Investing Billions - E5: Jonathan Hsu | Why Companies Fail Without PMF, Examining Tribe Capital’s Proprietary Data Source, and What Experienced VCs All Regret

Starting point is 00:00:00 That's kind of the underlying thesis of the firm. What really matters in the long run is that pattern of growth and that pattern of product market fit. It sounds kind of obvious, but it's actually like a bit contrarian, right? Because the vast bulk of investors are like, the founder matters, the team matters. We think the founder matters, we think the team matters. But in the long run, remember my story about Facebook, in the long run, what matters is that pattern of growth, not the founder. Not an approval. Not an approval. Welcome to the Limited Partner Podcast, where we talk about venture capital through the lens of limited partners. I'm your host, David Weisberg, co-founder and head of venture capital at 10X Capital.

Starting point is 00:00:53 The world of LPs can be notoriously private and discreet, but on this podcast, we speak candidly with limited partners about their true feelings on venture, the ecosystem, and 2023 and beyond. Jonathan Hsu, pleasure to be introduced to you. Welcome to the Limited Partner Podcast. Special shout out to Abe Othman, head of data science at AngelList, who previously interviewed, who was kind enough to make the introduction. First, Jonathan, we're going to start with your origin story. Every great superhero has an origin story. So what is your origin story? Yeah, yeah. Happy to walk through it.

Starting point is 00:01:18 Thanks so much for having me here. You know, my own background, I actually started life as a physicist. I did my undergrad and PhD in physics. I did my PhD was actually studying quantum gravity at Stanford, truly sort of the height of uselessness. And towards the end of it was just clear I didn't want to be an academic. So ended up going into technology. This was like 2006 era.

Starting point is 00:01:38 Spent a little time at Microsoft. Ended up co-founding one of these early social gaming companies. This was sort of in the era of, the era of sort of crappy games on Facebook. Co-founded one of them back in 2006. It was called Super Poke, if you happen to remember that. And we ended up selling that to a company called Slide, which is Max Lepchin's company. In between PayPal and Affirm, he had founded a social gaming company and he acquired us.

Starting point is 00:02:04 And so I went to run data for Max Lepchin for a couple of years before being recruited over to facebook in 2009 joined facebook at that time i was about the fifth data scientist roughly at the company really ended up building out the whole data science and analytics org for facebook by the time i left five years later in 2014, you know, company was like, you know, 15,000 people. I think the data science org was 200 people at that point, half of it, you know, reporting to me, you know, at that point had been spending a lot of time in big companies. So, you know, really wanted to try investing, was interested in finance and knew the folks over at Social Capital, which is, you know, a firm run by some ex-Facebook folks, and joined Social Capital in 2014,

Starting point is 00:02:46 which is where I was for four years before we spun out to create Tribe. So going to Berkeley, so tell me about that experience. What did you study, and how was that experience? I went to Berkeley initially wanting to be pre-med, wanting to be a medical doctor, but very early on sort of fell in love with math and physics and sort of just wanted to be a physicist. I wanted to be Richard Feynman. What about physics intrigued you? There's not many venture capitalists that are in the physics space.

Starting point is 00:03:15 Why physics and why VC? Back when I was, you know, coming up sort of working with Max Lepshin, data science wasn't a term yet, right? What happened sort of in the late 2000s, there was sort of this big data, big bang that occurred in the mid 2000s, you know, in about 2005 to 2008, all of a sudden it became cheap to store and compute on very large amounts of data, right? You know, if you go in the 90s, like people didn't have a ton of data because it was expensive and impractical to store it. But then starting in the mid 2000s, all of a sudden people had tons of it. And so if you have a bunch of data, what do you do? Well, you hire some PhDs to throw at it, let the PhDs figure it out. And I was one of those PhDs in that era, that early era

Starting point is 00:03:52 of, you know, of data science. And really what we were trying to do is we were just trying to figure out, we have these big data sets, how do we leverage them for business purposes, right? Whether it be for building product, generating strategic intelligence, right? Helping us with product strategy, whatever it is, how do we use the data to get to the next level? And sort of, there was a small group of us, you know, who were right of that era. Several of us happened to be PhDs in physics, you know, who were all exploring these things. And invariably, you know, when you have a bunch of physicists around it, they try to make models out of it. Well, let's create, let's use this model, that model. And, you know, when you have a bunch of physicists around it, they try to make models out of it. Well, let's create, let's use this model, that model. And, you know, we developed a bunch

Starting point is 00:04:27 of these different analytical techniques to leverage that data. This really sort of came to its really full expression at Facebook. This was an early stage, move fast and break things. Mark Zuckerberg famously called that era of Facebook. How was the early days in Facebook? And what did you learn from that experience

Starting point is 00:04:45 of working with Shamath, working on the growth team and working at Facebook as an organization? Yeah, I would call it, it was kind of middle days. It wasn't super early, right? This was sort of the era of a few thousand people at the company. So it wasn't like tens of people, but it was a few thousand. And it was nonetheless a very interesting,

Starting point is 00:05:01 it was sort of in that interesting zone between, because it had gotten to that scale of employees very quickly, right? And so it still had a lot of these behaviors internally that were much more small company oriented, you know, which was included the ability to move super fast, which was really good. And, you know, there were sort of, it was an important episode there, right? In 2009, when I joined Facebook was just the era when it was just barely clear that it hit, well, it was more or less clear that it had defeated MySpace and that it was going to win. Right. At that point, those other social social networks were still bigger, but the speed was there. Right. For Facebook. What allowed Facebook to really leave the pack?

Starting point is 00:05:38 A lot, as you mentioned, it's not just MySpace. MySpace, people forget, had a good outcome with a half billion dollar exit. There was other social networks. What really differentiated the team at Facebook? Yeah, you know, I think this is this is sort of the fundamental question and the fundamental driving question for all of my work since Facebook. Right. And I think there's a bit of intellectual humility of knowing of noting that there's no one thing and it's kind of foolish to think it's just one thing.

Starting point is 00:06:02 Right. And I think that's there's sort of an aspect of it. That's the team, right? The team was special. The era in history was just right, right? Like they were able to sort of catch, there was this moment, right? They sort of did a really good job through the desktop era. And then when the phone happened, when the mobile phone happened,

Starting point is 00:06:16 Facebook did a really good job of taking advantage of it. Facebook was able to take advantage of a lot of that, right? And WhatsApp, so on and so forth. And, you know, what we believe here at the tribe is that the underlying bit is really about the sheer momentum of product market fit, right? Once you get the right pattern of product market fit, once it starts really moving, it'll just keep going regardless of anything, right? And it just had the right pattern. And when I say that, you know, I guess the thing, the things that I'm thinking of when I say product market fit and, you know, something that just keeps growing when i say that i'm i'm in i'm referring to things like

Starting point is 00:06:48 religions right like if you look at early christianity it behaved the same way sort of grew from you know when when jesus christ died it was just a few hundred people right and then sort of you know 300 years later you know by the time of constantine it's like 30 million people it just happens it just it's able to maintain its momentum through a long era. And obviously, Facebook's probably not able to do it over a 400 year era. But over that era, it was able to just generate this momentum of adoption and product market fit that was just right. And it was able to drive its own growth. And it was able to do that all the way out. So we'll move on to social capital in a bit, but first let's cover SuperPoke. So behind some of the best venture capitalists,

Starting point is 00:07:30 there's always founder experience. Tell me about the SuperPoke experience. Me and a couple of friends founded it while I was at Microsoft, while we were at Microsoft. We kind of did it on the side as a little side project. It ended up growing really quickly, sort of, this was in 2007 and we didn't, we never raised any outside capital. This was in 2007. And we never raised any outside capital. We started building it. We were all fairly new graduates. None of us had really any experience in startups.

Starting point is 00:07:55 It wasn't like today where everybody knows everything about cap tables and investing. Back then, we didn't really know. We just built something because we thought it was amusing. And it grew really quickly. And then it was actually Keith Raboy that reached out with an interest to acquire us. I think we were only two months old or so. We were like, oh, this looks cool. Let's go work at a startup in San Francisco. We didn't really think through, oh, did we build something super valuable here, blah, blah, blah. It was more like, oh, these guys, Max Lepchin, he's a big guy. They want to acquire us. So we were kind of starstruck by that and we took an offer and we went and got

Starting point is 00:08:23 acquired really quickly. Let's go from the transition from Facebook, which is obviously a once-in-a-generation company to Social Capital. How did you get involved? How did you get recruited in Social Capital? And what was the original vision for Social Capital? Yeah. So Social Capital was founded in 2011, after Chamath had left Facebook. It was him and his two co-founders, Ted Maidenberg and Mamoun Hamid. Ted is now one of our co-founders at Tribe and Mamoun is over at Kleiner. And I joined them in 2014. And really the idea when I joined them, there were a couple of folks there from Facebook. And the idea of me joining was really to sort of continue building out this data-driven growth mentality and applying it to

Starting point is 00:09:03 the venture capital world at large, right? The vision back then was, this was 2014, right? So Andreessen Horowitz at that point was a couple of years into building out their market development team, right? Prior to Andreessen's market development team, venture firms were all kind of the same. They were all just kind of like small boutique investment firms. Andreessen Horowitz had this idea of, oh, we're going to build a capability. We're going to invest in this capability, market development. We're going to help our portfolio companies with enterprise sales, right? And it became a thing, right? And so, and that social capital part of the idea was, oh, let's take the growth team thing that we

Starting point is 00:09:37 did at Facebook. Let's make that like our special thing that we give to portfolio companies. And so that was kind of the initial bit. What does that mean, give the growth team to startups? Can you give an example? Yeah, so we would sit with companies, portfolio companies, and dig through their data and help them, you know, sort of build out that analytical motion of developing that internal tension, right, between their desire to build a beautiful product and sort of balance it with something that's sort of analytically backed.

Starting point is 00:10:02 We tried to help them create that tension, help them build that mindset. And, you know, without overriding, obviously, their, you know, pre-existing product mindset. And so we ended up working with a bunch of portfolio companies in that regard, sort of helping them build out their data capabilities, build out their analytical capabilities. It very quickly naturally became, oh, well, if we're going to do this data work on portfolio companies, we can actually do some of this stuff before even investing in them. Right. And so it turned into this broader question of, well, how do we use data as a firm? You know, we can use it. We can go into portfolio companies, help them with their data, but we can also use it in the investment process. And it wasn't just us. You know, you got chatted with

Starting point is 00:10:36 Abe, you know, before. There are several firms that sort of in the mid-2010s were playing around with using these data techniques to do various, to help with various aspects of venture investing. We happen to start from the lens of portfolio company, you know, sort of value add, but it's very clear that you can do many other things, right? Sourcing and evaluating and such. You know, also with my partners here at Tribe, as we spun out, is that when you work with people who have been venture investors for a really long time, like five, 10, 20 years, all of their regrets, their regrets that say, oh, I saw this thing, but I didn't do the right, I didn't make the right move, right? Like

Starting point is 00:11:11 it's not, they don't, they don't usually say, I wish I had seen that thing, right? Because people who've been in venture for a long time, they see enough good stuff, but they make the wrong call, right? And so from our point of view, using data, the question was, okay, well, we could use data to generate more sourcing, but that's actually not the structure of your regret. The structure of your regret is more about the decision-making on the things you saw. So speaking about those regrets, why do VCs pass over good opportunities? You know, what are the most common biases and what has the data showed you are biases that could be corrected? They're sort of infinitude of biases, right?

Starting point is 00:11:47 People pass for all sorts of reasons. That's the nature of VC, right? Is that you pass on like 98% of the things you see. And so, you know, maybe the question isn't so much like, what are the reasons you pass just because your default is passing? You know, one of the things that we really came to see is that, you know, in particular in the cases

Starting point is 00:12:03 that we were looking at, the passes were usually situations where they didn't take the time to really look at the data to make sure to see what the actual pattern of growth was, right? Because in almost all the cases of regret that we were dealing with in social capital, there were cases where when they saw the company was already working, it was like the series A and the series B, something was happening. Then we had the shot and for whatever reason, they didn't take it. They were like, oh, they didn't, you know, they thought the person who sent it over to them was an idiot. The referring VC, they're like, oh, that guy's a moron. I don't want to look at his deal. Right? Like dumb shit. Right? And so like, we're like, found this founder.

Starting point is 00:12:36 I did, he did it. I saw him at his last company. That company was crap. There's no way he's going to build a good company. So they don't even look. And so the question is, when do you do the work? Right? And that's usually the structure of the regret. The regret is that I didn't even do the work. It's not like I did the work and I pass and I regret passing. It's like, I didn't even do the work. Right. And so the question was like, how do we do the work and make sure we do that really well and do it at a high volume? You find a bias that's not actually at the beginning and not the sourcing and not in the diligence phase, but somewhat in a first look phase, first or second look phase,

Starting point is 00:13:11 where the mistakes being made, the false negatives are on which companies to actually double down on. Well, to spend time on. Because remember, like, the thing is an investor can only really look, like really look at like 50 companies a year. You just can't, like you can't dig deep.

Starting point is 00:13:27 You know, you can't call the customers, do all the work. You can't do that like hundreds of times a year. You can only do it a few dozen times a year, realistically, right? And so the question is, how do you open that aperture, right? And that's really what we ended up doing, right? Like what we ended up sort of prototyping at Social Capital

Starting point is 00:13:43 and really have built to its full fruition here at Tribe is basically the ability to dig deep into hundreds of companies a year, not dozens. So that we can get, you know, what we do at Tribe now, you know, we're able to get like extremely deep on hundreds of companies a year to make sure we're not missing anything, right? So basically tell us, okay, should we pay attention or not? Because the reality is we still limit, you know, you can't call the customers and do the, you know, founder reference checks. You can only do that a few dozen times a year. I think this is an important lesson for everybody to understand whether they're a startup raising capital or a venture fundraising from LPs. If somebody's in their data room, if somebody's

Starting point is 00:14:21 doing diligence, if somebody's checking references, they're interested. It might sound obvious, but when you look at it from their perspective, there's a finite amount of companies and funds they could be looking at. So let's move on to Tribe. So you met Arjun and Ted at Social Capital. What is it about Arjun and Ted that made you want to partner with them? You know, at Social Capital, we had sort of developed the early versions of this process of using data to like quickly understand and diagnose the pattern of growth and be able to give that back to founders, give them a really good experience. And we had sort of a glimmer of what it would be to build a firm around it. Hey, we'll continue our interview in a moment after a word from our sponsors. The Limited Partner Podcast is proudly sponsored by

Starting point is 00:15:03 AngelList. If you're a founder or investor, you'll know AngelList builds software that powers the startup economy. AngelList has recently rolled out a suite of new software products for venture capital and private equity that are truly game-changing. They digitize and automate

Starting point is 00:15:16 all the manual processes that you struggle with in traditional fundraising and operating workflows, while providing real-time insights for funds at any stage, connecting seamlessly with any back office provider. If you're in private markets, you'll love AngelList's new

Starting point is 00:15:29 suite of software products. And for private companies, thousands of startups from $4 million to $4 billion in valuation have switched to AngelList for cap table management. It's a modern, intelligent equity management platform that offers equity issuance, employee stock management, 409A valuations, and more. I've been a happy investor in AngelList for many years, and I'm so excited to have them as a presenting sponsor. So if you're ready to level up your startup or fund with AngelList, visit www.angellist.com slash TLP. That's AngelList slash TLP to get started. Back to the show. It was clear to us that if we just parted ways and went and joined other firms,

Starting point is 00:16:08 you wouldn't be able to build this kind of thing in an existing venture firm, right? Existing venture firms, they all have sort of an existing process and existing motion. They're not really going to change it, right? They just don't. And so the only way to really make it live would be to build a firm around it. And so, you know, the three of us were all sort of aligned, that's what we wanted to do. And we spent many years doing it together already. And so we partnered up to go for it in 2018. So you started a new type of firm in 2018 focused on first principles

Starting point is 00:16:34 investing. Tell me about what is that first principles and how do you look at companies at Tribe? The focus at Tribe is really all around being experts at understanding, measuring and amplifying that pattern of growth and product market fit. That's really what the focus is. You know, we meet thousands of companies a year, but we end up doing like really intensive data work on about 400 of them a year, where we literally, they will literally send us their data, their first party data, that analytics data that I kind of referred to at the beginning, right? They will send us that data and we will generate, you know, 50 to a hundred pages of really hardcore benchmark analytics for them, right? And we do this 400 times a year. As you can imagine, we're passing on the vast bulk of these, but even when we pass, you know, the whole point is when we pass, we say, hey, we're passing,

Starting point is 00:17:16 but here's why, you know, this number here is median. This number here is kind of bottom quintile. Give them really detailed feedback. And the reality is the founders don't know that stuff. They may have a sense that their burn is high, but they don't realize it's bottom decile high, right? They might think their growth is good, but in their mind, they might think it's like top decile good, but no, it's like just above median, right? So to be able to give them this really detailed information. And so founders, you know, what they say is that I'm bummed you're passing, but thanks so much for this. You did all this work for me, even though you're not investing. Can I, you know, can I, can I come back for my next round? Can I, you know, can I, can I share it with my, with my, with my investors, with my board and so on and so forth.

Starting point is 00:17:54 And the point, the point is really that we're giving them something valuable. We're helping them, even though we're not investing. We generate about 20,000 pages of analysis per year, but like 80% of that is just straight pass. Obviously collecting data the whole time. 20,000 pages of analysis per year, but like 80% of that is just straight pass, obviously collecting data the whole time. But then in the top quintile, when the data looks pretty good, that's kind of the signal that we should go dig in and do the work. We're calling the references, calling the customers and so on and so forth. But it's really all about like generating that value upfront so that we make sure we're looking at the right companies. And then if we do the diligence,

Starting point is 00:18:21 if we pass and we mess up, at least like we saw, we knew what we were messing up versus like, versus not even doing any work at all. You provide a lot of data, another form of feedback to the entrepreneurs. Do you find a correlation between the entrepreneurs that take the data to heart and those that are successful ultimately? Not very much, but not for the reason you think. It's just because there are many founders who are very good founders who do a good job of taking the feedback. But the reality is that if their company doesn't have a good pattern of product market fit, it's not going to be successful. Right. And that's that's kind of the underlying thesis of the firm. Right. Like the underlying thesis of the firm is that what really matters in the long run is that pattern of growth and that pattern of product market fit. It sounds

Starting point is 00:19:00 kind of obvious, but it's actually like a bit contrarian. Right. Because the vast bulk of investors are like the founder matters. The of investors are like, the founder matters. The team matters. We think the founder matters. We think the team matters. But in the long run, remember my story about Facebook, in the long run, what matters is that pattern of growth, not the founder. Yeah, I think you have two different camps in Silicon Valley.

Starting point is 00:19:18 I think the reason the narrative around the founder is more pervasive is because it's a form of marketing for the VC firms themselves. Yes. Back to my Christianity example is the question of Christ himself versus the machinery of the church, the early church and their evangelical behavior. So going back to Tribe and the way that you're able to assess the 400 startups per year. So tell me a little bit about, Is your data moat essentially recursive in that you're using all the data from the startups to create these benchmarks and to create these medians? Is that the right way to look at it? Yeah, that's right. I mean, at this

Starting point is 00:19:55 point, we have like 1500 companies in our database. And when I say companies, I mean like utterly raw first party data. You cannot buy this data. You can't go, it's not like the kind of thing, like another firm can't just be like, ah, I got set up. I have this, you can't buy this stuff, right? Like the companies have to give it to you, trust you with it, one company at a time. The only reason why they trust us with it is because we give them back something so valuable that it's worth their time to come back. That's the fundamental piece. The goal for us, right? The underlying driver is for, we want to understand the nature of growth and product market fit. It's bizarre that some things grow and some things don't. It's like a weird thing. And that's fundamentally what we're studying. And obviously we're trying to invest in companies that

Starting point is 00:20:33 have the most exceptional patterns. It's not solvable, right? Like why do some things grow and some things don't? You know, if you study this in macroeconomics, you study this from the microeconomics point of view, you study phenomena like religions. It's not really clear. Like there's many reasons why something's growing, something's don't. And like, that's fundamentally what we're interested in. And in terms of this data set, one argument might be that the data becomes,

Starting point is 00:20:55 has a pretty quick half-life given the disruption of industries, the disruption of platforms. But the reality is this, like if you're top quintile in this stuff, you're basically going to be top quintile. Maybe things move around on the edge a little. Maybe, you know, maybe the thing that was median is now like 58th percentile, whatever. But we're not looking for the 58th percentile. We want to make sure you're loosely top quintile. So in which case

Starting point is 00:21:18 we start paying attention, right? So over time, things shift around a little bit for sure. But we're not trying to be ultra precise, right? We're trying to because there's no point in being ultra precise. Right. We're trying to pick a company who's going to live over 10, 20 years. There's no way you're going to know. You're not underwriting to academic research or 95 percent confidence interval. You're underwriting to to return the fund. Well, yeah, that's right. I think the way we think of it is the data work is there to help us really ensure, okay, is this something that's worth spending our time on? Remember going back to the other thing that we discussed earlier, the real scarce resource is, you know, your ability

Starting point is 00:21:53 to do all of that traditional venture diligence work. That's pretty scarce. You just can't spend, you only have so much finite time to do that. And so the data work is there to help us figure out where should we direct that resource. And so the whole point of doing the data work is there to help us figure out where should we direct that resource. And so the whole point of doing the data work is so that we can do as much as possible through the data, make sure that we can direct the human diligence piece to the places where we think it's going to be best used. And then most of the data tells us where not to invest and then tell us where to pay attention. And then when we do invest, it's a combination of the data and all the traditional diligence work folding together. So another way to look at that is that the VC is essentially making investment at the time of diligence, not at the time of capital deployment. That is one way of looking at it, right?

Starting point is 00:22:38 Because that is like, you know, it's a key scarce resource internally. To be clear, like our 50 page analytics, it would be actually basically impossible for like whatever a traditional VC principle to do something equivalent. Like they just don't have access to the data. They don't have the analytical skill. It would take them a year or more to generate the level of analytics that we generate. And we generate that level of analytics in like a few hours. So a lot of our listeners, we like to brag, we have over a trillion dollars in assets listening to us. And a lot of our listeners, we like to brag, we have over a trillion dollars in assets listening to us. And a lot of them come from other assets and other asset classes that have much more efficiency than I would argue venture capital has.

Starting point is 00:23:14 How efficient is the venture capital market? And specifically, looking at the data that you have versus the top quartile investments and the investors that follow on and what you perceive to be the best companies? How efficient is that? Well, it's an interesting question. I actually, you know, I would say that I don't think venture capital as an asset class is particularly efficient. And the reality is you just don't know what's going to work at the beginning, right? Like you kind of have a loose sense. But I mean, the whole history of venture is stuff that completely surprises us. I promise I wouldn't delve too much into your secret sauce, but our listeners have been listening now for a bit.

Starting point is 00:23:47 So can you open up the Komodo a little bit and tell us a little bit about what the juicy data says about what's important for startups? Yeah, in fact, it's all on our website. It's not secret at all, right? If you go Google Tribe Capital or you look at our website in the essays section, there's an entire article

Starting point is 00:24:03 about how we think about product market fit, quantitative approach to product market fit. And there's a second article that's entirely our view on unit economics, right? And between those two articles, it basically outlines the entire thing. It's, you know, the whole point is to create a company, right? Where you can find some pattern of growth, some pattern of product market fit that is economically reasonable, which is to say has unit economics that makes sense and such that that pattern will continue to hold as you scale the thing up by orders of magnitude, right? You look for a company that expresses some unit economics, that expresses some growth pattern. Then the question becomes, okay, if I plug 10, $20 million into this thing and they go spend that money on more engineers, on more sales and

Starting point is 00:24:45 marketing people, is the pattern that I see, is it going to stay the same? Is it going to get worse? Is it going to get better? Right? And at some level, we obviously want to see ones that are going to stay the same or get better, right? Over time. Realistically, they almost all get worse, right? They get worse because your early customers are early adopters, right? So your acquisition costs are lower. They're going to be more fervent users of your product. So their lifetime values are going to be higher. So your unit economics are only going to get worse from the beginning. So when you see a company that already has like weak economic structure at the beginning, you're unlikely going to be in a situation where it's going to magically improve.

Starting point is 00:25:19 All the data that we do is all oriented around an implementation of this thesis around growth and product market fit. It's all out there. So let's talk about a topic no one is talking about today, artificial intelligence and AI. How does that affect your strategy? And do you see Tribe ever going 90, 95% AI driven? Yes. To be clear, we use it, right?

Starting point is 00:25:42 I mean, like a lot of our technology, you know, I told you that, you know, we have this giant technology stack that we've built. We're generating tens of thousands of pages of analysis per year. More and more of that right now, we're actually utilizing AI to actually help us write those memos. I mean, you're basically, there's no way to generate that volume of analysis with a small team, right? If you don't use a lot of machines to help you. And now the machines include, you know, some of these AI components. You know, we definitely believe it's an important aspect of it. Do I think that AI is going to do the full investing? Well, I mean, the thing is, remember, the decision making is only one small piece of the investing, right? Most of investing is like, you know, meeting founders, giving them a good experience, meeting yet more founders. And then like, yes, there's a moment of

Starting point is 00:26:20 decision making, but then there's all this stuff that happens after, right? You have to sit on the board. You have to like be a good partner to, but then there's all this stuff that happens after, right? You have to sit on the board, you have to like be a good partner to the founder. There's all this stuff that occurs that is a bit unrelated to that decision-making piece. You know, where AI could be really good at that piece, on the other side, it's less clear how it's going to help. Much less talking about the LP side, right? Like remember that a big part of what VCs do, right,

Starting point is 00:26:42 is fundamentally we're there to help implement, you know, views for our LPs, right? Make returns for our LPs via whatever mandate it is that they're giving us, right? And so in so much as asset management itself could be taken over by AI, perhaps we could. But since I don't see that happening anytime soon, I doubt that it's going to fully take over our work. So we're in the middle of a correction right now in 2023. Where do you see the future of VC? Are there going to be more firms like Tribe that come across or is there going to be amalgamation of AUM alongside the top, the Sequoias and the Andreessons of the world? It's been a fairly good past half decade or past decade to start VC firms. There are a lot of firms

Starting point is 00:27:24 that came up over this last decade that are sort of in the early part of their life. It's kind of incumbent on us to go find capital and to go develop our mandate and make good investments, right? Like I said, I think the underlying driver is really the growing appetite of venture capital as an asset class amongst allocators. And while it's true that they all may want to be in benchmark or whatever, it's just not going to make sense. For the largest funds that, you know, for the very big mega funds that believe that they can provide, you know, this product to that whole variety of LPs, I think they can. But the thing is that from those LPs point of view, they will find over time that they want some more specific mandate to be executed. And it's unlikely the case that the mega fund will

Starting point is 00:28:02 be able to execute their specific mandate because they're going to have preferences. And it's unlikely the case that the mega fund will be able to execute their specific mandate because they're going to have preferences. And that's really what it is. It's about venture being able to provide venture as an overall asset class to provide LPs with various tools, various mechanisms. Yeah. And I think the nature of what is a return has changed in 2021. It was TVPI. In 2023, it's DPI.

Starting point is 00:28:23 I think that's interesting. I think another factor to add is, of course, diversification, stage, hold period, all these things. Well, Jonathan, it's been a pleasure. You came highly recommended from Abe. I could see why. Venture capital, ultimately, when it's seen as an asset class over the last 10, 20 years. I think venture capital will ultimately be seen as driven as a science, non-art. So it's a pleasure to meet with one of the top data scientists in the space. Thank you and hope to chat soon. Yeah, absolutely. Thank you. Thanks for listening to Limited Partner Podcast. If you like this conversation,

Starting point is 00:29:02 please like, subscribe and review on YouTube, Spotify or Apple. Thank you for your support.

Investing Billions - E5: Jonathan Hsu | Why Companies Fail Without PMF, Examining Tribe Capital’s Proprietary Data Source, and What Experienced VCs All Regret

There aren't comments yet for this episode. Click on any sentence in the transcript to leave a comment.