Alex Mittal – Early Stage Investing - [Invest Like the Best, EP.119]

I know firsthand how complex the tech stack is for asset managers, and seemingly every new tool and data source makes the problem even worse, adding more complexity, more headcount, and more risk. Ridgeline offers a better way forward, one unified platform that automates away all that complexity across portfolio accounting, reconciliation, reporting, trading, compliance, and more, all at scale. Ridgeline is revolutionizing investment management, helping ambitious firms scale faster, operate smarter, and stay ahead of the curve. See what Ridgeline can unlock for your firm. Schedule a demo at ridgelineapps.com. Hello and welcome, everyone. I'm Patrick O'Shaughnessy, and this is Invest Like the Best. This show is an open-ended exploration of markets, ideas, methods, stories, and of strategies that will help you better invest both your time and your money. You can learn more and stay up to date at investorfieldguide.com. Patrick O'Shaughnessy is the CEO of O'Shaughnessy Asset Management. All opinions expressed by Patrick and podcast guests are solely their own opinions and do not reflect the opinion of O'Shaughnessy Asset Management. This podcast is for informational purposes only and should not be relied upon as a basis for investment decisions. Clients of O'Shaughnessy Asset Management may maintain positions in the securities discussed in this podcast. My guest this week is Alex Mittal, co-founder of The Funders Club. Following past guest Jeremiah Lohan, Alex is my second elementary school friend to appear on the podcast, A Trend I Hope Continues. Funders Club is a unique venture firm because it builds around a network of investors and entrepreneurs who submit deals for consideration and invest together. But as you'll hear, Alex and his co-founder Boris aren't just building an open platform for early stage investing. They also then take a very traditional venture approach, making investment decisions themselves when it comes to building a centralized portfolio. Our conversation is about what Alex has learned investing in almost 300 early stage companies over the past seven years. Please enjoy. So Alex, you're the second and we could call this like a sub series of the podcast, which is like my childhood friends that get to appear on the podcast after Jeremiah. It's an honor. It would be fun to begin since probably...

A lot of people won't be familiar with Funders Club, with the idea itself, and maybe the kind of origin story around the inception of the idea. And then there's a million little categories that we're going to go into. There's kind of two sides to the story of the origins of Funders Club. So one is the personal one of Boris Silver, who's my co-founder and investment partner, and my own individual entrepreneur story as well. And the other is the story of venture capital. So in the quickest nutshell that I could possibly do on both, one is Boris and I, previous to FC, we're entrepreneurs. I still think of us in the present tense as entrepreneurs, of course, as well, but had started our own technology companies and from the lens of a founder thought that there could be potentially a different and we hope to better approach to venture capital. And that's sort of where it started from. And then as far as the lens of just VC as a discipline and a field, if you look across the history of venture, it's maybe a generation old, 80 something years old. especially at the time in 2011-12 timeframe when Boris and I got together and started thinking about, okay, what would a VC firm that fully embraced today's technology and landscape, what would that look like? What we came away from was recognizing there's a lot to learn from the past successful VC firms. In fact, they wouldn't still be here and have ridden the ups and downs of multiple economic cycles if they weren't doing something right. On the other hand, Somewhat shockingly, the people providing capital to the innovators that disrupted other industries, you know, like Amazon disrupted e-commerce, et cetera. They themselves hadn't necessarily fully embraced software and technology at that time. What's really cool is over the last five years, you're starting to see pretty much every venture firm incorporate software, data, et cetera, in some format or the other. But that was our origin. Almost more of a community and co-op powered model where you leveraged a community of thousands of people around the world for deal flow, for sourcing, for vetting on the tail end, the value add piece, and still had the discipline of a traditional venture firm where returns matter and due diligence matters. And that was the spark and that was six and a half years ago. Can you describe the basic setup? So I'll try to use a spectrum.

And then you can say where on the spectrum you might be. So one would be just a small partnership. They do everything themselves. There's no network involved. Another end of the spectrum might be something like AngelList, where you can sort of open up and crowdfund or something like this. My sense is that you're somewhere in between those two. So maybe describe literally how this process works. Sure. Yeah. I think the advent, like as I was explaining over the last five years, the advent of the adoption of technology has, you know. come to light. And so AngelList is a great example of maybe more of the shovels model, where they're providing infrastructure that almost anybody can use. I think there's over 100,000 companies listed on AngelList. I believe there's tens of thousands of investors on there. And it's really utility versus, say, the other end of the spectrum you're describing, where it's a partnership. In the case you were giving, I think it's literally, as they say, Five GPs. Yeah, two investors and a dog, that kind of a model. And so what's fascinating is if you look at the most successful firms, they actually much more closely resemble the latter than the former. There's a potential future where the future firms adopt those shovels and those tools and anything that'll happen and is happening. But at its core, venture does require some... Not some, actually. Venture requires an intense level of discipline, an intense level of due diligence. And if it's just sort of left, we didn't seek to create something that kind of executed the venture by itself. One of the differentiators of our model is while it is a network model, there is a centralized standard for our standard bar. for what qualifies for investment closer to the partnership model then. So describe the role of the network in funders clubs setup and success. So put some numbers around this. What are the type of people that are involved? How are they involved? How much historically is it like a nice sounding idea versus something that's actually produced the things that have gone on to do really well? Just some stats around the network would be fascinating. Sure.

So since we started investments in Q4 of 2012, we've backed about 300 startups. We almost always enter at the seed stage, although we'll follow on and follow on financings as late as so far the Series D round, or actually the Series E round. And the network is over 20,000. accredited investors on the investor side and several hundred entrepreneurs. So it's like founders and co-founders together. And the startups are predominantly, we're based in Silicon Valley here in SF and portfolio is concentrated here in the Bay Area. But the long tail of the portfolio stretches all around the world. We have companies headquartered in over 20 countries and similarly for our investors as well. And then as far as just where we've seen the numbers play out on the model side, we're a bit more transparent on performance than I would say most other VC firms. We've always felt we have to hold ourselves to a higher standard because of our online presence. And so if we publicly publish our returns, we update them every quarter. There's no... rule requiring us to do that. We just do that. And so those are publicly available and we're proud of our returns, although that is reflective of work done in the past. And so there's an old saying, you're only as good as your next deal, basically, or your next partnership. And we believe in that very much. And so while it seems like the model is working, we seek to hold on to that original sort of questioning that we had way back in 2012, where it was like, what if this was the case. Would it be fair to characterize this as trying to harness the power of networks? So you mentioned 20,000 people. That's a lot of people, right? Credit the investors in this case. Harness the power of networks, but maintain a pretty strict rigor. in a centralized way around what investments get made? Is that a clear characterization? Absolutely. I think we live in an era where, of course, as we were early in on the blockchain, and so obviously believe in the merits of decentralization, but there are plenty of good things about having centralization in certain circumstances. And this is one, and maybe we can get into this later, but just...

Because of what we do, which is looking into the ether of what's coming up down the pipeline when it's really unknown, really a lot of risk, and attempting to predict what will be the next multi-billion dollar generation defining disruptive technology companies. It's really difficult to do with a consensus, decentralized approach. And we actually did spend a lot of time scratching our heads and chasing down, can you have some sort of crowd-based venture model that's good at picking outliers? And I think we've come up with something that gets at that in the ways that's possible, but it's certainly not from consensus. The needle that we're threading with our model is how do you leverage the value of an immensely scaled network, but also systematize and implement what has worked well for the highest returns in venture capital, which is that outlier effect. Basically, all of the returns of the venture industry come from a handful of companies, and it seems like that's the gravity. of the industry. That's the laws of venture capital, this outlier effect. I'd love to get into some of the key lessons that you've learned and really peel away at this outlier effect at systematized ways of positioning yourselves in companies that could become the whatever billion, $10 billion plus companies. But I'd love to begin with data. So you mentioned 290 or 300 companies that you've invested in. God knows how many you've seen. I always love people that just get to look at a ton of businesses. Is there any just pure quantitative data or rules that you've derived from that process since 2012 that you think people would find interesting? Yeah, so this opens up a whole avenue of conversation. We probably could do the whole podcast just on this topic. See how far we go. Well, yeah. Perhaps we'll do like a little preview. So basically, like obviously with any application, your ability to leverage data for useful insights is dependent on quality of your data and quantity of your data and the applicability of your data, like all this. And so one of the challenges for venture is, especially early stage venture, is one could very credibly say, OK, if your whole job is to identify sort of disruptive outlier, like no one thinks this is possible companies, how could you possibly have data that's relevant? Right. And so that's a very strong argument. And actually, that's the consensus argument. What we realized is that, OK, that's probably true. Like, actually, I believe that also.

But what if it's simultaneously true that there are certain business models in venture, in tech startups, that have been used over and over and over again where you actually could potentially have good data? And so we thought about that. And for instance, in enterprise, like... companies selling to businesses. There's the SaaS business models, subscription as a service, software as a service. And so in the consumer side, you have consumer marketplaces. So think of Airbnb or Uber. And so what we realized is, as you mentioned, we've funded hundreds, literally hundreds of companies. We've seen literally thousands of companies. And again and again, certain business models are repeating. So like breaking apart, say like SaaS, what are example metrics that or parameters that a SaaS company might have. These would be things like ACV, annual contract value, average annual contract value, or retention. or churn. Do the companies stay with the firm long-term or do they leave? Later on, sales efficiency, how efficient are the sales teams and so forth? Although you don't have that data at the seed stage, by the way. I was going to say, both those things sound like Series B type evaluations. Yes and no, right? We'll get into that. And then on the consumer marketplace side, metrics like what's the percent of organic acquisition versus paid acquisition, LTV, lifetime value, customer acquisition cost, CAC, cohort retention. There's a discipline, almost a science, around some of these metrics that you truly can translate. And we've seen this done extremely well. And so once you realize that, it's like, OK, cool. Now we know some of the parameters that are applicable. And we know that some of the business models where we can apply this data to make informed venture investments. But the challenge is that, OK, in the public market, you can download all this data. You can have gigabytes of data. You can pay people for the data in the private market. And in particular, at this stage. There's no such data set. Companies are notoriously protective of their KPIs, their metrics. And so it's how do you actually pull this off then? Basically, you need a lot of data, but you also need trust, right? So you need scale and you need trust. And I think there's a lot of VC firms that have one or the other.

even not just classic VC firms, just institutions that are in the business of working with startups that have immense scale, but they may not have that level of trust that's required in order to have the quality of data that's useful. And then there's on the other side, there are very thoughtful, very, very selective investors who have immense levels of trust, like they're on the board of these companies and so forth, but they don't necessarily have scale where data analysis could actually. where you could take data points and draw a line, right? Or derive a formula. And so you have to have both. And that's something that we both scale and trust. And that's something that we invested in from the beginning. And actually, it's not something we had on day zero. So this was kind of like a ambition where we looked. into the future and said, OK, what if we were at scale? What would that look like? And so one of the cool things that came out of this is obviously for us, that's great. But being able to flip everything I just said to the entrepreneur. So I think there's a value to investors if they can, if they're in a position to to being quantitative about whatever data they do have, assuming the data is good, which is a huge assumption and applying that to their portfolio as as a tool, like not not in a dictatorial way, but saying, hey, did you know that, you know, the top decile of SaaS companies, this is what their net revenue retention looks like when you're in. This is what yours looks like when you're in. Or did you know that for small businesses, the retention or the churn rate of X percent is this on average versus with ACVs that are higher for true enterprise customers, it's this other number. And so that can be immensely valuable to even the most talented founders because they don't have all those other data points, right? So that's one way to flip it. The other way is like, Even there's other places where it's even easier to apply data or it feels easier based on our experience where things like deal terms, we actually have a model where we just plug in certain parameters and we can tell them what their valuation is going to be on the term sheet that they're going to get. before they've gotten it. And it's been remarkably accurate for these discrete business models that we are able to collect and have a repeatable model around. So just so I understand that, so that might be something like you check a box that they're an enterprise SaaS business and the retention is this, their average order volume is this, and then that basically spits out, like, here's roughly where you should be. Yeah. And it sounds so anti-romantic compared to the sort of the myth of the...

venture capitals riding through, you know, Silicon Valley and looking the founder in the eye and saying, I knew I could see the resolve of the entrepreneur then and there as opposed to, but like the thing that I, what I'm saying is, is actually not that this is the opposite of that because I, the very, yeah, the very immediate next thing I was going to point out is unfortunately. at least right now, not everything can be distilled to that. And so like a good example is moats. If we could quantify moats, that'd be awesome. We're working on it. It'd be a pricing problem. The whole idea is finding like mispriced moats, basically. Yeah, yeah. You could frame all of investing somewhat like that. Yeah. And clearly we know what moats are. And so you can kind of like create a semantic language and a way of organizing companies and putting them in a tier of, okay, strong moats versus weak moats versus... not actually a real moat, but people think it is kind of a moat. And so it is both. Let's talk about the very early stage stuff now. So I want to weave this data idea maybe throughout the conversation. But when we first chatted about this idea, I don't know, six, seven months ago, you mentioned this idea of a very particular and important Venn diagram. So maybe you could describe what that Venn diagram is. Sure. Yeah. So if you can, in your head, imagine two circles, like a Venn diagram circle, one being ideas that seem like bad ideas. And the other circle being ideas that are good ideas. Then that little sliver in the middle is ideas that seem like bad ideas, but that are actually good. And, you know, this is a concept, a construct that I've heard, you know, early on, when we were, when Boris and I were just dreaming up this model, we talked with like some of these venerable folks in venture, but who had. kind of disrupted ventures. So they were like a little bit of outsiders themselves. Very like an Andy Ratcliffe type idea. Yeah, or like Paul Graham, Peter Thiel, that kind of mentality. And so we consulted with PG and actually Josh at First Round Capital, Josh Koppelman and many others. And so what we realized is that that sort of seems weird because, you know, we live in a society where, like, if you think about the top 40.

You know, for music or on Instagram, you know, in your feed, you get whoever's popular or even like take like machine learning algorithms. Like by definition, they're basically it's a consensus driven type of a model, a type of an algorithm. And so we live in a world where almost everything is telling us that actually the good thing is whatever what seems like the good thing. And actually what this Venn diagram is saying in a very contrarian way is the very best things do not at all look. like the very best things in the very early days. The most non-obvious thing, I think some people, some of your listeners have probably kind of heard this concept before to some level. It took me like six and a half, I've been on the entrepreneur side and the investor side for over a decade. I've been investing for six and a half years now. And it took that long to realize like one of the more non-obvious parts of this Venn diagram is that the level to which the intersection grows or to the level to which basically as the company progresses and gets older, the level to which it becomes obvious isn't that much. Put another way, like when Facebook had a market cap of $2 billion, like if you go look at the press from that time, most people did not think it was worth that much. And they're actually like reporters. I remember an article that it was like, dear Mark, please accept the acquisition offer from Yahoo for $2 billion. Like, please, please, you're too young to know what you have and this kind of thing. And so it's so misjudged, basically, even as time goes on. And just to put it in like sort of other terms of like how easy it is to misjudge, true outlier. businesses and ideas if you go back to i guess by our silicon valley pace standards of time the old days of like google when it was first starting this is like larry and sergey as researchers not a lot of people know this but like they actually shopped the technology around to excite and Lycos and search engines that I'm just barely old enough to remember using. And they actually reached a deal to sell it for $750,000. The numbers are all over the place. I think the Google founders remember it as $1.6 million. And it doesn't matter whether it was $750K or $1.6 million. But the reason I think this is powerful is

It's the innovators themselves even that misjudged the future value of that technology. And to their credit, I should say one of the reasons, other than price, that I think a deal didn't happen is it feels like maybe the technology itself was misjudged even by the buyer. So even at that price point, it was like an acqui-hire. It was like, all right, cool. We got two smart people. We're just going to throw that algorithm in the closet. It doesn't actually work as well as ours or better, at least. And you see that even today in the blockchain space, Vitalik, creator of Ethereum, along with the community of developers behind Ethereum, early on in 2016, sold 25% of his stake of Ethereum at $10 per Ether. And clearly, he was very public about that. He said, as of very young, I think he might have still been in his teens, he needed diversification. There's no fault on him for that. But if he had just waited 18 months, it would have been 30 times. more of value. And I should point out I'm a strong bull on Ethereum. But my point in sharing this is actually that if the innovators themselves cannot exactly quantify the future value, how can investors even do that? It's a very, very non-obvious and challenging skill set. And one of the answers that we seek to teach ourselves in doing this, in mastering this, is basically deprogramming a lot of the heuristics that we've been taught by society. over our lifetimes. So let's get into some of those. Yeah. Let's hear as many as possible. Sure. Okay. So like I mentioned PG earlier and he had recently tweeted about if Darwin had published the origin of species today, he would probably, you know, he'd probably be fired from his job because that's how controversial it was at the time. And now, at least within scientific circles, like the idea of evolution is very accepted and it's actually what's behind a lot of art. advancements in science. And so that's behind that simple observation is just the idea that I was referencing earlier around like the top 40 hits. It's just we live in a society that your brain is just hardwired to assume that like consensus is good. And for many things it is. But if specifically for this idea of pushing the edge, the boundary of what's possible next.

It's bad. And so it's just reminding yourself that just because consensus says this is bad doesn't mean anything with respect to one's discipline of trying to invest in the future outliers. You kind of have to ignore that noise and sort of the supposed wisdom of experts. And it's so challenging because almost you could ask yourself, like, who are you or who am I to cast a vote? when there's when there's experts in the field who've dedicated their entire career to this thing so it's very abnormal to even engage with that so so that's like that's just sort of a mentality thing i think it's important to embrace that i want to dig in even a little bit more maybe start using some examples of companies and like your experience at the time for seeing them to illustrate some of the points around things that seem like bad ideas. So some common hallmarks maybe of why we think things are bad. You've mentioned some, obviously, non-consensus, things like this. We were at a dinner last time I was out here together. And I can't remember the guy's name, but some guy had been successful, I think, in like the biotech space or something, had delivered a big exit for his investors. And he was talking to us about he even said, like, you're going to think this is crazy. And it was like some crazy video game idea that he had he had come up with. And like, I saw your ears perk up because you're like, oh, that's good. Like, I like stuff that's like it needs to kind of sound crazy. So, you know, picking. Maybe some examples, obviously Coinbase, you were an early investor in Coinbase, and that's a huge obvious example. So it'd be fun to talk about that. But maybe even ones that haven't worked out, just to hone in on what it is that you look for and see at the early stage in this realm of seems like bad. I completely agree. We love things that seem crazy. I will say we love things that seem boring also, which almost sounds like the opposite. But in its own way, it can often end up being somewhat contrarian as well. And therein lies opportunity as well. But just to walk through some quick examples, and we'll start with Coinbase, but we'll go to some other less well-known examples as well. Coinbase is just such an epic one in the sense that in 2012, the price of Bitcoin was like $6.

Most people did not know what Bitcoin was, had never even heard of it. And if they had, they knew of like the Silk Road, which was at the time a black market that was associated with illicit drugs and weapons and this kind of thing. And so most people, when they did their due diligence on the market around, say like Coinbase, would find that. It seems like no one's talking about this. And also, it seems like the only thing I can find about it is that people are using it to buy drugs and weapons. And that's bad, right? That's not a good thing. But when you step back and think about, okay, well, there's US dollars. And in fact, I don't know this for a fact, but I wouldn't be surprised if the number one currency used for drugs and weapons is US dollars. Sure, yeah. It's the whole black market globally, right? Yeah. And so we're like, okay, well, from just... An objective viewpoint, that's irrelevant. The fact that the Silk Road exists is completely irrelevant to this decision of like, is this industry, is this technology potentially foundational for the future? And so it took that kind of thinking to put blinders on basically for the seems bad factors in order to understand and dig a little deeper at what people often call first principles analysis. I almost hesitate to reference first principles though, because I see it thrown around, especially in this town, as a precursor to an opinion. It's like from first principles, blah, blah, blah. And it's just like someone's opinion. So I think it's very, very difficult to truly be first principles. But if you can do it, you might just see the future, right? And so that was one piece. Can we pause there just because I have a question that might help me clarify and understand a bit better. So Coinbase obviously has been... a hugely successful business, largely because people use it like crazy. In the same way that Uber is amazing, because I've used it five times in the last two days, and it's indispensable to me. Coinbase is a company built around what to this point has been the primary interaction with cryptocurrencies, which is to buy and sell them. It's the gateway. It's the gateway. And so I totally understand how...

You could even at the inception, if it seemed crazy, you could foresee a future where if it did work, like people are going to use that platform to do stuff. So I want to take that same idea because you said earlier you're a large bull on Ethereum, the currency, not any company wrapped around it. And try to understand why that is the case, because the common screed against Ethereum today would be that, you know, most of it has been price speculation and that the actual use of it. So people use Coinbase like crazy, like no one's actually using Ethereum. to do anything. And it's actually like horribly suited just because of distributed compute to really accomplish basic functions. So talk to me about that bull case and like what I'm missing in that comparison. I think the phase of the industry is in its infancy still. And so I think what you're seeing when you look out of the landscape is tons of parallel experiments that are all playing out, some that are just atrocious and others that are interesting and others that may very well be like the future Google. And what I think is like this is an opinion, but what I think is one of the strongest, most interesting aspects of blockchain and other people have said this, I don't think this is a shocking insight, is the fact that with blockchain technology, you can actually have trustless transactions, trustless computations. I'm getting goosebumps when I'm saying that in the sense that that is a foundational technology that literally did not exist before blockchain. How will that be leveraged and used will be defined within our lifetime? Is it going to be like all these things that got a lot of buzz like over the last year? Who knows, right? Like it's too early to say. But that's actually specifically what personally I'm most excited about, about Ethereum is its ability to truly be a decentralized, trustless form of interaction, which has applications too. All sorts of things. And I could point to specific examples of maybe early experiments that look really interesting. Yeah, that would be helpful, yeah. And I think these projects have subsequently got some traction, and so it's somewhat becoming more mainstream. But like the MakerDAO project, right, sort of the idea of a decentralized stablecoin with DAI, DAI, their first major token to get adoption. And they'll be introducing multi-collateral DAI soon.

That's really interesting. That's like an entire financial institution, basically, that's completely self-running, decentralized, trustless. And will that still be here next year? No, I have no idea. Is it completely novel and is there any precedent for that? Yes, it's completely novel and there's no precedent for that. And I think that's amazing. So that's why I'm bullish on it. And then when you look at, I actually have a newspaper clipping from the Financial Times from like six months ago when the recent crypto winter hit, where it said something like, I'm going to ruin the title here, but it's something like, questions over the utility of Ethereum rise as price plummets. And I just love that title because it captures the... inability to decouple price from utility in the mainstream press's mind. And I don't fault any particular person. I think it's very easy to fall into that bias. But going back to the Coinbase example, one of the things that we really got as bullish is like no matter what the price was doing, their usage was growing. you know, week over week. That was phenomenal. It told us that this is a real thing that people get utility from, and it doesn't matter which way the price is going. And then when you look within the current space, we're in like a crypto winter. It's a pretty bad one. And yet projects that have been around forever, like Gollum, the decentralized compute network, usage has 5X'd in last year. despite the drop in price, you know, like 90% or whatever it has been. You see real adoption from blockchain from real companies, you know, like DocuSign, I believe, announced integration with the Ethereum blockchain for e-signing as a ledger for that. There's many other examples. I would say it's still experimental phase, but you're starting to see that. So that's really the uniqueness of the technology, the theoretical utility of it, and the early signs that it's actually being adopted by credible people. And the early use cases that people are actually using these things for gives me a lot of hope and confidence in that somehow, in some way, this is going to be big. Yeah, but I'll admit this is like super primordial stuff. It's something that obviously people that listen know I'm interested in.

usually skeptical of, but there's no denying that if it's real, like this price drop is the kind of hatred and fear, I guess, that you would want as an investor. Yeah. I've never publicly shared this before, but our peak of investing in the blockchain space, and we've never, by the way, we've never purchased tokens. We don't own any cryptocurrency. We've only ever invested in equity, like in companies like Coinbase, Chainalysis, and others. The peak was like in 2016 for us. We did zero blockchain investments. This year, we did zero new blockchain investments last year. We did one follow on. It was a co-investment with Benchmark and Chainalysis. And so there was a moment where there was just crazy amounts of speculation where it almost didn't matter if one was bullish about it. As an investor, we took a pause. And so I haven't publicly shared that before, but I do think it's an interesting insight. And it goes along with this almost. duality. Like earlier, I was talking about the intersection of ideas that seem bad that are actually good, that little sliver. It's sort of like, yes, I'm bullish on this technology, but I'm going to stand aside and not invest because the price is crazy. It takes almost like... believing, understanding like both sides and being okay with that weirdness. So let's go to other bad ideas. We'll leave crypto and Coinbase to the side, maybe some more esoteric ones, maybe the companies that people haven't heard of, or even ones that didn't work out ultimately. But back to this kind of the origin of the origin story is awesome. Like Instacart, we already, I already brushed on earlier, but the idea that, okay, really, you're going to take another go at Cosmo and Webman, which failed and raised a billion bucks and failed on demand delivery. Like that's crazy. failing to understand the nuance of like actually totally different business model, totally different asset structure, like lack of assets, actually pure software business and thus potential for much stronger business. And in fact, that's what played out. There's a great podcast actually with Apoorva on how I built this, where he talks about the story where a VC drops the business plan on a floppy disk for web band and be like, you don't know what you're doing. You got to read this thing. So there's nothing much more to be said there. It's obviously they're doing phenomenal and it's great to see play out that way.

There's a company called GitLab, which most engineers who are tuning in, and I hope there are plenty of engineers tuning in, would have heard of. And at the time, so this is a company that's kind of in the vein of GitHub for lack of just for being brief, but it's much more. At the time, they were espousing two things that were super like unsexy. One was open source, which at that time. So the things have totally changed now. The pendulum has completely shifted where suddenly that's attractive because I believe SAP acquired open source project for a very high premium. And now people are excited about open source all of a sudden. So anyways, one was open source. GitLab is an open source code repository. And so if you wanted to, you could fork GitLab. You could host it on your own servers, all this stuff. And by the way, I'm with you on this one. Open source seems like a bad business idea. Yeah, it's like, what? Wait, you could just give it out? And people can fork it. Someone could start another company around it. You did all the hard work. Yeah, it sounds absolutely bonkers. So that was one really crazy thing that the founder was espousing at the time. Another really crazy thing was this was a completely remote company. And by the way, today, this is a billion-dollar business that has hundreds of employees that's still completely remote. And so either or both of those things would be good reasons to just be like, sorry, you're a cool founder and all, and the market's interesting, but we don't believe that giving away your code to your competitors and customers makes sense. Because why was that the prevailing wisdom? Cloud SaaS, proprietary SaaS, that was what was in and what was cool. And just to show how difficult this is, this was like the one across 300 investments we've done. This was the one investment where we reversed our decision. We will almost never do this, but it was a case where we ran our process, which again, as you can tell from our discussion earlier, is pretty algorithmic. It's meant to be so. It's meant to almost... erase all emotion from, I say almost only because we can't possibly do all of it. I wish we could, but it's meant to really distill it to as much of a science as possible. And we had reached a no, but that was one where Boris couldn't sleep that night and got a phone call from him. We talked about it. We changed our decision. And the other thing about this company is even at the later rounds, so at the series A and B, we were able to actually increase our ownership at the B round.

which is almost unprecedented in that kind of a company where people are fighting for allocation. And part of it was just the relationship driven with the company, but part of it was actually just, yeah, like the best companies, even at that stage, sometimes are still not very obviously, like in the minds of some investors, going to be those later outliers. I would say that's becoming less true with the influx of... capital like into venture at later rounds especially at the growth stages but yeah it's it's fascinating so it's there there's there's a lot of um and i'll say like by the way like it doesn't always work like i gave three examples that all at the time were seed stage companies that we invested in in some cases it was a one-person company when we invested And that later became multi-billion dollar companies, right? So when it works, it's amazing. But the power law dictates that frequently, most of the time, in fact, it doesn't always end up that way. And, you know, I can recall, and these are actually some of the data points that teach ourselves. And also, I think the founders come away learning from those experiences. And it really helps the whole ecosystem develop. So one of the things I love is when founders are open. on what went wrong and if they understand that. And it helps us also do our own analysis and say, okay, like what went wrong? And it was a company that I was actually personally like a huge customer for. It was on-demand hot food delivery, right? And this was in the era when, right before on-demand got hot. And again, another pendulum swinging thing where first it was like nobody liked it. Then Uber came around, everyone. Totally threw their money at it. And then they're like, wait, this is hard. People ran away from it. And so people were learning lessons in real time about that. And one of the lessons that was learned is it's incredibly difficult to the point where you need a computer science degree and a mastery of algorithms and data to get the unit economics correct on on-demand delivery. That was one learning lesson that was learned the hard way. The other is just like...

Food is hard. Launching a restaurant, if you talk to any chef, like they'll probably agree with this or anyone who's tried to like invest in the restaurant business, being able to serve like on time, good service. good tasting food that's priced right, that in and of itself is hard. So couple that with like the on-demand thing, that's like really, really, really, really hard. And so that's an example of something where it's like consumer, you know, like NPS off the charts, right? Like people love that. But the nuances of it, it took some experimentation live in the market to really figure out that, yeah, you know, despite mobile and all this other stuff, it's really freaking hard. Let's go back to this. I love this. crazy and boring as two words, as sort of signposts for something you might be interested in and apply it to today's... landscape. So we talked a bit about blockchain. You sent me some stuff ahead of time. Like one example that I would probably put in the boring category is like retail and logistics. So talk about your thoughts on that. And again, through the lens of your Venn diagram. When we first came to the market as investors, we had been for almost a decade as entrepreneurs. But when we came there, it was almost on the tail end of this e-commerce pendulum swing to the bull side. And I'll refrain from naming specific names, but long story short, there's a lot of money that went at these branded e-commerce companies that many went under. And so the prevailing wisdom was e-commerce plus tech is bad. And it's not a good place to put your money. Meanwhile, Amazon just was destroying everything in its path and growing and scaling. So clearly e-commerce plus tech is working. There's different ways to tell the story of e-commerce, but I'll use the Amazon lens, which is basically like, all right, it's working for them and it's kind of terrorizing everybody in their path. They need tools to keep up with that because basically Amazon at the time especially was one of the few entities that really was ignoring the noise and just doing their thing and winning because of it. was like a pressure, a forcing function for who cares what the VCs were thinking. It was a forcing function for everyone in the industry to like do something. And you could also say, OK, it's nothing to do with Amazon. It's just like modernization of tech and like the consumer population growing and more people buying things online. All that stuff is equally as possible as a forcing function. But what it led to is end to end everybody in the supply chain from how product leaves.

It's country of origin and reaches the shores of its country of consumption to how it then is stored there at the port to how that makes its way from the port to this store, from how it's actually bought and purchased in the store itself to if it's in an e-commerce setting, how it's actually sold to the consumer. There's a business at every one of those intersections. There's a multi-million dollar business opportunity for multiple businesses in each of those categories. Almost without exception, we've never anticipated. these trends before the data told us that they were happening. And like, so as a consequence, like we first invested in blockchain in 2012, and we basically stopped it when everybody else jumped in in 2017 and 2018. And similarly with e-commerce and logistics, this shift, I think that in 2000, like I said, it was actually seemed like a bad idea back then. But we started seeing some really weird things like, for instance, there's a company called Chippo, which was seeing some pretty eye-popping numbers on their growth, week over week, double digit. like growth. Basically, in short, they help e-commerce companies manage their shipping and returns, reverse logistics processes and what is going on here. And we kind of through that lens realized, you know, the platform of Shopify and the whole market for mid-market e-commerce players and the whole lack of infrastructure for them. And it was this very eye-opening process. And through that realized, wow, there's like a whole like going back to the how product. leaves the port and that whole thing with our investment in Flexport, which... Earlier, you were talking about you guys seem to have a bias or preference for crazy sounding things. And I was like, well, actually, we like boring things too. And Flexport, until later on, was really boring of a sounding company. But that's cool. And Ryan has acknowledged that as well. Their whole thing is trying to make logistics sexy and exciting. So they're literally just software that helps move stuff around in port? Data plus freight forwarding, logistics, air.

ocean and this goes all the way i mean like embark trucks for example where we were i believe their first or one of their first investors at the time like when it was sort of what was very popular maybe like five years ago was about like self-driving trucks right and sorry self-driving cars i should say but what's interesting about self-driving trucks is the bulk of the trip and the value of the trip happens on a pretty linear route like the freeways the highways on pretty predictable patterns. Like highway traffic is a lot easier to navigate than city traffic. In a strange way, there was all this focus on the hardest possible problem at the earliest stages of the technology when in fact the lower hanging fruit with like a very real tangible value there. And I mean, even societally, the trucking industry is one of the- It's huge. It's like the number one job in the US, right? Yeah, like it's one of the most dangerous jobs also is what I was going to say. The fatality rate is actually like really high. And so from a societal standpoint- There's something to be said for like, there are certain jobs that humans are currently executing that, you know, maybe they shouldn't be. And I'm not saying there's not a room for very skilled truck drivers to continue doing their thing. In fact, with Embark, it's enabling them to focus on that skill set. So once the truck reaches the city, a human driver takes over because technology is absolutely not there today to enable that, to drive an 18-wheeler around a city, right? So that's a human job currently. And even riffing off of that. idea of like boring sounding but actually like really great companies um gecko robotics is another one that comes to mind where people may not realize this but like the power that lights our lights and powers our gadgets and stuff that comes from power plants obviously and inside of every power plant that like say is like a gas or coal power plant there's boilers that once a year have to be shut down And people put up scaffolding and climb them and inspect the walls by hand. And this is both very expensive and very dangerous. Again, people die every year from doing that. And this company was like, hey, we're going to make little robots that climb the walls of these things so people don't have to set up multi-story scaffolding and risk their lives. And so we don't have to literally shut down the entire power plant for a week to do these inspections. And so these are boring-seeming.

but actually very exciting opportunities as well. Talk a little bit about another category that I found interesting was chip and electronic design. So with chip and electronic design, it's funny because I keep referencing pendulum shifting, but there's certainly been like this pendulum swinging that's happened in that field as well. And so my personal lens into this came actually as a consequence of being very close to some of the... cutting edge blockchain compute initiatives that have been happening. And so in the very, very early days, in other words, so Bitcoin is secured by proof of work, which basically means fancy computer computations to secure the network and run it. And so there's been this race basically for compute. And so it started with CPU, then it was just totally not profitable to compute with CPU. So then it became GPU. And then ASICs, which are like basically highly specialized hardware that's just designed to do one algorithm. And sort of somewhere either before that or ahead of that is FPGAs, which is like kind of like an ASIC, but sort of programmable. And I believe that there's a window into the future of compute because Bitcoin was like this petri dish of what happens when there's just this crazy amount of demand for the most efficient compute possible. And when you look at what's happening in, okay, where in business is compute demanded the most right now? It seems like it's in machine learning. It seems like it's in AI and these fields. And that's a very like, That that application of compute bears a lot of analogies to the application of compute for Bitcoin. And so what is happening is in the old days, we had mainframes. They did one thing. They ran one application. Then we got like the personal computer ran multiple applications. And so it kind of went general in the same way. You know, in the beginning, there's like CPUs and GPUs that are general. It's swinging back. It's swinging back to like specialized hardware. So that's one shift that we're seeing is the swing back to hardware that is designed to really just execute one function really, really well, really, really fast. There's obviously risks to that. Like if you invest in hardware and that algorithm goes out of fashion, it seems like you're kind of screwed, right? So can you create custom hardware designed to execute an algorithm and shift really quickly as the algorithms change? Is that even possible, right?

That's one area. The other is just because of the growing volumes of data and data sets getting really, really big, just being able to process and compute huge data sets really, really fast is another avenue that's opening in that space. You mentioned at the beginning material science, and I just finished a couple months ago this fantastic book, just basically a history of energy by Vaclav Schmiel. frames everything through what he calls fuel sources and prime movers. And that those are like the primary categories through which we, you know, go from human muscle to inanimate muscle and things like this. And that effectively what we have been doing for the most part since like the late 1800s with some exceptions has been eking ever more efficiency out of existing systems. And that's amazing, right? Like if you can be 90% efficient instead of 20% on coal, on extracting energy from coal, like that's fantastic. We haven't really, for the most part, found the next coal. And it sounds like a lot of what you've described, effectively better and faster, more well-specified compute power, better, faster, more efficient logistics, things like this, is that we had all these latent assets in the world and we're becoming more and more efficient at coordinating them and that there's huge businesses to be built around that. Is there any industry or company or people that you've funded that are in the other category? greater and greater efficiency through software, but through like foundational, like almost like super crazy sounding, like hard science type stuff. Absolutely. I hope that all made sense. Yeah. Yeah. I, you told me if this answer makes sense, it responds to it because, because we absolutely have backed truly like hard science out there companies. And so, you know, like one that comes to mind is actually, I mean, going directly off of material science was a company called OTI. that's applying quantum computing and just also non-quantum algorithms to materials discovery. And so as the founder of a material science startup myself, I know that like, this may be almost the opposite of what you're talking about, but bear with me for a second. The history of a lot of material science is more empirical based. It's like, all right, let's try this. Let's see what happened, observe, and then refine versus the way things typically work in like mathematics or physics or computer science, where it's like...

These are the laws of the world and like, let's compute this. And so obviously progress is orders of magnitude faster in that realm than in the theory realm than in the on the bench realm. And obviously theory has to be checked with reality, but at least you have the theory. And so there's basically like certain ways that people have known for a long time that one could apply that computation to material science, but we just like didn't have the compute power going back to like the need for computing and what's happening there. the algorithms and to execute them to do that. And so there are certain companies on the cutting edge now that are saying, wait, for the first time, we actually can do that. The reason I referenced them in particular is they're not just a pure software company. They're actually introducing innovative lighting and display technology to the world. They're partnered with Fortune 500 companies for delivering those to the world. And so they're applying. fundamentals of material science, applying the tools of machine learning and even quantum computing to then actually implementing in the hard sciences realm in their labs and their pilot lines in collaboration with industry and delivering working devices. So that's super exciting to see that end-to-end happening. In the life sciences realm, we're investors in companies that are similarly applying the tools of software, but in a hard sciences fashion. So there's a company of notable labs that is able to effectively test thousands of combinations of drugs, some approved for the illness, others not, to derive therapies both for industry and also for patients that people didn't realize were even possible at a scale that wasn't possible before. So a lot of our hard science investments, they're still implementing some of the tools that the last, I don't know, 10 years of innovation around machine learning and data has brought us and implementing it. So, you know, another really fascinating one, we're investors in a company called Neurable that, I mean, you should demo this at some point, but you have to try it to believe it because other, you know, there's many companies that have tried doing brain computer interfaces, but few have built ones where you try and you're like, wow, like this.

working is magic. Yeah. Yeah. And, and this is different from like wearables that go on your wrist or something as a separate sensor. This is meant to be invisible technology that is integrated into, you wouldn't know it's there basically. So like some, you know, some of this is like super stealth, but as far as what's publicly disclosable, it's a combination of let's call it proprietary technology and just like traditional, like the fact that there are signals that are discernible from outside of your head that your brain emits right and and that combination paired with a different type of algorithm than has traditionally been used is leading them to have 20-fold faster speed or like 20-fold lower latency in responding so so in other words your brain like if you were to think right now like a word or or if you were to color or something yeah like there's a latency there to pick that up and to like do something with that and there's plenty of like academic researchers that are showing these really cool but like really impractical implementations of stuff like this and so What's amazing is this thing actually works. You can play a little standard demo as you go around, and it's a first-person shooter game, and you can just fire a gun on command with your brain. Just thinking about it. Yeah, and pick up different tools. I'm just curious. You don't have to give away any of the technology, obviously, but let's take that use case, for example. The action is fire a gun. How does it train? I'm sure it doesn't just know. it can't read the signal fire. Yeah. So, so, and that's the other remarkable thing is I think that you would think that something like that would have to really get to know your specific brain, but you know, we're kind of all, you know, there's these old statistics about where like 90, 95% of our DNA is like overlaps with the gorilla or something like that. Like within people, there's even more overlap. And so, so actually like there, there's a lot of common. commonality between people. And then, but there is like a short, brief training thing, but it's very brief. That's the other impressive thing. It's like a minute to get acquainted with yourself and to be able to have that then refine itself over time, obviously is very powerful. And so, so we're, I would say that something like 70% of our portfolio is strictly pure software, maybe another like 10%.

is basically software, but it touches the real world or 15% or something. So it's only like 10% that's like this kind of hard science first thing. So we're very, very careful where we decide to partner there. But obviously, a lot of this sounds like almost impossible or like on the cusp of sounding like just sci-fi. But a lot of this, the future is already... Here, it's just not evenly distributed. That's very much true for these kinds of technologies that we're seeing. So we spent almost all this time talking about trends, ideas, and Venn diagrams. We really haven't talked about the people behind any of this stuff. And I'm just curious what, and in my, let's call it like 15 VC specific conversations I've had, you hear a lot of the same things over and over again when it comes to evaluating founders. So I am curious what you found is important or not important, but I'm especially curious if there are things that you think people care too much about or too little about. So the theme of the conversation is this, things that we over or under index on. So when it comes to the actual people themselves, the ride in the horse and staring them in the eye, any observations there from all these companies and founders? Yeah, well, I mean, like, first of all, as founders ourselves, we have a lot of empathy for the founder. And we've sought to be as thoughtful as possible about that. And our model keeps getting updated as we collect more data on. character traits, personalities, et cetera, about founder. However, what I'll start with by answering that question is almost like an anti-answer, which is that, and some other investors will agree or disagree with this, but in the spirit of trying to help my fellow entrepreneurs, what I will, especially first-time entrepreneurs, is what I will say is the market can make a team that maybe is lacking in many dimensions of, say, being the perfect founder or something. it can make that kind of a team incredibly successful because the market is on fire. And then the converse is true as well. You can have the perfect team battling a market. And so the force of the market is so strong that what you really want to do if you're an entrepreneur is be very thoughtful about what market you're attaching yourself to and your company to because that'll like...

lift your strengths and where you're weak and seek to compliment yourself, at least it'll be a little bit less hard. And so I'll start with that. But like going into, I think the number one trait that matters the most is relentlessness. And so a good example would be like Wonderschool in our portfolio. Like Chris Bennett actually started that business together with his co-founder, Arl, originally in a completely different field and industry. They were actually helping merchants sell through Instagram and refused to give up and refused to stop following where the heat was in the market and ended up in a completely different ancillary field of sort of Airbnb for preschools and daycare. And that was over the course of years. And you see that also like in Slack, I was mentioning Slack earlier. I mean, this is a company that started off as like a gaming company and then failed at that, realized they failed and then decided to like launch Slack, which is like not a game. Although, you know, they would say, yeah, there's plenty of small examples of where you can see, oh, yeah, that kind of makes sense with all the emojis and stuff. But the reason I cite those is not because I'm against, say, like a founder calling it quits, because I think there's Fred Wilson wrote a recent post about I thought was good about like knowing when to quit. But it's more to emphasize that like, hey, here are two examples of. true relentlessness like they could have easily given up and instead you know decided to keep going and trudging along no matter how hard it got to ultimate success and that's really really important and i think that might even be the most important character trait of what makes founders who are successful that successful pairing raw relentlessness with Amazing market. But there are also other character traits that I think are immensely important. So Brian Armstrong, who is the co-founder and CEO of Coinbase. As you get to know him, I've had the pleasure of being both a friend and also investor in the company. But he has like a next level ability to soak up like a sponge things that he reads and implement them. And also to ask questions of. everyone he meets in a very inquisitive way where it's not small talk. It's actually, you can tell that it is being incorporated, logged, put in the library for later pulling out. And what I would call that is ability to learn basically and to be curious. I've heard David Swenson at Yale describe that exact same way, which is kind of interesting, totally different field. Yeah. And so like, I don't think it's a coincidence that that founder has been so successful because he has a combination of all three of those traits that I just mentioned. So I didn't think like,

For sure, it's important. I also think for sure it's important to recognize that as strong as one could be, you could be perfect, basically, whatever that means, and still fail if the market's not there. The last question I ask everybody is for the kindest thing that anyone's ever done for you. Without taking absolutes, because you're putting me on the spot. People have done so many kind things for me. I'm basically indebted for life to those who've been close to me. I don't want to avoid acknowledging, accidentally acknowledge people who have been so kind. But one of the kindest things that's happened recently is one of our own entrepreneurs actually... took me aside, or it was actually over a text message. And we have a great relationship, right? And maybe that's what enabled them to do this. But they're like, you know, I think that in certain contexts and conversations, when you have a lot of, you know, pre-existing insights and you get like somewhat excited and I noticed that you're not. asking the question or pausing. And you could really, you should take a look at active listening. I was just like, as the investor, you're used to being the professor. You're used to being, you should do this or whatever. And obviously we try not to be like overly dogmatic and listen and try to learn as much as we can. But I think it's an incredibly kind thing for somebody who's in the position of being the entrepreneur to be like, hey, like. I care about you. You should invest in this thing. Yeah. And that was like, A, it was very eye-opening, but I think it was very kind. And I'll never forget it. And actually, that just only made me really respect that person that much more. Cool. Well, this has been as interesting as I'd hoped. So thanks for all your time. Thank you. If you're a book lover, you can also sign up for my book club at InvestorFieldGuide.com forward slash book club. After you sign up, you'll receive a full investor curriculum right away and then three to four suggestions of new books every month. You can also follow me on Twitter at Patrick underscore Oshag, O-S-H-A-G. If you enjoy the show, please leave a quick review for us on iTunes, which will help more people discover Invest Like the Best. Thanks so much for listening.

Thank you.