The Exchanges

Q Do you think Zoom helped that? It removed the, um, the fear actually when you're sitting in a room and you feel that. I think Zoom helps. You're just a square now. I can say what I want. I'm in my home.

A Right. There's a bunch of reasons that virtual makes improvements, right? The one you mentioned, which is like everybody gets their own square. Like that alone is a big deal. Now everyone literally is on equal, uh, Pixel, pixel footing to everyone else. Um, And you, and some people, you know, interact or, or, or best bring their best selves in different contexts. So having like time to be on mute and like maybe have a doc open and like type your thoughts down and then present, like maybe that works better for some people. I think that's great. Um, but I think there are trade-offs, right? On Zoom, I don't know about the meetings you join, but like often you just sit there and there's all these squares. Nobody says a word until like the person running the meeting joins and then everyone unmutes. Um, I think you lose a lot of culture building. By not having those awkward down times, right? In fact, like when, when COVID first hit, and we had to, like, abruptly switch to all remote, um, one of the, and everybody was brainstorming, like, how do you keep your team culture? How do you keep people, like, knowing each other, all that? The technique that I found, like, the most successful, it's the simplest thing I could think of, is for every meeting that I was kind of, like, the top person in, I would just intentionally show up, like, five to seven minutes late. And I would, it creates…

AI assessment note: “There's a bunch of reasons that virtual makes improvements, right? The one you mentioned”

Q Sorry, final, final one on product reviews. Um, you don't want it just to be like a great discussion, and then what? What's the right post-meeting process to ensure that the discussion items are implemented in some way?

A Yeah. Um, I mean, I think the simple thing is like whatever you decide to do in that room, somebody needs to be the singular owner of each of those things happening. Um, and again, this, this, this is probably broader than product reviews. It's more like how do you run a larger scale team? Um, and what I have found again is you lose a lot of value the minute responsibility gets sharded across multiple people. Um, if there are like three people who say they're gonna clean up the floor in the office, who's gonna actually do it? Like, it's less likely that the floor gets clean when there's three people responsible than when there's one person responsible. So I always loved operating in this model. You know, Apple calls it the directly responsible individual, which is just a person, and it's not necessarily the boss of all the people, just a person who's designated as, like, your job is to make sure this thing succeeds. If there are any roadblocks, you escalate it to whoever you need to escalate to so they can help fix the problem. I think that's the way you should come out of all these reviews. Like, who's responsible for each piece of this? And like, when are you going to get it done by? And getting them to sign up for that.

AI assessment note: “somebody needs to be the singular owner of each of those things happening”

Q Did you, did you write the same check size every single company?

A Um, well, at the time, I, it was, there's one company I could invest in, and I have no money because, like, we hadn't yet really gotten any liquidity from Pomp, um, so that was, like, the only company I could invest in, right? Uh, later on, yeah, I, I've generally tried to keep my check size reasonably static, um, because, I don't know, there's lots of different opinions on this, but, um, it's very hard to know up front which companies are going to be successful, and in fact, Often your gut will tell you the opposite. This is interesting. We'll tell you the opposite of what actually will happen. The companies that seem the stupidest or the craziest, and you're just like, I don't know, but I like the founder. I'll invest. Those are the ones that turn out to be big wins. So if you modulate your check size based on like probability of success or like how big the market is or whatever, you'll probably be doing the exact opposite of what you should do.

AI assessment note: “I've generally tried to keep my check size reasonably static”

Q When do you think about bringing in CPOs? If we're a scaling company, I often see the transition required. When would you advise founders on now you need a CPO slash head of product, nor just jack of all trades?

A Again, most people do this probably too early. So that's like the common theme here. So I made this mistake myself at Bump. We got to like, maybe it was 30 employees. All of our vanity metric graphs were like going way up. We were feeling really good. And I thought our company is going to need to like figure out our business model. Eventually we're going to need to keep scaling and hiring. I'm going to have to do more fundraising. I better hire a product lead to take over from me and I'll go focus on those other things. And I did. He was awesome. And then we realized, oh, actually we have not Really solved product market fit. Sure. We have a lot of people who use our product, but like we can't figure out a way for them to pay us. We can't figure out anybody else in the world who will pay us for access to those users in some way. So really we don't have product market fit. Whoops. And I realized the success or failure of my company actually has to do with what this person I just hired is going to go figure out for me. Generally, that is a very dangerous place to be where the founders or the CEO of the company is not working on the most important problem for the company's success. I just Found at some point that I need to be doing that. It's my responsibility for this company to be successful. I better be working on this. And now all of a sudden we have two people who want to be …

AI assessment note: “Again, most people do this probably too early. So that's like the common theme”

Q self-selecting talent. When I spoke to so many of your team, what really shone through was just what a incredibly kind of loyal and. And really bought in Google you, you, you are. And I think the thing that kind of shocked me, and we chatted about it a little bit before the show, is why did you decide to leave Google most recently, and what was behind that thinking?

A Yeah. Um, well, I, so I, I had been at Google nine years. I left last week. Um, and I would say the trigger for me finally. So, so the reason I stayed first, the reason I stayed for nine years, right. Um, is I just really wanted Google photos to exist. Um, it was a thing. It was an insight that we had at our startup that like, there was this big problem and we thought we had an idea for how it would work to solve it. Um, and I just really wanted it to work. You know, I would say like, Uh, maybe this is a slight digression, but I had a chip on my shoulder. Uh, we had like worked on our startup for five years. It was like this, in many ways, like a very big success. We had a hundred fifty million users. We were number one or two on the app store in almost every country. Um, And yet, like, we didn't figure out a way to have it be a sustainable business, um, and so we decided to ultimately sell it. Um, but, like, I didn't want those five years to kind of be all for naught. For everybody on our team, everybody who contributed to the success of Bump, I wanted something, like, long to come from it. And so I just had this chip on my shoulder that, like, I needed this next thing to be successful. Um, and so I stayed for a long time to make sure that would be the case. Last year, I, um, was diagnosed with surprise leukemia. Uh, and so I was out for a year doing chemo and fighting that. U…

AI assessment note: “Last year, I, um, was diagnosed with surprise leukemia.”

Q Can I ask, when does one rely on that gut intuition, more human incentive Versus, let's look at the numbers, let's look at the data, and let's go to the cohorts, because I'm sure there's a time for both. When is the time for data?

A Data is a fantastic tool. Um, it tells you a lot of information all at once that can be very useful directionally about where to go. Um, I dislike most aggregated statistics. Like, you know, yesterday we had this many users upload a photo, or there were this many comments on this post yesterday, um, because they can be very misleading about what's actually happening for individual users. So I much prefer looking at data around individual users in an, in a, um, Large scale way. So like, let me explain what I'm talking about. Uh, you mentioned cohort graphs. Like, I think that's the number one thing you should look at when you're building a new product is what is your cohort retention curve look like? If a hundred people start using your product today, who's using it tomorrow? Who's using it the next day? Who's using it the next day? And what you want to find is those cohort graphs get flat. You want the same set of people eventually to keep using it over time. Once the other folks who've decided it's not for you have left, um, because that tells you about individual users, right? Um, The other tool that I really like, um, and there's a variety of-

AI assessment note: “when you're building a new product is what is your cohort retention curve look like”

Q When do you think about bringing in CPOs? If we're a scaling company, I often see the transition required. When would you advise founders on now you need a CPO slash head of product, nor just jack of all trades?

A Again, most people do this probably too early. So that's like the common theme here. So I made this mistake myself at Bump. We got to like, maybe it was 30 employees. All of our vanity metric graphs were like going way up. We were feeling really good. And I thought our company is going to need to like figure out our business model. Eventually we're going to need to keep scaling and hiring. I'm going to have to do more fundraising. I better hire a product lead to take over from me and I'll go focus on those other things. And I did. He was awesome. And then we realized, oh, actually we have not Really solved product market fit. Sure. We have a lot of people who use our product, but like we can't figure out a way for them to pay us. We can't figure out anybody else in the world who will pay us for access to those users in some way. So really we don't have product market fit. Whoops. And I realized the success or failure of my company actually has to do with what this person I just hired is going to go figure out for me. Generally, that is a very dangerous place to be where the founders or the CEO of the company is not working on the most important problem for the company's success. I just Found at some point that I need to be doing that. It's my responsibility for this company to be successful. I better be working on this. And now all of a sudden we have two people who want to be …

AI assessment note: “most people do this probably too early. So that's like the common theme here.”

Q I want it. There's bits of your career, which I want to kind of unpack, but I do have to ask and Juan asked this one as well. He said, what are the ones, two biggest lessons for you from your time founding bump and the transition to Google photos?

A Yeah, I would say. For startups, like when you've done a startup, you realize that kind of the definition of a startup is you start with absolutely nothing and you have this big vision for like what it could be one day in the future. But to get from there to there, um, requires like you to do basically everything, uh, at least at the very beginning. So I think that the number one takeaway from the startup is just this like ownership mentality that you have, which is like, we have nothing. There's no one else. It's just me. I must make it work somehow. I must do everything that's required. So in the early days of Bump, like, you know, myself and my two co-founders, we wrote all the code. We did all the designs. We thought about all the marketing strategy or hacks or ploys we could use to get anybody to hear about it. Um, every PR, every press interview, like that was all just us. We had to learn it on our own. And I think that's a skill, this like generalist skill, uh, that generally people don't have at big companies because they don't need to. Uh, there's, there's experts at big companies that do everything, but, um, I think magic happens when you bring all that stuff together in a very small number of people's heads and they can do like pretty amazing things.

AI assessment note: “number one takeaway from the startup is just this like ownership mentality”

Q instinct is the primary tool, Dave is the master. He wins. Um, brilliant. Uh, I hope I have team members that say the same about me. Um, but, uh, my question to you is like, how do you think about using your own gut instinct? How do you think about downside protection given that, you know, you are willing to have the variables and how do you think about that?

A Yeah. Um, so people often refer to like using your gut or trusting your gut. Um, and I think often they say it in kind of like a, A derogatory way or like, oh, your gut is just like this thing on a whim that you should be careful about. Um, I have a different view. I think your gut is the world's most sophisticated machine learning model ever created. Um, it is like the best supercomputer we've got. Um, and there's billions of them on the planet and they've been trained over like tens of thousands of years with tons of training data. Um, the people who wrote the code, you know, there's like generations of people before you that have been tweaking this code For you. And it's now yours. You can now use it. Um, what an amazing thing. And I think we don't, we discount it. Um, but instead we should leverage it. My God. So what I try to do is put as many inputs into this machine learning model that you can learn as much as you can about your customer. Try as many experiments as possible and put it all into your brain and then see what your gut says you should do next. Uh, and I think that is actually like a quite, um, quite analytical way to approach the world. You're just using a tool that people aren't used to using.

AI assessment note: “put as many inputs into this machine learning model... see what your gut says”

Q I want it. There's bits of your career, which I want to kind of unpack, but I do have to ask and Juan asked this one as well. He said, what are the ones, two biggest lessons for you from your time founding bump and the transition to Google photos?

A Yeah, I would say. For startups, like when you've done a startup, you realize that kind of the definition of a startup is you start with absolutely nothing and you have this big vision for like what it could be one day in the future. But to get from there to there, um, requires like you to do basically everything, uh, at least at the very beginning. So I think that the number one takeaway from the startup is just this like ownership mentality that you have, which is like, we have nothing. There's no one else. It's just me. I must make it work somehow. I must do everything that's required. So in the early days of Bump, like, you know, myself and my two co-founders, we wrote all the code. We did all the designs. We thought about all the marketing strategy or hacks or ploys we could use to get anybody to hear about it. Um, every PR, every press interview, like that was all just us. We had to learn it on our own. And I think that's a skill, this like generalist skill, uh, that generally people don't have at big companies because they don't need to. Uh, there's, there's experts at big companies that do everything, but, um, I think magic happens when you bring all that stuff together in a very small number of people's heads and they can do like pretty amazing things.

AI assessment note: “the number one takeaway from the startup is just this like ownership mentality”

Q instinct is the primary tool, Dave is the master. He wins. Um, brilliant. Uh, I hope I have team members that say the same about me. Um, but, uh, my question to you is like, how do you think about using your own gut instinct? How do you think about downside protection given that, you know, you are willing to have the variables and how do you think about that?

A Yeah. Um, so people often refer to like using your gut or trusting your gut. Um, and I think often they say it in kind of like a, A derogatory way or like, oh, your gut is just like this thing on a whim that you should be careful about. Um, I have a different view. I think your gut is the world's most sophisticated machine learning model ever created. Um, it is like the best supercomputer we've got. Um, and there's billions of them on the planet and they've been trained over like tens of thousands of years with tons of training data. Um, the people who wrote the code, you know, there's like generations of people before you that have been tweaking this code For you. And it's now yours. You can now use it. Um, what an amazing thing. And I think we don't, we discount it. Um, but instead we should leverage it. My God. So what I try to do is put as many inputs into this machine learning model that you can learn as much as you can about your customer. Try as many experiments as possible and put it all into your brain and then see what your gut says you should do next. Uh, and I think that is actually like a quite, um, quite analytical way to approach the world. You're just using a tool that people aren't used to using.

AI assessment note: “I think your gut is the world's most sophisticated machine learning model”

Q Now we're going to move to David Lieb, founder of Bump and has run Google Photos for the last 10 years. How do you structure your hiring process for a product team? What is the process?

A Well, if you're early stage, your product team maybe is one person. Maybe you're hiring one PM. This is another meta mistake is people think like I've raised my seed round. I have four million dollars. Like let's go hire a team of 20 people. That's what I see all my cohort members doing. That's probably the wrong thing to do unless you really have product market fit already, but let's assume things are on fire. Like, oh my God, we've got to build all these things. Like great news. I would say at the early stage, the number one thing I would Filter for and find some products that you think are good and some products that you think are bad. Ask them about them. See if they've ever gotten to the point of thinking about why a product they love is great versus why a product that they don't love is bad. And can they explain it to you? Do they have an opinion on it? These are things that are very squishy. It's very hard to, like, how, how would you, like, write a multiple choice test that would test that? It's very difficult. But these are the things you're looking for because they show that the person has, like, Thought from first principles about why a thing that they see is true in the world is true, and the why is what matters, and the skill of asking why and learning why or figuring out why is the number one skill for building early stage companies, because that's what you're doi…

AI assessment note: “find some products that you think are good and some products that you think are bad”

Q Final one. What recent company product strategy, other than Google or Google Photos, have you been most impressed by?

A This is not, the first one that comes to mind is not recent at all, but I'll, I'll say it anyway, is the, the Tesla master plan. I think it was written in, like, 2006, when Elon just decided to, like, write down the master plan for how Tesla would be successful, and then he did the master plan part two, I think, like, 10 years later, there might be a new one coming. Um, I love that he can articulate, like, the set of things that might become true, that if true, yield, like, a big change to the world in a positive way. Um, I think that's what every product leader needs to do is, um, articulate, like if we succeed and to be fair, like some things have to fall in our favor, um, how can this be so important and so big in the world? Cause that's what gets people excited.

AI assessment note: “the first one that comes to mind is not recent at all”

Q Can I ask, when does one rely on that gut intuition, more human incentive Versus, let's look at the numbers, let's look at the data, and let's go to the cohorts, because I'm sure there's a time for both. When is the time for data?

A Data is a fantastic tool. Um, it tells you a lot of information all at once that can be very useful directionally about where to go. Um, I dislike most aggregated statistics. Like, you know, yesterday we had this many users upload a photo, or there were this many comments on this post yesterday, um, because they can be very misleading about what's actually happening for individual users. So I much prefer looking at data around individual users in an, in a, um, Large scale way. So like, let me explain what I'm talking about. Uh, you mentioned cohort graphs. Like, I think that's the number one thing you should look at when you're building a new product is what is your cohort retention curve look like? If a hundred people start using your product today, who's using it tomorrow? Who's using it the next day? Who's using it the next day? And what you want to find is those cohort graphs get flat. You want the same set of people eventually to keep using it over time. Once the other folks who've decided it's not for you have left, um, because that tells you about individual users, right? Um, The other tool that I really like, um, and there's a variety of-

AI assessment note: “I think that's the number one thing you should look at when you're building”

Q Now we're going to move to David Lieb, founder of Bump and has run Google Photos for the last 10 years. How do you structure your hiring process for a product team? What is the process?

A Well, if you're early stage, your product team maybe is one person. Maybe you're hiring one PM. This is another meta mistake is people think like I've raised my seed round. I have four million dollars. Like let's go hire a team of 20 people. That's what I see all my cohort members doing. That's probably the wrong thing to do unless you really have product market fit already, but let's assume things are on fire. Like, oh my God, we've got to build all these things. Like great news. I would say at the early stage, the number one thing I would Filter for and find some products that you think are good and some products that you think are bad. Ask them about them. See if they've ever gotten to the point of thinking about why a product they love is great versus why a product that they don't love is bad. And can they explain it to you? Do they have an opinion on it? These are things that are very squishy. It's very hard to, like, how, how would you, like, write a multiple choice test that would test that? It's very difficult. But these are the things you're looking for because they show that the person has, like, Thought from first principles about why a thing that they see is true in the world is true, and the why is what matters, and the skill of asking why and learning why or figuring out why is the number one skill for building early stage companies, because that's what you're doi…

AI assessment note: “find some products that you think are good and some products that you think are bad”

Q I think that singular accountability is so key. Final one before the quick fire, and it's, you know, you've invested in some of the most unbelievable companies. Your track on the angel side is astonishing. When you think to the incredible companies you've invested in, how has your mindset on product changed? Through those angel investments and through that portfolio building.

A Yeah. I guess the biggest learning is companies are very different. There are like totally different. I've invested in such a broad scale of companies that like things that I thought were like, ah, that was the reason like we succeeded on Google photos. I then look at another company. I'm like, well, they didn't do any of that. Um, and yet there are companies worth billions of dollars. Like, wow, there's different ways to solve these problems. Um, I would say. That's probably the number one learning is like, there's so many different ways to do this. The things that I've found are common across all the successful companies that I've invested in. Like, that's an interesting question. What, what is it about these companies? Um, generally I would categorize it as there's a person with a chip on their shoulder that wants to prove to the world that they can succeed at something or prove all the haters in the world wrong. Um, that's one. Number two is that person, usually the same person has to be like incredibly charismatic Because they need to get really smart discerning people to choose to work for them when it probably doesn't make any sense for those people to work for them. Uh, that's gotta be true. And, uh, what else? They have to approach the world with this mindset that they will succeed. Um, and I would say, like, the best founders I've invested in are people that when I me…

AI assessment note: “The things that I've found are common across all the successful companies that I've invested in.”