenny Rachitsky** (00:32:11):
Yeah.

**Ryan Singer** (00:32:11):
No problem.

**Lenny Rachitsky** (00:32:12):
Shape your gut, put your gut.

**Ryan Singer** (00:32:13):
I've seen some Shape Up adoptions that looked like that by the way, and that's not the way. The shaping step is crucial. And what you mentioned with your landing page example, by the way, it's so seductive because we can imagine, oh, Parkinson's law, right? If you give me six weeks to do the landing page, I'll find a way to use it, but if you give me two weeks, then I'll stop after two weeks. But when it comes to real product work, where there's some functionality that we have to figure out how to make it exist, we can't just cut the scope if we run out of time. So, what it means is that the shaping work is really working hard together to figure out what are the main moving pieces of this thing. How do we narrow down our understanding of the problem and how do we identify what the moving parts are of the solution and what actually connects together for this feature to work?

**Ryan Singer** (00:33:19):
And when we really get to the level where we can say, "Oh, we need to do this, this, and then the engine is going to turn," that's the place where we can say, "Oh, this is well-shaped." And it's a different kind of work. In shaping in real life, we call it, we actually teach it as doing live shaping sessions, and this was how I did it for years with Jason. We'd get into the room and I had both the technical and the UX side, so both sides were represented there in one person in that case, but for a lot of teams today, we actually teach them how to bring the senior engineering person who isn't just senior in title, the one who actually knows where the bodies are buried, how the old stuff works and what's truly possible and what's hard and what's easy in our infrastructure, like the person that really knows.

**Ryan Singer** (00:34:12):
You bring that person together with the product person who deeply understands the backstory of why this is an opportunity and what we're trying to solve with this. And then a designer in the room and they're whiteboarding and wrestling with each other to get to what's a version of this thing that we believe in that's real that we can actually finish in that time.

**Lenny Rachitsky** (00:34:30):
This is great. Let's go one level deeper on this shaping session. So, a few tactical questions. How long are these sessions? It sounds like the people that join are a designer and an engineer and an NPM. So, add anything else there. And when do they happen is at the end of a... Do you call it a cycle by the way or sprint, the six-ish week period?

**Ryan Singer** (00:34:51):
What I actually like is time box actually, because the thing is that some teams need regular cycles because they have parallel teams and they need that cadence in order to reduce management overhead. But if you're small and you only have one or two teams, you might not need to be on a fixed cadence or a cycle plan. You might be able to just set one time box after another. So, the key thing is actually that that time is pushing back at you and that you're being intentional about, what's my time budget that I need to shape into?

**Lenny Rachitsky** (00:35:24):
Let me take a quick tangent because if you're, that's so interesting that the time boxes can be very different lengths. Imagine at a larger company, this gets complicated when other teams are trying, there's dependencies and timelines launch and go to market dates and all these things. What's the largest company this approach has worked at? What's the ideal company stage for Shape Up?

**Ryan Singer** (00:35:47):
It can function in very large companies. We have, for example, I have some friends at a, what is it called? They're doing clinical trials. So, they're in the pharmaceutical industry and the companies, thousands of people, and it's not that every team is doing this, but they have a few teams that are working in important areas and they're doing this and it's completely possible in that context, if you have someone who's at a senior level on the engineering side who is able to make the right architectural choices and also do some negotiating and be the backstop to make sure that someone isn't going to get pulled away onto something else, if you can carve out, oh, this system can be worked on independently of that system. This was actually what David at Basecamp has always been amazing at is this dependency, how...

**Ryan Singer** (00:36:47):
It's actually not. It's not. So many people are used to it and they think that it's just how it is, but it's actually not. It is possible for engineering leadership, good engineering leadership untangles things, so we can work on this system without having to be thinking about this other system somewhere else. So, when you have some untangling with your infrastructure and with your architecture from an engineering standpoint, then you have some freedom. And then if you can also figure out the capacity management side of I'm going to protect this team from that other work for this number of weeks, you can really get a lot done.

**Lenny Rachitsky** (00:37:23):
This insight that you can operate this way at a larger company and the whole company doesn't have to operate this way, I think is really freeing to a lot of people. What's the adapter? And I want to come back to the actual shaping process, but I can't help but ask this. Say the company's operating a quarterly cadence or six month cadence and then there's a team operating in a two week, sometimes six weeks, sometimes four week cadence advice on how to, what's the adapter that connects those two cadences?

**Ryan Singer** (00:37:50):
So, there's two different things. So, I've seen cases where they've decided on a four-week plus two-week or so they'll do five-week and then one week of cool down in between and then they time it so that it adds up to a quarter. I've seen that. The other thing I've seen is when the team is just continuously delivering meaningful things, it doesn't have to line up because from the executive level, if you are CP or CTO or in these bigger cases, it's more like a VP in some area, but you're coming to the table where you're supposed to be reporting of what your group is doing. And when you are consistently saying, "We said we were going to do this and nothing finished and now, we're doing this and it's going to finish," and the next time you say, "We said we were going to do that and it's finished, without excuses and without, well, maybe another few more months or we're working at it," that's what everyone wants to see is that movement.

**Lenny Rachitsky** (00:38:52):
Yeah, if you're doing great, people will leave you alone. That makes sense.

**Ryan Singer** (00:38:54):
For sure. For sure.

**Lenny Rachitsky** (00:38:56):
I love that. I love that point. Okay, coming back to shaping, maybe one way that would make it real easy for people to understand, what's the output of the shaping session?

**Ryan Singer** (00:39:05):
The output of the shaping session is, and by the way, about shaping session, maybe we can talk a little bit about what shaping is not because we need the contrast sometimes. So, very often, when people try Shape Up, what I see is a product team creating either a lot of Figma files or maybe a lot of documentation, like a PRD with a bunch of requirements and a bunch of backstory and good reasons why we're doing this and stuff like that. And what you see is that when you give that to a team as this is what we shaped, what happens is it blows out. So, you probably know about what happens when the Figma file makes first contact with the engineering team. There's a reality check that happens there and very often, there's a back to the drawing board. So, when there's a lot of solutioning all the way down to detail without engineering involved, usually, that's a painful recipe and then it's like, "No, we can't do that," or, "Actually, it doesn't work like that."

**Ryan Singer** (00:40:16):
And then on top of it, the other big challenge is that there's so much that you can't see on the surface of a UI. How do we flow from here to there? What are the different cases of logic? In which case do we move from here to here to here in the flow? What is happening behind the scenes? It's like the engineering team, they have to put on their x-ray goggles and study this thing to try and understand what's happening underneath. I often use this analogy of if you're doing a home renovation, you can have the most beautiful rendering of here's the new bedroom, and we're going to have these lamps on the side of the bed that are coming out from the wall, and you can have the perfect rendering and the perfect lamp and the perfect color, but if you haven't checked if there's electricity in that wall there or not, it's going to drastically change the cost and the time and everything if you're going to have to rip open those walls to feed some lines up to those lamps.

**Ryan Singer** (00:41:15):
So, what we need to do in a shaping session when it's going well is we come out with some kind of drawing or diagram where engineers, product, and design are all looking at that and they're saying, "We understand that. I know exactly what to go build." I'll use the example of the calendar from the book. So, what is a calendar? So, first of all, there was this work that we had to do before we could even shape it, which is like, can we actually narrow down this problem? In shaping in real life, we call this framing. And in the book, there's a chapter called Setting the Boundaries where we get into this and it's like, look, we are not going to just build calendar, which is Google Calendar. Who knows where it ends? We narrowed it down to we understand that for our specific customers who are requesting this again and again, it's more about I need to see empty spaces and in the existing agenda view, I can only see things that are already scheduled and I can't see free spaces where I could book something.

**Ryan Singer** (00:42:21):
So, we got to that point of what we're trying to solve here is the empty spaces. So, that's a good frame. Then what are we actually going to build? We came to, here's a good rule of thumb. If it's shaped well, you can usually describe it in less than 10 moving pieces. If I can say, "It's going to have this, this, this, this, this, and this," and that's how we're going to let people see the empty spaces, that's a good indicator that it's clear enough that it's shaped well. So, in this case, when you go to an airline and you want to book something, you see this two-month grid. So, it's like there's going to be two months side by side, but they're just going to have dots in them to indicate if there's a free day or not, if there's something in that day or not, like the iPhone calendar, I think still has this where it's just dots on the month view.

**Ryan Singer** (00:43:17):
And then if you tap a day with a dot in it or without a dot in it, there's going to be an agenda view that slides underneath, which is going to show you what's scheduled in that day. And then there's going to be navigation to go forward and back in the months, there's going to be a create button to create an event, and that's more or less it. So, what you can see here is it's not like, what is a calendar? It's not a calendar. It's a two-month dot grid with scrolling agenda view underneath and the ability to hit new when you're looking at an empty space to create something in what you're viewing. So, that's the kind of thing where that's shaped and we can talk about what that means and what it entails, and we can have a really practical, realistic conversation about, is that a thing we can do in six weeks?

**Ryan Singer** (00:44:12):
That's going to be a real conversation and not looking at a whole bunch of mock-ups and trying to x-ray to figure out what's actually the intent here and what's really real and what's not and what's possible and what's not.

**Lenny Rachitsky** (00:44:23):
That was a great example. This is really helpful. So, if I were to try to describe this, essentially what you're coming out of it with a shaping session with is like the user experience with just wire frames/sketches of the screens and the key buttons and flows. So, it's like the architecture with key components, not like a dock of spec and not final design, and also not just a user story. As a user, I need to be able to see empty spaces.

**Ryan Singer** (00:44:56):
Exactly. So, because the thing where it goes wrong. If we're going to commit engineering time and it's like we believe there's some way to see empty spaces, but the way is a question mark, it's a really risky way to spend that time.

**Lenny Rachitsky** (00:45:11):
Because you're committing, right? It's like-

**Ryan Singer** (00:45:12):
Yeah, we're committing and that time is really valuable. That's six weeks of engineer's time, and that time wasn't easy to get in the first place because, of course, there's all these other forces in the company that want to be doing something with the engineers. So, if we want that team to be really using that time well where they are moving, they understand what they're solving and they're creatively engaged because they know what it's supposed to be doing, they need to have that clarity both on the problem side of this is about the empty spaces and on the solution side of it's a dot grid with two months and a scrolling agenda view and a button. There's still a million interesting creative tasks there in the actual high fidelity design in the code. There's so many things to solve there, but that is something that they can all hold in their heads and understand and work on.

**Lenny Rachitsky** (00:46:06):
**Ryan Singer** (00:47:26):
That's really interesting. I got to tell you, the dominant failure case that I see in the real world is always, again and again, not enough detail. And it's also the most common failure mode where the engineers run back to the product folks and say, "I'm not getting enough from you." It is really like that, but I can understand why the hair stands up on the back of the neck a little bit thinking about it because, of course, if you give a senior engineer like, "Here's how I want you to go implement the schema for this database change for this model," they're going to be like, "What do you think? Who are you? Who are you?" You know what I mean? But what's really interesting is it's not a universal thing. The amount of detail that the team is going to feel helps them is a dial that we can turn that depends on who's on the team.

**Ryan Singer** (00:48:28):
So, if you have a more junior person who's on the build team and then you have a more senior engineer who's involved in the shaping, they can make that junior engineer much more successful with additional detail. So, we're going to do this and I would suggest approaching it like this, this, this, and this. That junior person is, when they don't know how to do it, they're not going to ask because they don't want to show that they don't know, and they're going to hide the fact that they're lost and it's going to blow up later in the project. And on the other hand, if they are given more guidance, they're going to be able to be successful. They're going to learn about this is how this person who knows well, kind of approached it and then in the next round or a few projects later, you can dial it back and say, "Well, let's give less detail and see how they handle it."

**Ryan Singer** (00:49:19):
So, you can really give people bigger shoes to grow into and help them to be successful. And then, of course, you can also do it the other way around where if you've got some really stellar talent on the team and you have a long history and you have a lot of trust that they are going to be able to understand and solve it, then of course, you can leave things out.

**Ryan Singer** (00:49:40):
But the thing that I often see is if there's someone on the build team who really feels that they should be involved in the fundamental decisions about the approach, then a better solution would be to actually bring them into the shaping and have them play that technical role in the shaping session. If they have the right skills and the right perspective and the right knowledge to play that role well, then just bring them into the shaping. So, it's all about how do we bring people into a moment where we are using their strengths and then we're giving them an input, so that whatever their work step is that they're able to apply the maximum creativity, but also have the maximum clarity, so that they can really use that time well and also feel good about what they're doing.

**Lenny Rachitsky** (00:50:29):
It feels like the core of this is de-risk the biggest risks and address the biggest unknowns as much as possible. There's probably this 80% of here are the risks. Let's just understand them deeply before we commit to this appetite.

**Ryan Singer** (00:50:42):
That is exactly right. There are these, we can call them rabbit holes, we can call them time bombs. There are these things where we say, "Oh, it'll be fine." One example, simple example, I worked with a team and they had a step of onboarding in a FinTech product and there was this step of onboarding and they would lose a lot of people in the funnel at that step because you had to fill out a whole bunch of information, and they figured out that they could actually pipe that data in from one of the partners that they had. They were partnered with banks and they're like, "Oh, we don't even need to be asking people this. So, we're going to fix conversion. We're going to eliminate a step from our user experience. It's going to be great."

**Ryan Singer** (00:51:26):
The thing that they didn't look at was if you go into the code on that step of the onboarding, it's not actually one step. There's like three different branches of that step depending on which bank the customer is integrated on. And so, that's the kind of thing where it all sounds so great and simple, and then you get into the weeds and you realize like, "Oh, wait a minute." You know what I mean? So, now, we have decisions to make. Now, if you're in the middle of a project and it's already been resourced and people are already on the hook that we're supposed to be doing this, and you already got the alignment that the engineering time is happening for this, and you're finding that out in week four. You know what I mean? You did all these beautiful drawings, by the way, and now, you're finding this out in week four, that's a bad place to be.

**Ryan Singer** (00:52:14):
But if we're in the shaping room and we didn't kick this thing off yet, we didn't even green light the project yet, and we have an engineer there, not just the product people, not just the designer, but we have that engineer who really insists on sometimes, I like to think of it like the grumpy old plumber who's seen everything and he insists on opening up the walls and looking at the pipes before he'll give you a quote. So, it's like when you've got that person in the room, they're saying, "Yeah, that all sounds great. Let's take a quick look at the code and figure out what screen you're actually talking about. Just let's just take a quick look." And it only takes a moment to open up the code, find this thing that we're talking about, and really look at it and say, "Oh, it's more complicated than we thought."

**Ryan Singer** (00:52:59):
And now, it's not like, "Okay, now, we're screwed and the project is going to be bigger." Now, we can have a really cool conversation about trade-offs. So, let's say we've got three different integrations here, three different segments integrated into different banks. How big are they relative to each other in terms of the deals we made or the percentage of customers who flow through those three different conditions? If we just did this on one of those branches, would it be a win? And if we did it on all three, how much more time would we have to negotiate for and would it feel worth it? You see, we're getting into this horse-trading of what is important, what's worth it, what do we need to get out of it? And that's really productive. And when you're doing that before the project starts off, that feels like, "Oh, we're talking about the important things. We're not failing right now. We are engaged in the hard questions that are going to enable us to really ship something that's successful later."

**Lenny Rachitsky** (00:53:54):
Well, let's go one level deeper on this shaping session, because clearly, that is so core to this working, and I know you have a whole book about how to actually do this. So, we're not going to-

**Lenny Rachitsky** (00:54:00):
... to this working, and I know you have a whole book about how to actually do this, so we're not going to answer all the questions, and there's a lot of detail and nuance. But a few questions, how long do these usually take? It sounds like a whole day experience, and then it sounds like you invite as few people as possible, but not too few people. What's your guidance on who should join?

**Ryan Singer** (00:54:21):
In this shaping and real life course, we've been doing workshops where we try to help people to learn what it's like in a shaping session. One of the things that's always interesting to me is how... So Kathy and I will be running the session and we have to... People aren't used to working so fast. What are we actually doing right now? What's the decision? What's an idea right now? We're not going to go away and draw something, and then I'll comment on a document and then come back and get together tomorrow. What ideas do we actually have right now starting from zero? So imagine, we've narrowed down the calendar problem too. It's about empty spaces. We are willing, from a business standpoint, to spend six weeks on a whack at this where it's solved. We believe there's a way that's possible, so what can we come up with?

**Lenny Rachitsky** (00:55:23):
And that's the input to the framing session or sorry, the shaping session.

**Ryan Singer** (00:55:27):
Exactly. Having a narrowed down problem from the framing work, and this is a whole other topic of very often the PMs are sometimes just taking something at face value and not negotiating down to really narrow down what is this really, and where is the value in this? But let's suppose that that's happened, that we've narrowed down the problem, so now we've got a narrow problem. So now what we need to do is try out different ideas, and this is the real thing. We have to try to break them. So I want to draw an idea, and then I want the technical person to find, oh, this isn't going to... You know what I mean? This isn't going to work because of this reason. Or the product person is going to be looking in and saying, "I get that that's really an easy way to do it technically, but I don't think that we're actually delivering the value if I play through the customer scenario here." You know what I mean?

**Ryan Singer** (00:56:24):
So there's these different angles where the idea can fail, and one of the things that we also coach people to do in a session is not just to go down one path and then be stuck in one idea and now you're going in circles around little details of one idea for three hours, but to really step back and say, "Here's an approach. What if we had the scrolling agenda view, okay? And that's idea A. Then what's a very different way of doing this?" Do you know what I mean? If we didn't want to have the agenda view there, is there a way to do this where it's just a month view? Let's see if we can draw that. So that's the thing that's happening. You asked about the time, and I started with people aren't used to just going all the way in to actually trying ideas.

**Ryan Singer** (00:57:16):
So there is a little bit of learning how to just face that blank page and start trying things together. What we find is that a three-hour session can be very, very productive to help you figure out what do we already have that are possible approaches to this? What are some major missing things? Like, okay, I've got the calendar dot grid, I've got the agenda idea, but what about multi-day events? Do you know what I mean? So there can be these things, these what abouts. So then maybe we break and we think for a little bit, and somebody sketches some ideas on that and does a spike or two on something, and we come back again for another three hours or we come back the next day. And what I would say is if the project that you're trying to do is doable with, let's say, your existing technology, you're not inventing a new algorithm, you're not inventing some new database or... You know what I mean?

**Ryan Singer** (00:58:25):
You're not doing a new AI model. It's more about how do I use the APIs and the frameworks and the tech stack that we have? How do I put that together to build something? Then if the problem is clear and the time is now, you will be able to come to a conclusion about what's possible to build in three sessions, something like that. The key thing is leaning into those sessions and really wrestling with each other. If you have just the product and the designer there and then it's like, well, we'll show this to the technical person later, then it can all blow up. And then you find out it's more complicated than you thought and you have to go back to the drawing board. We need to have all the necessary information in the same room for these sessions to go fast.

**Lenny Rachitsky** (00:59:17):
There's so much genius built into this approach, and it sits on top of human nature. One is just, you need to actually spend... go into the deep edge cases and nuances and not just-

**Ryan Singer** (00:59:31):
Yes.

**Lenny Rachitsky** (00:59:32):
Yeah, that's fine. Let's go with [inaudible 00:59:34].

**Ryan Singer** (00:59:33):
More concreteness.

**Lenny Rachitsky** (00:59:34):
Very concrete, very in depth, and then the appetites are... There's just so many elements of this that just connect with the way humans work versus the theory of just like, "Yeah. Well, let's see how long this will take. It'll be a great... and we'll figure it out as we're building. We don't need to really figure this out. Yeah. We don't have time for that."

**Ryan Singer** (00:59:52):
And we're solving a puzzle together, if it needs to be doable in this amount of time. But it also has to hit these points in terms of the problem we're solving. You know what I mean? It has to do these things, but in this time. One other thing that I would mention is that we can't be drawing Figma files. By the way, I'm being very mean to Figma so far in this conversation. There is a time when it's the right time for the Figma file. What we want to do is have that clarity around the... Let's say, we already know where the sink is going in the kitchen and now we can make final calls about the tile and the exact fixture-

**Lenny Rachitsky** (01:00:38):
Grout color.

**Ryan Singer** (01:00:38):
... and stuff like that. Right, grout color. We don't want to have to throw it all away every time something changes. So there's a time and a place where Figma is amazing and feels good and it's like, oh, now it's beautiful. Now, it's amazing. But in a shaping session, you can't collaborate on something so high fidelity. So we need also some ways to collaborate, and this is where you see these techniques in the book, like breadboarding and fat marker sketching. These are tools to help us express an idea very, very clearly in detail. We're going to hit this button and from this button, go to here. This calculation runs, then we get this answer, and then we have this choice to go here or here. That's the thing that we need to be seeing and that's the level of detail where we can move fast together but still see something real as more this breadboarding level.

**Lenny Rachitsky** (01:01:33):
Fat marker session is very evocative of what this whole session is about, is very high level sketches. That's a great term.

**Ryan Singer** (01:01:41):
The danger there that I often see is that what we don't want is to say, "Oh, Figma isn't the right thing at this level. So instead, we're going to do fat marker sketches," and what you get is the equivalent of a blurry Figma. Do you know what I mean? Just less detailed. What we need from a fat marker sketch for it to be valuable to us as builders is it has to really communicate the idea. When I look at that, I'm like, "Oh, now I get it?" And if it's more this general wire frame of dashboard goes here and there's going to be four reports, and it's like I still don't know what to build, right?

**Lenny Rachitsky** (01:01:41):
Mm-hmm.

**Ryan Singer** (01:02:23):
So if it's not telling me what to build, so maybe this is a way to come back to your question about what's the output of the shaping session, it's like it's shaped if we can give this to a technical person and they say, "Yeah, I know what to go build now."

**Lenny Rachitsky** (01:02:37):
I'm very happy with our overview of the process. I think we've done a really good job giving people the gist. And obviously, if they want to actually implement it, they can get the book and dive in or work with you, which we'll point people to. Say someone's like, "This is awesome. I want to do this." What would you say is a good first step for a team that's currently... let's say, they're in startup land, and they're just shipping every two weeks, maybe every week? So maybe for that bucket of folks and then also for a larger company, I don't know, Agile Scrum SAFe, and they're just like, "Oh, let's try something different."

**Ryan Singer** (01:03:07):
Sometimes dev teams, they like the idea of not having the Scrum ritual, so they want to take in the six weeks, but unless the engineering and product come together to figure out how to collaboratively shape, like we talked about before, that time box isn't going to go well. So I-

**Lenny Rachitsky** (01:03:28):
They think they're going to not have to do standups, but now it's a day of hard thinking.

**Ryan Singer** (01:03:33):
Well, it turns into even more meetings, because we don't know what to do.

**Lenny Rachitsky** (01:03:39):
And the more meetings is just that shaping session specifically. Right?

**Ryan Singer** (01:03:44):
No, what I mean is that if we didn't-

**Lenny Rachitsky** (01:03:45):
Oh, right. I [inaudible 01:03:46] right.

**Ryan Singer** (01:03:46):
If we only adopted the six-week cycle and said we're going to figure it out and we didn't adopt the shaping, then we just don't know what to do. And then we reached the end, and we're basically scrambling to shape it as we go. And then we run out of time, and then we feel like this wasn't... It was nice to get a break from the Scrum rituals, but we can't say that we are knocking the champagne bottle on the boat because we're celebrating the launch or whatever, again and again. Right?

**Lenny Rachitsky** (01:04:13):
That's a good, actually, time to maybe talk about there's obviously the spring kickoff in Scrum. What's main difference for people, because they may be thinking, "Oh, this is just like a spring kickoff." That's-

**Ryan Singer** (01:04:22):
Oh, that's a good one.

**Lenny Rachitsky** (01:04:23):
... big deal.

**Ryan Singer** (01:04:24):
That's a good one. So what you often see in a Scrum team is that there's somebody who creates these tickets of these are the things that are going to happen inside of the sprint. Really, in my opinion, too many cases, it's not the person who's doing the building who's creating those tickets.

**Lenny Rachitsky** (01:04:45):
So your product owner.

**Ryan Singer** (01:04:46):
So there's a big, big gap there. We could talk a lot about that, but there's gaps in context. The person who's writing the ticket doesn't actually understand the work that's involved. You know what I mean? So there are so many unknowns and time bombs waiting in those tickets that sound reasonable, but they weren't really created by the person who understands the work that needs to happen. So the key difference is two things. So in Scrum, you have a person who's not the builders making tickets, and this is what in the cruel picture is the paper shredder. In the shape-up world, you have a single idea that was shaped. This is the thing we shaped with the two months in the agenda view and da da. Go make your own tasks, because you're the professionals. Do you know what I mean? So the contractors, if you're building a house, they have to know the plans, but you don't have to tell them, "Now take the hammer and go over here."

**Ryan Singer** (01:05:52):
That's their, right? So in a shape-up world, a kickoff is, here's the well-shaped idea, and now this time box starts. At Basecamp, it was really, really loose because they are just stocked with senior people. They have so many very senior engineers and all the designers are coding. They're very technical, really, really skilled people. What I found is helpful when the team is a little bit more mixed, if they're not all super senior, is a simple exercise of at kickoff, take whatever was shaped and just draw a grid with nine boxes, and give me nine boxes of the nine major chunks that you think have to get implemented from an implementation standpoint. So translate this into nine major scopes of implementation that need to somehow happen over the time box. So really, really useful exercise to kick things off, and we have a lot of teams doing that now.

**Ryan Singer** (01:06:52):
If you take six weeks, that's 30 business days. You divide that by nine, it's four days per box. So you're going to get a lot of clarity from a quick exercise. And again, this is done by the builders. This is a really also good exercise for them to notice like, "Oh, wait a minute, we think there's too much scope here. Even though it seemed reasonable, when we put it into nine boxes, it's like, I don't think we can do this all." Or it's also a good moment where somebody who's more junior might describe their implementation approach, and then someone who's senior can review that and say, "Actually, we've done that before, and we'll run into this trouble if we don't use this other thing." So those really nice coaching moments can happen.

**Lenny Rachitsky** (01:07:39):
If you were to try this approach and have a shaping session, this is a sign you're heading in a good direction, is if the output, the team that's building it can come up with nine... Does it have to be nine? Is six cool, 10 cool?

**Ryan Singer** (01:07:55):
What I found is if it's more than 10, then you just get into ticket land of, here's a million things I have to do. You know what I mean? If you have 100 things, that doesn't tell me anything. But if it has to be nine or less.

**Lenny Rachitsky** (01:08:10):
Nine or less. Okay. Okay, cool.

**Ryan Singer** (01:08:13):
I actually think... I'm speculating here, but the UX designers in your audience will know about this rule of seven, plus or minus two. It's this cognitive science principle that was found about how many things someone can hold in their head at once. So this nine is the upper limit of seven plus or minus two, and it basically... It's like, do we actually have a picture of what it means to build this that we can also hold in our heads? Can we see the whole castle?

**Lenny Rachitsky** (01:08:41):
So what I'm hearing is if you're on a, say, Agile Scrum team today, if you want to start trying this out, it's schedule a shaping session, assume it's six weeks to start, try to come into it with a framing of here's the problem we're trying to solve. Is that a good way of thinking about it, like the problem we're trying to solve?

**Ryan Singer** (01:08:59):
Yeah, for sure. The question is what problem are we trying to solve, because the shaping work is more what are our options for the solution? And if the problem is too fuzzy and big, if the problem is just calendar, then the shaping is going to be this ever-expanding, never-ending thing, and we're not going to be able to get anywhere.

**Lenny Rachitsky** (01:09:16):
Yeah. Okay. So you spend three hours, maybe six hours in the first session. Would you recommend try to keep it to this many hours when you're first trying it up?

**Ryan Singer** (01:09:26):
No, I wouldn't do that. I think the key thing is actually if you get to the point where you're trying to hold a shaping session and you manage to get product and engineering into the same room to do it, you are far along. You're doing great if you're at that point. Oh, so much of the challenge is getting to the point of aligning between product and engineering that we cannot have projects that are dragging and dragging and dragging. We can't keep ending in a place where this is the end of a sprint or the end of a cycle and we still can't see the end of it, or we have to make so many cuts and compromises at the last minute that it's not the quality of... or it's not really matching what it was supposed to be in the first place. When those problems are happening or... Also by the way, this is surfacing at the exact level.

**Ryan Singer** (01:10:22):
Imagine, you're the CPO, you're the CTO, and you are supposed to be answering to, "How's that work going?"

**Ryan Singer** (01:10:30):
And it's like, "Well, actually, we're working on it. I can just think of a couple of times when I needed to go to Jason, and he expected me to be making progress on something and I had gotten nowhere on it. And that feeling when you are with top leadership in the room and you don't have a good answer for where you are on something is like... Oh, it's brutal, right? And then from the CEO's perspective, it's like, "Where's the movement? We're running a business here. Really, nothing is shipping still?"

**Ryan Singer** (01:11:03):
This can't just keep happening. So there's some recognition somewhere either at the higher levels or within the team of, we don't want to keep dragging, we don't want to keep being lost in the weeds, and then this can be the activation energy. You gather the power to be like, "Okay, we actually want to try something different."

**Ryan Singer** (01:11:26):
And in that case, what I would say is what usually works best is, okay, we're going to try a pilot project, and what we want to do is, as you said, choose a problem that's important enough to all of us that we think it's meaningful, it's going to be worth trying to do well. And it doesn't have to be six weeks. It could be something that is a little bit smaller, maybe you feel comfortable taking on three week thing for the first time. What's important is just matching these things together.

**Ryan Singer** (01:11:55):
Here's a problem that we actually care about. It's timely, something that we would like to be shipped soon. It's not so small that we're not going to actually learn this new muscle, and it's big enough that it's going to feel like we really achieved something. So maybe that's going to be four weeks, maybe it's going to be six, maybe it's three, I don't know. And then getting to a place where we wrestle a bit with the problem to get the problem narrowed down. We get into our shaping session, and then we do our best. Do you know what I mean? And usually, what happens too is if we have an engineering team that's going to become free to do this work for those X number of weeks, that's the upper limit on how long we can spend to shape, and that's another real life thing, is sometimes we talk about if...

**Ryan Singer** (01:12:51):
On the one hand there's this universe of never ending documents back and forth to get feedback and comments, and then on the other hand there's like the team is going to be available. We're trying to actually do this, so actually, we've got a week to shape because engineering needs to kick off next week. Do you know what I mean? That's a little bit more the real scenario when you're actually in this aligned world of we want to ship something now.

**Lenny Rachitsky** (01:13:16):
Yeah, real life constraints. That's a really helpful way of telling you how much time you have to do this. For people that are just like, "I don't know, any friends that are using this. It's like weird, this way of working. It's not a thing I hear about all the time." What can you say to them to help them be like, "Okay, I should actually give this a try. Here's how many people are using it. Here's impact that you've seen." Anything you can share that would help them get over that hump?

**Ryan Singer** (01:13:40):
I would say wait until it hurts more. If the unfamiliarity is the big problem with it, then maybe the things are fine. Because it's not like this is the only way. It's more like, changing is really hard, and if there's a good reason to do it and it's like, look, we've done it the old way. We've tried different experiments. We've even already churned through a new head of product, or we've got a different CTO in and we're still having the same problems, then there comes a point where it's like, I know that this is uncomfortable, and I don't know somebody who's done it, but I think we need to try something different because we can't continue this way.

**Lenny Rachitsky** (01:14:30):
That is a great answer. Following that same thread, just what are signs that it's time to try something? What sort of pain do you often see that's like, okay, you shouldn't look into this seriously?

**Ryan Singer** (01:14:44):
There are pains all along the journey. So I think the place where it's most obvious is at the end of the line when we thought we were going to be done and this thing is just dragging and dragging and dragging. The teams, we're not shipping things. We're running in place. We keep going in circles on this like we don't see the end. Of course, that's the culmination, but there's also a lot of pain points along the way.

**Ryan Singer** (01:15:11):
So if we go all the way upstream, if we go to the source of a project, sales talk to a customer... You know what I mean? Or sales talk to a lead, and they have this idea of this thing we need to build, or the CEO had this idea in the shower the other day, or the product team did a whole bunch of research and they have a big case for why this is the thing that's important to build next. Whatever it is, there's a source from the business perspective of this is the thing we should do next.

**Ryan Singer** (01:15:44):
If we just say dashboard and we don't negotiate what that means, if we just say calendar and we don't negotiate what that actually is, then what do we experience? This fuzzy thing where it's really hard to get to a conclusive answer about, yeah, that's what we need to go do. It's like the ever expanding blob. So if you've felt that before, that's already a first pain. And then of course, where does it go from there? So we say calendar, so we don't know what it means, but we say calendar, so now we give it to product and we've either got a whole bunch of Figma files or we have the PRD with a million requirements about what a great calendar is. Of course, I don't want to be cruel to the people who are putting their hearts into that work, because the Figma file is beautiful. It's just coming a little early. And the PRD is full of a lot of true things that are probably really important for decision-making in the project, but the way that it's packaged at that moment isn't something that gets absorbed. You write this document and I'm sorry, who actually reads it? You know what I mean? I know it's painful, but it's like that. And then even when we try to read it, because it wasn't shaped and we didn't get down to it's this, it's that, it's that, and that's how it works, it's hard to walk away from reading that and have anything that's in your head as, this is what we're going to go build. It's like a million puzzle pieces that you're going to have to solve. So what we see is either there's the Figma file and then there's the pushback from engineers. There's the PRD, but then it's like, okay, but we still don't actually know what to build.

**Ryan Singer** (01:17:40):
There's all those things where, instead of moving forward, there's more and more questions, more and more pushback, more and more going back to the drawing board. So that's another big indicator that something is going wrong. And then when we're in the building and we thought we knew what we agreed to, we thought we all said, "Yeah, this is what we want to go make," and it's just more and more questions coming out. More and more unexpected complexity, things that we didn't anticipate, and it just doesn't feel like we're getting warmer and coming closer. It just feels like it's getting harder and harder. Those are all the signs that whatever process you use, that there's a lack of clear shaping and there's a lack of clear framing because there's a lack of clarity around what it is that we're doing.

**Lenny Rachitsky** (01:18:26):
Before we started recording, you made this interesting point that there's always talk of feature factories and that rarely are they actually efficient factories. They don't actually work. Talk about that.

**Ryan Singer** (01:18:38):
Yeah. Well, I understand what the feature factory critique is supposed to be. It's actually to the framing point of we're not negotiating the value and the outcome we're trying to get from something. We're just taking it and building it. And then of course, in the end, according to the feature factory critique, we just built it because somebody said we should build it, and then people didn't use it and didn't value it, and the product is just getting bloated. The thing is that, I would say if you have a feature factory, meaning you're continually cranking out features, you're probably quite healthy. All you need to do is feed a different input to the beginning of the factory.

**Ryan Singer** (01:19:18):
What most teams are struggling with is that they don't have predictable repeatable shipping of things. At least from my experience, the bigger really widespread struggle is, stuff isn't moving, it's dragging. I can't see the end. I'm losing my... I'm feeling burned out. You know what I mean? It's not exciting to work on this anymore, all that a thing.

**Lenny Rachitsky** (01:19:41):
Maybe last question here is what's the sweet spot stage for a company to start using Shape Up? You basically worked in this way from the very beginning when it was just three people. What do you find... Should startups that are just starting out start working in this way, or do you find it's more useful later on?

**Ryan Singer** (01:19:58):
We didn't formalize it until we had to, and there was a long time where there wasn't a fixed length for projects. There was just an understanding of the urgency and a feeling of what too long felt like. And it didn't actually click into, oh, this is a cycle length and this is six weeks, and then we pause, and this is who comes together to make the decision of what's the next project, and here's who is mainly doing the shape. You know what I mean? All that stuff didn't get solved until we had reached a certain size. Usually, the main tipping point if we start from smaller to big is there's a phase when the founders are still involved in everything, and so it doesn't matter what your process is, it's going to be fine.

**Ryan Singer** (01:20:40):
But then you start to hire the first other people and then for the first time you try to delegate some of those things and the founders try to be less involved, and that's often where a lot of these problems start to appear. And the founders start to ask themselves like, "We used to be fast, and now we hired people because we needed to scale, but now we're slow. So how do we be fast again? Because we know what-

**Ryan Singer** (01:21:00):
... well. So like, "How will we be fast again?" Because we know what it's like. If we just got back in there as founders and got our hands dirty, like we could make this go. But how do I get the people that we've brought in to make these trade-offs and make these decisions and how do I get the work to flow again? So that's something that we definitely see there. So that's a really good moment. I'm onboarding new people. I don't know what to tell them how to work. I don't want to introduce a bunch of scrum rituals. Just winging it on Kanban isn't working, because they don't have enough clarity around what to go after. So I have to babysit them all the time. You know what I mean? Like these kinds of things.

**Ryan Singer** (01:21:40):
There is another extreme, which is I... We've already gone past that. We've been scrum or whatever for years. The company has been growing, like revenue is coming in, like sales is doing their job, like things are running. But man, nothing is getting out the door. And we're years in and we have an entrenched development. We have like an entrenched engineering team, which is a wall away from an entrenched product team and everybody's apart. And this thing is like, we're like stuck.

**Ryan Singer** (01:22:19):
And that is more where there's going to be some tensions that are starting to appear at the executive level. There's going to be some finger pointing. There's going to be some like, "Why isn't this moving? Why isn't this happening? How can we be spending so much money in all these engineers and we don't have anything to show for it?" And that's a point where there can be kind of a... Some hard conversations need to start happening about, "How do we actually start to negotiate around how we spend time?" And we can't just have endless refactorings and infrastructure projects, but we need to be actually building things that we can sell again.

**Lenny Rachitsky** (01:22:54):
What an idea.

**Ryan Singer** (01:22:56):
Yeah. You know? But it can... There are a lot of engineering orgs that have been standing around for a while and it's all refactoring all day and tech debt and stuff like that. And there's reasons why all those things got there, but there comes a point where we have to figure out how to cut through it and make some hard choices so that we can carve out time to build the stuff that's actually going to be needle moving again and not just sustaining us where we are to run in place.

**Lenny Rachitsky** (01:23:24):
I imagine this latter bucket is who you mostly work with, the kind of companies that bring you in.

**Ryan Singer** (01:23:33):
It's been a lot of the folks who still remember what it was like to be fast and they're kind of newly too big and they don't like being slow. I've had a lot of that. I think that your intuition is right. That the market for the last category is the biggest, but it's hard to reach them. It's not easy to talk about these things. These are sensitive topics. Do you know what I mean? Like, "Our engineering team isn't shipping," and it's happening at leadership level. There's a ton of complaints happening deeper in the org, but nobody down in the org can change anything. At the end of the day, it's actually the interface at the executive level of being able to say, "How are we using our time? We have to change something."

**Lenny Rachitsky** (01:24:19):
To make it even more concrete in that first bucket, what's the size of org that you find is most in need of this? It's like, "How many engineers?" Or is it like when they hire the first PM? Like what's kind of the-

**Ryan Singer** (01:24:29):
I sometimes have the like, "How the heck do I hire the first PM and what do they do?" conversation. But usually, it's later than that. It's after they hired the first PM, after they hired the second PM, and maybe even the third. And they're getting to the... Product and engineering together are like 30, 50 people and it's like, "We thought we put everybody in the right roles. We kind of did what we were supposed to do and everything is just grinding. And why are we so slow?

**Lenny Rachitsky** (01:24:57):
Perfect. So 30 to 50-ish people seems like a good time to... Basically, you're finding that's when things start to really break down.

**Ryan Singer** (01:25:05):
That's when they show themselves and I think... I mean, if someone hears this and it all starts to make sense and they're earlier in that wave, then of course the earlier that you can anticipate it, the better, right?

**Lenny Rachitsky** (01:25:16):
Yeah. That's a good point.

**Ryan Singer** (01:25:16):
So if you're-

**Lenny Rachitsky** (01:25:17):
When it's too late is when they come out so-

**Ryan Singer** (01:25:19):
I mainly hear about it when it's too late. That's why they're reaching out-

**Lenny Rachitsky** (01:25:22):
Got it. So maybe closer to 30. Okay.

**Ryan Singer** (01:25:26):
Honestly, I think it starts the first project where, for example, the founding engineer is hands off and then the new hire is taking over responsibility. Or the person who was like sort of founder/CEO is first giving it to a PM to kind of thinking they're going to carry it through. And then, it's not exactly meeting their expectations of what they thought was going to happen. I think that's when those disconnects actually start. It's the first step away from the work where the seeds of all of these problems actually start.

**Lenny Rachitsky** (01:26:00):
I want to talk about Basecamp and how maybe not every company can operate like Basecamp. Before we get there, is there anything else along the lines of Shape Up that you want to add or share?

**Ryan Singer** (01:26:10):
Yeah. There's one key thing, which is the role of the PM. I think what we see today, out of necessity in a lot of teams, is that the PMs spend a lot of time chasing around inside of the build phase, inside the time box, to try and make sure that people aren't stuck and getting lost in the weeds and try and keep things moving. And it can sometimes be too close to project management rather than product management.

**Ryan Singer** (01:26:43):
And what we see in Shape Up teams when they hit their stride is that the PM moves upstream. So the PM is less busy with, "How do I get this project to not be in a bad state when it's getting built?" And they're way more in, "How do I understand the business context? How do I narrow down the problem? How do I negotiate back and forth with maybe the CPO who brought this to me to understand where the core of it is?" That really getting the deeper understanding of the business and the problem and the customer domain and like what problem is worth solving and what's even slice of that problem is the valuable slice to argue that we should spend a few weeks on. That's the place where the PMs can really contribute a lot in the Shape Up world. That's kind of what they do, rather than shepherding the process or being a ritual master or something like that.

**Lenny Rachitsky** (01:27:42):
That sounds pretty wonderful. I've been doing some thinking about what an AI-oriented world does to the role of PM and it feels like very similar to that actually, where the building now is going to happen for you with AI tooling. And that means the bigger question now is like, "What the hell should I build? And is the thing I've built right and correct and likely to work?" And it feels like this is similar, it's like the PM spending a lot more time upfront thinking through what to build. And then, the building is a lot more hands off. So hands off it gets done in like five minutes when you're just like, "Well, build this thing for me." "There it is."

**Ryan Singer** (01:28:19):
Yeah. Let's see. Let's see. Yeah.

**Lenny Rachitsky** (01:28:21):
Let's see.

**Ryan Singer** (01:28:22):
I'm also very curious. Yeah.

**Lenny Rachitsky** (01:28:25):
Oh, man. What a wild time. Okay. Let's talk about Basecamp. I think we talked about this ahead of the podcast that... You want people to know that Basecamp is very unique and not everyone can work the way Basecamp works. Just talk about your insight there and your advice there when people see all this advice coming out of Basecamp.

**Ryan Singer** (01:28:45):
I got to tell you, I had no idea how unique we were until I was outside and there are so many things. For example, it's a lot of the things that people ask me about that are not in the book that started to reveal those things to me. That's so many things that I was just taking for granted. I mean, every designer codes.

**Ryan Singer** (01:29:05):
Imagine, if every designer codes and I don't just mean HTML. I mean, like running the app locally, going in to the place where that view is rendered to make that thing look the way that they want it to look or whatever, right? I mean, like really codes, every designer. So every designer codes, where's the wall between design and engineering? Where is the moment where you arrive with the Figma file and then the disappointment and all of your hopes get destroyed because the engineer is telling you no, right? Like those moments don't even exist in that world.

**Ryan Singer** (01:29:42):
And then, also, I think also there was this lack of distance between sort of the business objective, the thing that we're trying to... The reason we want to maybe do this project and the blessing of the founders and the... Like, there wasn't this kind of executives far away with some big targets and then some layer of PM and then some building. I mean, the founders were always there, right there in the problem definition still.

**Ryan Singer** (01:30:14):
I mean, I can't say today, but I mean up until 2021 when I was still there. So it meant that there was so much clarity all the time around what we're solving and why and why we're making time for it. And then, of course, on the engineering side as well. I mean, imagine, you have no sales org, you have no marketing. That all of selling and marketing is happening by the unicorn founders. So it means that there isn't contention for engineering time, that there isn't like all these different sources of requests that you have to wrestle with,

**Ryan Singer** (01:30:50):
And David did such an extraordinary job of... I mean, the more I see the real world, the more I'm amazed at how every six weeks, there was clear runway in engineering of like, "We have time for whatever the... Whatever we'd agreed together is the most important thing." Just blank check like six weeks at a time. Not a blank check, but you know what I mean? Like a blank six weeks, yeah?

**Lenny Rachitsky** (01:31:15):
Yeah.

**Ryan Singer** (01:31:16):
Again and again and again, years without end. Keeping that engineering capacity focused on readiness for product and totally leaning into what's exciting to do to build for the product. And not getting lost in all this refactoring and new infrastructure and technical debt and stuff like that. I mean, those are amazing. So those are some really big differences. And it doesn't mean that you have to be Basecamp to do Shape Up. But what it does mean is that when we say, "Oh, just have a shaping session and if you have the pressure of the time box, then you can make trade-offs together." It's like, "Well, if we are used to having a big wall between, for example, engineering and design, then we're going to have to learn..." Somebody who wants to start shaping is going to have to learn like, "Well, oh, I need to figure out who to bring together and how to have that session and how do we interact with each other. So that we are combining all of that knowledge that maybe at Basecamp was all in the same head in a lot of cases."

**Lenny Rachitsky** (01:32:18):
This is such an important point for people to hear, because there's so many people that come on podcasts like this and share, "Here's how to do it," based on their experience. And there's so many just assumptions about their resources, the people they hire, the way the founders operate. Like no sales team, I think that's like... I don't even think about that.

**Ryan Singer** (01:32:36):
Yeah. Imagine, no such thing as a request from sales, yeah? No such thing as pressure of like, "We need this thing in order to upsell or to close this deal." Never.

**Lenny Rachitsky** (01:32:47):
It sounds like you're in this Basecamp... By the way, was it called 37signals? Like it's interesting you call it Basecamp not 37signal.

**Ryan Singer** (01:32:54):
Yeah. I mean, so it's just like the timing of when I left. We were originally 37signals and then Basecamp became so big that we renamed ourselves to Basecamp.

**Lenny Rachitsky** (01:33:03):
I didn't know that.

**Ryan Singer** (01:33:04):
Yeah. And then, so for example, on the book, it says Shape Up and there's a Basecamp logo on the bottom, not a 37signals logo. But then, since I went back, so it's 37signals again. So I sometimes struggle with I don't know what to call it but it's both. Whatever people can recognize, it's the same powerhouse.

**Lenny Rachitsky** (01:33:24):
Okay. Cool. I'm glad I'm not the only one that's confused. But 37signal is the current name. Great.

**Ryan Singer** (01:33:29):
Yeah. Yeah.

**Lenny Rachitsky** (01:33:30):
So you said that it felt like you left and it was like this bubble you got out of. Was there like a moment where you're like, "You wrote this book. Everyone..." You're like, "Hey, this is how you should work," and then you're like, "Oh, wait. This doesn't actually work in real life for a lot of people."? Is there a moment there?

**Ryan Singer** (01:33:44):
It wasn't that this doesn't work, I was just in a foreign country. It was like we tried it and it didn't work. One of the common things I would hear is the projects kept running over. "We weren't finishing them at the end of the cycle. They kept running over and running over." And then, I would be like, "Huh. So can you show me your shaping work?" And then, they would show me a PRD and I'd be like, "That doesn't look like what's in the book." And again like, "Can you show me your shaping work?" And they'd show me like a bunch of Figma files.

**Ryan Singer** (01:34:21):
And then, what I started to understand was like we have some people in a role who were used to making a certain artifact at a certain step and they just kept doing that. And I didn't appreciate... It took me a while to realize like, "There's no engineer in the picture here." And it was when we started to actually do the course, I said... Well, I did actually a couple projects where I helped teams hands-on and I learned that they...

**Ryan Singer** (01:34:50):
It was the first actually consulting project that I did where I helped a team who was stuck. And what we did was we chose the engineer who was best suited to come over to product and be there in the shaping. And that was the moment when it was like, "Ah. Now, I'm in the world I know again," when we had all of that mixed in the same room again. And so, that was kind of like... That was really something... I mean, it was a total learning curve for me and there's a lot of things like that. But that was, for example, a really big one. It's like, "Oh, we have to get engineering in there."

**Lenny Rachitsky** (01:35:24):
You're the type of guest I most love having on this podcast, because you basically work with many, many companies, study what's working, what's not. You're not in the clouds pontificating about something. You're working with teams to make things better. And then, you take all of that learning, put it into a book, and share with us all. And so, the ROI is just incredible for us all because you've spent so much time doing this and you've actually done the work. You're not just in theory about it. So this is amazing. But we're not done yet. One question I wanted to ask is, Jason was tweeting that there's... He's working on a follow-up Shape Up book. What's happening there? Are you involved in that? What's the story?

**Ryan Singer** (01:36:06):
So I also saw the tweet. And I have to admit, I was a little surprised when I saw this tweet, but I had had a conversation with Jason a year earlier. And he reached out and he said, "Hey, we're thinking about doing a second edition of the book." And my first reaction... Imagine, I was actually really in the middle of learning all these things that teams need to learn in order to catch up to what was natural for Basecamp to do. And so, for me, it was like, "Interesting. I have a lot of new things. I have a lot of new ideas. Maybe collaborating on the second edition could make sense."

**Ryan Singer** (01:36:46):
But what I understood was that what he wanted to do was to make an updated version of how they work, because that's always been a big thing of how... I should use the right name, 37signals, of how they market and also how they lead is they like to really show a clear example. Not like, "This is how you could do it. This is how some people do it," but like, "This is how we do it."

**Ryan Singer** (01:37:09):
And I think it's their strength that they are very, very clear like, "This is how we do it and take it or leave it." What I understood was that if I did another version of the book that was just how Basecamp does it, I think it would leave so much opportunity on the table. Like there's so many people where what they need to learn is more like, "How can it come closer to where I am? If I have the wall today between product and engineering, how do I bring the right people together into a shaping session? How do we actually do that? How do I overcome all of these little challenges because this is so far from our current way of working?"

**Ryan Singer** (01:37:44):
So the work that I'm doing with, for example, with shaping in real life is all about those gaps. And then, I don't know what's going to be in the second edition because they are... I guess someone there is going to be working on that. But what I'm guessing is it's going to be an update on kind of, "Up on top of the mountain over here, this is what Basecamp is doing." So hopefully, it'll be a cool thing to look at as like, "Here's a model of what they're doing." And then the question is, "What can I take from that and what do I need in order to actually be able to make it work in the real situation I'm in?" And that's kind of where... Well, that's my focus.

**Lenny Rachitsky** (01:38:20):
This is so interesting. Thank you for sharing. It sounds like a fork. You forked it and these are going potentially in different directions but inspired by each other.

**Ryan Singer** (01:38:29):
Mm-hmm. Totally.

**Lenny Rachitsky** (01:38:30):
So interesting. Ryan, is there anything else that you want to share before we wrap up?

**Ryan Singer** (01:38:37):
One thing I could throw out there is sometimes people reach out to me because their projects aren't shipping, there's a lot of struggle, there's a lot of lack of clarity. But the root cause is actually that the input at the very beginning of the process is too unclear or... Like we don't actually know what's important to customers or we're not actually sure where the value is or this kind of a thing. So there is this link, this framing step that we talked about of, "What is really the problem?", before we go into shaping.

**Ryan Singer** (01:39:11):
This is the link to product strategy also. And this is the place where it can be really useful to reach for a lot of, for example, Bob Moesta stuff with the Jobs-to-be-Done and the demand-side work, trying to get clear about big... So that's the tool that I reach for at that phase. And you can think of kind of this... Before the problem definition, there's this question of like, "What's the demand? Where are people struggling? Where is really the place, the itch they're trying to scratch?

**Ryan Singer** (01:39:42):
And then, a lot of the Shape Up stuff is kind of when I have something where I think there's an opportunity or I think there's something meaningful there because of what we learned from customers or the job to be done, research, or whatever it was. Now, how do I turn that into something that we can actually go do and ship in a reasonable amount of time? That's the supply side. That's where the Shape Up part fits in. So maybe it just might be cool for people to see a link there.

**Lenny Rachitsky** (01:40:06):
That's a great plug for a Bob Moesta episode. We talked in-depth about the Jobs-to-be-Done framework and how to actually apply it. What's the book you'd recommend there? It sounds like basically it's like Shape Up plus this book gives you a lot.

**Ryan Singer** (01:40:19):
Actually, the one that I recommend the most is Demand-Side Sales 101 and it's funny because it's like sales. Especially for a product person, you're like, "I'm the product person, not the sales person." But it's such a good dive into, "What are people really trying to solve?" And getting into that mindset of, "What's the struggle? What's the problem?" I think that's a really good entry point for that.

**Lenny Rachitsky** (01:40:43):
Yeah. I don't love that title. I feel like you could have done better there with that book's title because-

**Ryan Singer** (01:40:48):
It's-

**Lenny Rachitsky** (01:40:49):
Yeah.

**Ryan Singer** (01:40:50):
What's interesting about it is that it's very, very pointy for like if you are trying to make progress on sales, then it's this other kind of sales, this demand side sales. So I think maybe it's more for us who are kind of using it for different purposes. Like we're the product people trying to pull something out of it. That it's a little bit less aligned, but it's still useful.

**Lenny Rachitsky** (01:41:11):
Yeah. But basically it's like the Jobs-to-be-Done book is what-

**Ryan Singer** (01:41:15):
Yeah. It's kind of like the Jobs-to-be-Done book that's a bit more tactical. If you're really curious about the general spirit of Jobs-to-be-Done, then Competing Against Luck is a really good intro. That's the one that Clay Christensen wrote with a lot of... I think there's a lot of stuff that Bob worked on that's in there. But for a little bit more tactical like, "What's it look like to do the interview? And how to think about the struggling moment?" and stuff like that, this Demand-Side Sales is good for this strategy stuff.

**Lenny Rachitsky** (01:41:44):
Awesome. And we'll also link to this episode where you could get the gist of it in one hour's time.

**Ryan Singer** (01:41:48):
Oh, that's right. You did... That episode was great by the way. That's... Yeah-

**Lenny Rachitsky** (01:41:52):
Thanks, Ryan. Thanks, Ryan. Okay. We did it. This was amazing. I think this is going to help so many people-

**Ryan Singer** (01:41:58):
We got through it.

**Lenny Rachitsky** (01:41:58):
We're not done yet. Two final questions for you. Where can folks find the book, find you if they want to work with you? Anything else that you want to share? And how can listeners be useful to you?

**Ryan Singer** (01:42:08):
Well, they can find me at my website. That's ryansinger.co. I'm also on X on RJS. I'm on LinkedIn. So just reach me there. And how can people be useful to me? I love hearing from people who are having these problems. If you're having these problems where it's like, "Things are dragging," or, "We can't see the end and we're not getting the quality we need," and all this stuff like man. I mean, this is how I learned all this stuff is by talking to people who are in it. So even if it's not clear what's the next step yet. If that problem is real, it would be cool to hear about it. I'd love to chat.

**Lenny Rachitsky** (01:42:46):
Be careful what you wish for about Moesta. He was just on the podcast and he told me he's got over a hundred LinkedIn DMs with people sharing their struggles with their job search. So here you go.

**Ryan Singer** (01:42:57):
Oh, yeah. Job moves, that's a big one. I think that's a broad appeal. Yeah.

**Lenny Rachitsky** (01:43:01):
Yeah. That's true. I'm going to ask you to explain that when you do consulting work, just like how does that work? Who's that for? Just because I know that's something else you do.

**Ryan Singer** (01:43:10):
So it basically starts with uh, either often first CPO or CTO often reaches out first. And when it works well is when we actually get them together and then they understand that they need to change something or we have like a head of product and a head of engineering, that kind of a thing. If those two are both seeing eye to eye that there's a problem, then we can start a conversation around, "Okay. So who would be the right people for a pilot team? What are the things that are going on business-wise that could be a good pilot project?"

**Ryan Singer** (01:43:41):
And then, I can help to figure out like, "How do we actually..." So almost like guiding through, like narrowing down that pilot project framing so that they have the support that it's going to be successful in shaping. And then, coaching the team so that they actually learn those shaping skills so that they can get through a session and come out with much more clarity. Like how do they actually run those sessions.

**Ryan Singer** (01:44:03):
So it's kind of first working with leadership, "Who do we need to get to do this work? Who are the right people? How do we bring them into a pilot project?" Narrowing down, doing some framing work on the pilot, so it's going to be clearer in the shaping. And then, giving some guidance on how to get through that shaping with some feedback rounds. This is usually a good approach.

**Lenny Rachitsky** (01:44:22):
Amazing. And they can find this on your website if they want to explore this?

**Ryan Singer** (01:44:24):
Yes.

**Lenny Rachitsky** (01:44:26):
Amazing. Ryan, thank you so much for being here.

**Ryan Singer** (01:44:29):
Yeah. Thanks a lot. You had amazing questions. It's a subject that can go in so many directions and you kept bringing us onto some kind of a main track, so I'm really pleased. It was really nice. Thanks a lot.

**Lenny Rachitsky** (01:44:39):
I do my best.

**Ryan Singer** (01:44:39):
Yeah.

**Lenny Rachitsky** (01:44:40):
Thanks, Ryan, and bye, everyone.

**Lenny Rachitsky** (01:44:45):
Thank you so much for listening. If you found this valuable, you can subscribe to the show on Apple Podcasts, Spotify, or your favorite podcast app. Also, please consider giving us a rating or leaving a review as that really helps other listeners find the podcast. You can find all past episodes or learn more about the show at lennyspodcast.com. See you in the next episode.

---


</div>