The Exchanges

Q Would you say that it's, are the founders charismatic, or is there a myth of charisma that leads to this sort of people following you? You're seeing thousands of them a year, effectively. Like, what A percentage of them are using charisma versus idea. Like, how are you getting people compelled to follow you into this pursuit?

A Yeah, I think charisma helps, but I think it's more about the power of the vision. So, you know, Steve Jobs, super charismatic, obviously, but Bill Gates and Mark Zuckerberg, I wouldn't say are, you know, a standard deviation, more charismatic than most people. But I think that both of them believe they were working on something important, and that they were doing kick ass exceptional work, and they attracted people who believe their vision and wanted to work hard on a great project with them. And so I think that that's really the defining characteristic is they have a vision that compels a set of like-minded people, and they all feel like they're in on a secret together, and they believe that they're gonna convert everybody else to their point of view until one day everybody has their point.

AI assessment note: “charisma helps, but I think it's more about the power of the vision.”

Q How did you come up with the mental models?

A I wish I could say it was by being smart, but it was, it was as much as by being dumb. So like, if you look at what Charlie Munger talks about is great investors, they, they read a lot. They're genuinely curious and they try not to repeat avoidable mistakes. And so, so I went back in time And I looked at all the companies that we had passed on that we should have invested in, Airbnb, Pinterest, um, who else, PagerDuty, Anaplan. I did a forensic analysis of it. I found the initial intro email and who the source of that intro was. I looked at all the notes we took. I looked at the pitch deck at the time we saw it. And I even looked at the notes between us and the pass email I sent. And then I studied what happened in that round. What price did it get done at? What were all the future rounds? Who said yes? Why? And then I, when I get a chance, I actually interview the founder a few years later. And the reason this is so important, Shane, is that startups are so random that you can literally forget how it happened. So, for example, we invested in Lyft when it was Zimride back in 2010, and it would be very easy for us to believe we knew more than we knew at the time we made that investment. Same with Twitter, same with Twitch, same with Okta. And so in the ones that we did say yes, and it was a big win, we try to not breathe our own fumes too much. And we similarly try to go back in…

AI assessment note: “I went back in time And I looked at all the companies that we had passed on”

Q Can you walk me through those miracles in a little bit more detail? Like, uh, what is the insight? What does that mean?

A Sometimes people will ask me, how do you get a good startup idea? And the counterintuitive answer to that question is don't try to think of a startup. And people are like, what are you talking about? That's, that's a crazy answer. Well, it turns out that most of the great startups come from a great insight. And a great insight usually occurs when someone is living in the future and they notice something that's missing. So if you take somebody like Mark Andreessen, he didn't try to think of a startup. He was in a computer lab at the university of Illinois, and they were trying to experiment with the internet and share notes and documents that collaborate. And the internet wasn't useful to him. And so he built mosaic, both the browser and the web server. To make the internet immediately more useful for him. And so insights are kind of, um, a play on William Gibson's, uh, quote. Um, he wrote a bunch of sci-fi, cyberpunk novels. The, the, the future is already here. It's just not evenly distributed. And what, what you find is that most of the really great startups are people who are living in the future and they notice something that's missing and then their job, it's, it's almost like if, if the metaphor of customer development is experimenting, which we'll get to, right? The metaphor of insights are the time traveler and somebody is living in the future and they notice something …

AI assessment note: “a great insight usually occurs when someone is living in the future”

Q What are some of the exclusion filters? That you use to avoid the noise or avoid situations that you don't want to be involved in?

A Um, well, sometimes it could be a surprise. So one time we were pitched by somebody who said that they were in the Guinness Book of World Records for Tetris. And, you know, we had this great meeting, a bunch of other stuff, and, and we were going to have a second meeting. And just on the side, we're like, Huh. I wonder if he really does have the Guinness Book of World Records in Tetris. Let's just check that out. So we looked at the Guinness Book of World Records. The guy wasn't in there. And we looked around for more things. Wasn't in there. And so then we, we called and said, Hey, uh, I know this is kind of crazy, but we're doing our diligence. And, uh, you know, what year were you in the Guinness Book of World Records for Tetris? And, uh, they said that, you know, I'll get back to you and, you know, kind of didn't call us. And then they said it was at the challenge book of world records, uh, which I've never heard of. I don't even know if it exists. And then next thing, you know, they kind of go radio silent. They're like, ah, we raised the money somewhere else. But like, I find that when somebody lies about something that is in no way germane to their business, I'm like, that's, that's a red flag for me, right? Like, that's somebody who has a difficult relationship with the truth and facts and reality. Elon Musk, right, will sometimes say, well, this car's gonna be ready be…

AI assessment note: “when somebody lies about something that is in no way germane to their business”

Q That's a great way to put it. Let's start with insight models. So some of the insight models I want to talk about, let's start with the earned secret. What is the earned secret?

A Yeah, so a great example of the earned secret is Brian Chesky at Airbnb. So it's basically Some work that you did to discover something that most people would think is non-intuitive. So the case of Brian Chesky, he didn't have any money. He wanted to go to a design conference. He'd just come out of Rhode Island School of Design, and he couldn't afford a ticket. So he put his apartment and an air mattress up on Craigslist, and he got four to 500 people saying they wanted to rent his room. And, you know, keep in mind how counterintuitive that, that must have been in 2007. You know, a year later, there was the Craigslist killings, and, you know, who's gonna, who's gonna wanna stay in a stranger's apartment, and who's gonna wanna invite a stranger into their apartment? That, you know, that's crazy. But he had hundreds of people wanting to do this. But Brian took it a step further. He said, huh, that's interesting. Why did that happen? Oh, the town was out of hotel rooms. Why was the town out of hotel rooms? There was a conference. Huh, I wonder if that happens a lot when there's conferences. And by the way, what's up with a hotel business anyway? What's their business model? How long have they been in business? Why do people go to hotels? And he started to realize that the main reason people go to hotels is because of trust. And he started to realize, well, if you could have a trus…

AI assessment note: “Some work that you did to discover something that most people would think is non-intuitive.”

Q When you're telling the story and you're sort of assembling your team, how important is it to seek out contrarian team members?

A I think that that helps a lot. The way I like to say it is, uh, most startups don't realize that you have to have Contrarian insights, as well as contrarian recruiting strategies. So, for example, if you look at the PayPal mafia, none of them were established in tech before they joined PayPal. So you had Peter Thiel. He's a law school guy. You had Max Levchin, who's a, you know, a young engineer. They recruit Reid Hoffman, who'd been at eWorld at Apple, which was a failed initiative. And then he had done SocialNet, which is a failed startup. He recruited Keith Raboy from a law firm. You know, David Sachs, uh, all these people in the PayPal mafia were people that he knew in his Stanford network and people that Max knew in his University of Illinois network. But like discovering undiscovered talent is a key superpower of great startup teams. Because like, if somebody is making a ton of money at Google or Facebook, Why are they going to join your startup for less money and more risk and less status? Most people won't. But if I'm Keith Raboy and I'm in a law firm and I'm like, eh, this job's okay, but it's kind of soul crushing. And then Peter Thiel says, hey, we're going to go on this adventure. We're going to change how money happens. We're going to get to work together again. That's a pretty good story. And so like, I like to say that the great startups find people who are one d…

AI assessment note: “I think that that helps a lot... you have to have Contrarian insights, as well as contrarian recruiting strategies.”

Q How important is the team with the idea? Like when you're evaluating these opportunities, you could have this most like execution sounds like it's sort of multiplicative by zero, right? If you have a great idea and you can't execute it, you're Just, it's not going to go anywhere.

A That's right. So the, you know, the insight is a, is a necessary prerequisite because it really contains the potential energy of the idea, right? And you want to have, you want to have an insight that's so powerful that if you're right, you change the rules in markets and you create this massive movement with massive asymmetric upside. But to your point, you also have a team and Teams are different in startups than they are from companies because I like to say that in a, in a company, it's more like a marching band. People need their sheet music and they need to dance in formation and play their instruments in exact terms. Whereas, um, I don't know if you've ever been to New Orleans and seen improv jazz, but it's, you know, with improv jazz, you have a tune, but you're kind of going in a direction and usually the lead performer Starts to do a riff and goes in a different direction. The rest of the band just kind of goes with them. And so that, that tune is never going to be played exactly that same way ever again. And, and startups are a lot like that. They're these close knit teams of people who, um, enjoy the craziness of the ambiguity and they, they enjoy improvising and they understand that startups are ambiguous and that you're, you're trying to improvise your way to success rather than, you know, have Objectives and key results and a lot of the basic management frameworks…

AI assessment note: “the insight is a, is a necessary prerequisite... But to your point, you also have a team”

Q What are some of the other questions you try to hone in on?

A I, I try to find out how committed the team is. So, so for example, I don't react super well to, uh, we're, we're all still working at our jobs and we can't, we can't wait to leave as soon as we get funded. So I, I, uh, I'm like, look, if you're not passionate enough to leave your job and do this full time, why would I be passionate enough to write a check? And so, I like people who say, I'm doing this. And they would do it even if they couldn't raise the money, because they know they'd find a way. And it's like, I'm doing this. I'd love to have you on board, but this is happening. Are you in or are you not in? And so, um, you know, teams that are tentative, teams that are sort of, I'll, I'll do it when I get validation of the funding. Not, not my, not my tempo.

AI assessment note: “I try to find out how committed the team is.”

Q Wasn't that the, the mini mills and the rebar too that Christensen used as an example?

A Absolutely. Right. And the pattern's always the same. It's the incumbent says, oh, you're going after a market or you're charging with a pricing model, or you're, you're doing something that I don't really care about. It's not really a threat to me. And so they give the upstart time to satisfy the early desperate customers who are unsatisfied. And then over time, as they build their skills and their capabilities, they become More effective attackers in the core markets of the incumbents. When the incumbent sees you, the, the notion of your attack will be too obvious, and their ability to respond to the attack will be too powerful. And so part of the value of being in non-consensus and right is not just that you find desperate customers who value what only you can deliver for them. The incumbents are disoriented by how you go to market because the insight around your idea is novel, not just in the technology, but in how you deliver the technology.

AI assessment note: “Absolutely. Right. And the pattern's always the same.”

Q What do you do when you realize you've made a mistake backing a particular company? Do you sort of cut bait and like, okay, well, they'll sink or swim. Do you dive in? Do you try to sell it? What do you, what's your thought process?

A Well, I try, I try very hard to acknowledge, first of all, this was a failure on my part because I picked Either the wrong team or the wrong opportunity, which is my job on a very fundamental level, right? That's a very core part to how I create value in my business. But what I found is that when you replace the CEO involuntarily, you almost always make the company worse. So if a CEO says, look, I'm kind of in over my head and I need your help recruiting somebody. That can work. Uh, but if you have a fundamental difference of opinion with the founding team or the CEO, usually my default is, hmm, that's a shame. It's your company. I'm not gonna, I'm not gonna try to run it for you. Cause I think that if I tried to run it for you, it'd be even worse off. And so, um, I mean, who am I to think that I could spend two hours a week running somebody's company? You can't.

AI assessment note: “It's your company. I'm not gonna, I'm not gonna try to run it for you.”

Q the things that we had chatted about on the phone that I really liked about floodgates approach was the concept of mental models that you guys have developed. And we're going to, we're going to dive into those during this conversation, but can you give me an idea of how you landed on the idea of mental? What is a mental model and how did you land on this idea?

A Well, without, uh, flattering the interview host too much, I, uh, have been a big reader of Farnham street for awhile and have, have liked a lot of your podcast episodes. I'm friends with Naval and, uh, just got Into the notion of mental models. And then I went down the rabbit hole and I read books about and by Charlie Munger. I read this book that I'm sure you've heard of thinking fast and slow by Kahneman. And so I got, I got interested in the following question. There seemed to be a lot of mental models for investing or operating companies, but they didn't seem to apply to startups. So I looked at Companies that we should have invested in, like Airbnb, and we passed on it, and none of those mental models would have been helpful. And that's a problem, because had I invested 250,000 dollars in Airbnb at the time, it'd now be worth a billion dollars. And then there's other companies that we did decide to invest in, Twitter, Twitch, Okta, Lyft. Well, Okta was called Sasher. Twitter didn't know what to call itself. Twitch was Justin.tv. And so Lyft was Zimride. And so I started to realize that we didn't really have mental models for startups. And so I got really interested in the question of how, how would you come up with mental models for startups or is it just purely random? And so I spent a couple, three years trying to dig into that question.

AI assessment note: “I started to realize that we didn't really have mental models for startups.”

Q And then I say you have an insight, you have a team that are working towards this insight. And then the next thing you have to do is the value. How do you test the value proposition? Like just because you have an idea and you have a decent team doesn't mean that people are going to pay for it.

A That's right. So once you have a non consensus and right insight and you have a team of people who have the stuff, Then, then you go into this mode of, um, the value phase or what I sometimes call value hacking. And what you're trying to do in value hacking is figure out what you can build that is unique, that people are desperate for. And if your insight is truly powerful, it will connect with somebody desperate because you'll be solving a problem in a novel way that's never been solved that way before. And that's An order of magnitude noticeably better for those people. And so the trick in value hacking is to say, okay, in the insight phase, I posed a truth. Now in the value hacking phase, I need to discover the truth. I need to validate that it's, that, that my idea was true. And the way that I do that is I find the people out there who believe what I believe and who are desperate for the product that I'm going to deliver. And then and only then do you start to, you know, only when you've proven your value hypothesis do you start to test your growth hypothesis.

AI assessment note: “I find the people out there who believe what I believe and who are desperate”

Q Are there any other lies that, like, what are the different types of lies that we tell ourselves?

A Most of the lies that we tell ourselves, uh, Relate to not facing reality, right? So startups are impossible. And usually there's, there's kind of this big boogeyman risk that if, if we don't overcome this risk, we're going to fail. And so I'm a big fan of like getting that boogeyman right front and center and punching into the face and like going dead at it. And so the best founders I know, know what are the most important risks that stand between themselves and success. And they tackle those risks early, because if you eliminate those risks, you massively improve the probabilities of winning, as well as you make the company more valuable, you learn more faster, you know, a whole bunch of good things go your way. Whereas the, the, the reverse of it could be the company wants to give its product away, but they don't have a good strategic reason to give it away. And, and really what you, what you learn is they're giving it away because they don't have enough confidence or value proposition to charge a price for it. And it's like, okay, well, you know, we, we'd be better off charging a price when the product's half done, because, you know, then we're going to find out how visceral is the pain and how compelling is the value that we're here to deliver for them.

AI assessment note: “Most of the lies that we tell ourselves, uh, Relate to not facing reality”

Q Yeah. How do you say no in those cases and maintain that relationship?

A It's hard, right? And I, I, I'm probably too nice about it. I'd probably say yes too often. Uh, and, and, you know, in the end, you, you only have so many hours of the day and you got it, you got to decide how you want to spend it, who you want to spend it with. And, and I also think it's good to just have some serendipity and not be too dogmatic about who you're going to meet and who you're not going to meet. And so you just, you just try as hard as you can to meet with as many people as you can. The other hard part about it is the more people that you meet with, the more follow-ups you have. This has always been hard for me. You want to help everybody. You want to make intros for everybody. And, and, um, if you take more meetings, you're going to have to do more of those and you have, it's more likely you can't get it all done.

AI assessment note: “I'm probably too nice about it. I'd probably say yes too often.”