#58 - Miles Grimshaw

32:18It's writing code for you, testing. They now have bug bot on PRs. There's a lot to the development lifecycle to writing the world software that they can do and work on, many aspects there. And very strong ultimate business challenge, very sticky product, a product that can get better. The leader's product can get better where product depth will compound. And so you see a small team of five, six people in the beginning with a few users, And you're trying to imagine what it can become, how much it can compound. And it's going to look the same as some other startup in some sense. At that point, it's hard to differentiate.

32:57But you're thinking about, in many ways, that genetic potential unexpressed to grow up in over the arc of time. To that earlier point on dying of indigestion, this sort of fear that after the first product crests, the organization can't absorb a second or third or fourth act. something I learned from an investor I really admire or his his sort of recollection at least is that companies that go multi-product tend to do so early and those that don't do so early tend to never do so at all I've heard you say something similar I think it was Lattice that began its second act so to speak at something like five million in ARR how do you know if or when to go multi-product and that's a bit of a broad question but yeah I think that whoever told you that is absolutely right which is i think the um the companies who do go multi-product um and in many ways is in the genetics early even if even if unexpressed obviously um not not realized um spacex obviously famously has starlink now as a second business they were thinking about that a couple years in you had to get rockets that could land themselves to make deployment of satellites cheap enough and have a strong enough cool business to do that but they were thinking about that and actually made the first few highs on that um very early on lattice we were at a few million of revenue when we laid in a second product hubspot which you know is a great public software company started in marketing but laid in crm a couple years in um sometimes these things also take a while to play out you come back around to them uh square had its first wallet app i think it was about two years in didn't work out square had to grow more come back around obviously that's cash app uh many years later but it was in the ambition it was in the eyes that that that those founders um that team early on and i think when um there's a danger obviously of of sometimes trying to do too much too fast.

35:05And I think momentum is really important. Small amounts of scattershot progress is not really progress. You need the, to use a rocket analogy, you need the first booster system to really work and to get you into orbit for the second one to be impactful and take you further. A few boosters all in the wrong direction. You didn't really go anywhere. So I think the momentum on the core is really important. And a momentum on the core does allow and does empower in many ways having the bandwidth, having the potential, but it has to be there in the founders to dream of that second, third act. And I think building that into the muscle, the strengths of the company early when that first boost is really working is really, really powerful.

35:55Okay. There was this narrative not too long ago that the vast majority of value in AI markets, whatever that means, would accrue to the base model providers of the world, right? Your Open AIs, your Anthropics, your Googles, and so on. And that anything built on top of them was sort of on thin ice. it seems like you saw something different, which has become quite a bit more obvious as of late. But as we move from, as you were talking about earlier, software that helps you work to software that does the work, defensibility in a business shifts too. I've heard you talk about this as going from co-pilot to command center.

36:29But how do you think about building out moats in this new generation of AI-enabled businesses? And how do they make that transition from co-pilot to command center. Yeah. Our investing has gone on the journey in some senses, the market has from starting with the base models and now partnering with many great emerging application companies. And as we focus on this application layer, we think there are a few things that are important that we've looked for. The first is we talked a bit about the full stack teams. We don't think that just being an API call to the underlying model is enough. And so I think we saw that in some of the marketing tools in the first wave.

37:16They were a thin wrapper on top of the foundation-based models. But I think that shaping that intelligence into a breakthrough and differentiated product experience is really powerful. And so Cursus, as we talked about doing that with some of their own model works doji's doing that uh with with the diffusion models obviously the vertical models chai are doing that there but even rogo's working on um directing uh the intelligence with its own fine-tuning rl and the like and much more to come for financial queries really complex longer running financial queries and workflows where the product experience is manipulating the intelligence and exposing the intelligence to get the best product experience and so we really think full stack teams in that sense are important.

38:02The other is that the experience that they come to market with is a really magical experience, is a, what I sometimes think of as a can't unsee it experience, which is once you've been exposed, once you've seen it, you're not going back. Once cost is code for you the first time, once you've written a Waymo for the first time, once you've done a deep research or financial analyst career for the first time, like, yes, some days you will be doing it the old school way, but like you really, you can't unsee how slow you're being, how inefficient that is. And so the ability to deliver a can't unsee it experience to this moment in time, like you can get distribution fast on that because it is such a breakthrough.

38:46And that that product experience a bit to the conversation we were having earlier around great genetics, that is a high ceiling. There's going to be a lot of depth to that, which is it's not just going to deliver you simple marketing copy as an API call, but it's going to be a holistic workflow. It's going to integrate with maybe the rest of your dev cycle or take on parts of that. It's going to change not just a piece, but holistically command center, how you do a lot of the work. And then lastly, as an ingredient that we think is important to the application layer companies they can pace the race these teams can move really fast um there's a there's a ferocity to their ambition because um everyone is trying to win here and so i think these full stack teams that can deliver that the first holy shit experience um that can really move at speed of shaping intelligence into a product experience and and meeting customer demand which is it's great when there's so much demand.

39:45It's also hard to sometimes serve it really well, right? There is a lot of demand so they can meet the moment in that way and keep sending the pace of what the product capabilities should be and market leadership and adoption. It's hard to do. And I think will lead to these breakout application layer companies that we're starting to see in Cursor and Rogo and others. You know, on the investing side, common wisdom tells us that Startups often have two types of teams, those that get from zero to one and those that get from one to 100. Some say the same is true for investors, right? That someone can't be great at both seed and growth and that they're fundamentally different skill sets.

40:27You seem like a walking counter example. You're primarily focused on early stage, but how do you think about the lens through which you look at a company at seed versus one doing hundreds of millions of dollars in revenue? It feels like quite a context switch. It's hard, but focused on being great at it every day. And I think there is a juxtaposition there. And as we build the team at Thrive and support each other in being our best selves, we really think about being analytical dreamers and combining that analytic side with the ability to dream. and when you meet an early stage team, if you're all analysis, no dream, you're probably gonna have a small outcome.

41:16If you're with a late stage team and you're all dreams, limited analysis, you may have a disaster. At the same time, there's the challenge of not doing something early can be very costly. Doing something late can be very costly. And so there is a marrying of these that is hard to get right. But that's a muscle in some sense back to our genetics. The environment we want to have create genetic adaptation inside the team at Thrive is that of being great at it and pushing ourselves to become great, to be able to do both, to be those analytical dreamers. And we want to be great because we think we can be the very best partners to founders when we do that.

42:03And by focusing that way, we don't want to just find and support a great Series A company. We want to be a great company. And the same way we just want to select and invest in a great Series D company. We want it to be a great company. And I think the very best founders think that way. They're not thinking about how do I get a good Series B. You need to pass through that gate sometimes, in many cases, to be a great company. But that's not the goal. And that's not our goal when we partner early. And so I want to tell you something we partner late. It's because this company is still a fragment, even of its ambition.

42:40And in some sense, it's still an early company. And I think that analytical dreaming, when you bring it to early and growth companies, really allows for hopefully a great unique perspective, which is when we partner with the early companies, we want them to go far, not just fast. We want them to build a really strong team, not just win the first customers to get a Series A. We're imagining what it's going to take to write the S1 with them. You're not trying to be premature on those things, obviously, but that's what you're aimed at together, building that strong company. And at the late stage, as we were talking about earlier, the very best companies, the ones that are that next trillion-dollar company, they're sustaining their innovation.

43:27they may go down so to speak to go up they may struggle with building acquiring the next product line that's messy and hard to work through they may be working on a stripe with stable coins or vlad were thinking about a credit card for a while and and just because you haven't cracked it yet doesn't mean you want to give up on it and having a partner who yes could think about what the business looks like today and do the spreadsheet exercise and everything else but actually can keep the purpose of that company alive in their work with them, can keep supporting the investment in that, in their second and third acts, and the dreaming of that, the soul of that company alive in its inventiveness, and fan that flame, I think for the most ambitious founders is the sort of partner that they want.

44:16And so we really focus to be the best partners, the very best company is the one that will be category defining. We think no matter what the stage, you have to combine this analytics and dreaming. Silicon Valley, if we're talking about just the state of venture as a whole, has probably self-servingly told us that raising VC is the best way to capitalize any company. Many of the greatest companies, Microsoft, Dell, eBay, WhatsApp, MidJourney, more recently, were bootstrapped early on. In decades past, that's kind of just what you had to do if you wanted to run your own company and not be ousted by investors.

44:52But the dynamic has changed and capital is so abundant today that this isn't as much of a problem anymore. In your opinion, when is VC a suitable capital instrument for companies and when is it not? Yeah. I don't think everyone should necessarily take venture capital. I think those are probably great examples of cases that maybe didn't need it and are better off for not having taken it. I think it can be powerful in two ways. One is when you really want to try and pull the future into the present. This is easy to understand if you think about all the customer acquisition costs that typically go into building a business.

45:38You know that that product can deliver great value to the customer, but it costs something to go and acquire them and to onboard them and the like. You want to, in some sense, pull forward what would otherwise be organic growth from customer profits as a limiting reagent at the rate at which you can growth and almost borrow from the future to take market share, win market share, when you have to invest today to reap the rewards over time. I think the same can happen in product development as well. There's a scale of which you can probably invest in R &D based on customer profits at any point in time.

46:18And if the surface area and the market change is vulnerable and open to more inventiveness, more disruption at that point in time, you want to pull forward from the future of what would otherwise be customer profits at that moment in time. and use them today to take on those initiatives that you plan have really high rates of return. And I think in times of change, the chance, the opportunities for really high rates of return on that invested capital is very high. And so borrowing to take that market share to invent that future is, in some sense, people would say venture capital is expensive. It's much more expensive than debt.

47:06But like it's absolutely worth it because the returns of on that invested capital are very high. So that's, I think, one lens. There's another that parallels that but is easy to understand in a different lens, which is there are just some endeavors that are going to be a while till they are profitable or take a lot of money. And I think that is in many ways more classical, like very risk taking venture capital. Can we invest in a way that clears the hurdle in many ways of this change in the world being achieved? SpaceX might be a prime example of that. OpenAI has a certain amount of that. There's just a lot that has to be built to realize that opportunity.

47:51Even Monzo, which had the fortune of serving on the board of, a bank is capital intensive. Ironically, as you grow, you need more capital in the bank for regulators. We had the fortune of Josh founding OSCO, which is a health insurance. You need a lot of capital sitting there to have the regulators, to have a customer trust that you're good for the insurance. It is important that you have that. And so there are these bigger endeavors that are also capital intensive in those ways that have extremely high returns on that capital invested. Monzo now serves over 10 percent of the UK population in a couple of years as a challenger bank.

48:33Oscar's an incredibly impactful health insurer. Obviously, you can look at what's happening with SpaceX, et cetera. And so I think those are areas that almost the beauty of classical venture capital is coming to bear in really exciting ways again right now. And people should not shortchange funding those. And in some sense, if everyone tried to bootstrap, we wouldn't have all of these. We would not have Starlink. You know, we would not have the incremental competitive health insurance. We would not have the competitive bank. We would not have Nudge, you know, which is going to get to work on ultrasound, you know, therapy and more.

49:18It seems like Thrive was able to tap into a couple core truths about technology markets that really allowed it to win. One, that a generalist lens towards investing could really be a forcing function for focus and that you could win while investing stage sector and geo-agnostically. And two, that the enterprise value distribution or the extent to which monopolies or market winners dominate those markets was underappreciated and that everybody else was structurally undervaluing and underconcentrating into those market winners. What else do you think I'm missing? And as the fund grows larger and the capital product evolves, Are there any core beliefs about the world that you think it's going to lean into?

49:56I think those have been two very strong truths of Thrive, generalist and multistage. And the reason being that technology tends not to move uniformly through sectors. and there are moments in time where certain industries are more vulnerable to change. The ACA emerging makes healthcare more vulnerable as a case study. Mobile being the important form factor after social has gone and wired up makes referral and mobile acquisition easier for financial services. Even Plaid as an enabling layer for account authentication is actually an important prerequisite for some of the fintech being able to grow as fast as it is because it makes onboard easier.

50:54And so some of these things take time to work through their industries and being able to be sector agnostic has enabled us to work with the companies that will define those categories as those categories have turned over, have been disrupted, as amazing founders have come into with their authenticity, the depth of their ambition to shape it. And so if we'd stayed in social network investing, you know, we had the fortunate partner with Instagram and Twitch very early on. And if we'd have stayed just focused on social network investing, we might not have done anything else since. And so it allows us to do what we want to do, which is partner with the category-defining companies.

51:43And so we have to move across categories with them. the other is being stage agnostic. And I think the element there is twofold. One is the thing that you highlighted is that we would do what otherwise people might think of as late-stage investments, and we'd be doing them very early on. Even in some of our very first funds, one of the largest investments we made was in GitHub, when about$2 billion was considered a very late-stage investment, And we thought that was still early in collaborative coding and the role that GitHub played in the arc of cloud infrastructure. And we've continued to do that.

52:25And I think our view is for these very special companies that have a high ceiling to them, indigestion as possible, founders with, long arcs with ambition. what looks late to some by absolute size of the magnitude is still really early relative to their product ambitions and the financial performance that will come from the genetics of those products in those markets take stripe as we talked about earlier today processing now the same size as the whole market when we first invested but i think what is equally as important is um a huge focus So still most of us still spend most of our time meeting early stage teams, teams of two to 10, 20 people just getting going.

53:13The first early set of customers who are just embarking on their journeys, because those handful of teams each year that we are fortunate to end up partnering with for a really long period of time at these periods of tectonic change, which AI gives us, all the ones who will define, go bend those next categories to define them. And so as we have become fortunate to be able to lead some big rounds in companies that are so early, we have just as much focus on the two to 10 people just getting started that are going to shape the future. You know, Thrive as a firm talks a lot about being the most meaningful partner to its founders.

54:03What does this actually mean in practice, right? There's a sort of a spectrum of belief sets here from just giving people the capital and getting out of their way to being there in total lockstep, both of which have their drawbacks and benefits. But in what ways have you found yourselves working with founders? I noticed some level of intentionality, even in the words that you use. It's not a bet or a deal, but a partnership or a commitment. How does that materialize in practice? Yeah, we model our investing after the nature of the relationships we want, which is to be really concentrated. And so we're very concentrated in the number of investments we make so we can have a depth of support and relationship with those few teams.

54:47And you point out words. I think it is really important. Venture classically says that you're not putting chips at a roulette wheel. like you're committing to the success, the full potential and doing everything you can to realize the full potential of that company who's taking you on as a shareholder. And it's a high trust and total fortune to get to do. That's a commitment. That's not being at Vegas with some chips. And to that end, to support, we invest a lot and have amazing colleagues here at Thrive who focus on what we call impact, not portfolio services, but we call it the impact team because they're not there to just respond to services, respond to requests, though that happens, absolutely, but they're there to impact those companies that were partnered with success, impact them on recruiting and find the next great executive or the next great engineer or the next great recruiter for them, work with them on marketing, and comms and storytelling.

55:53We even have some engineers and designers on the team. We have colleagues who spend time on finance so that we can step in and be a key part of them in tough times and good times and denting even those. I have some fun stories. Monzo, which we mentioned earlier, I had the fortune of being on the board of, being a loss-making bank when investors pull out funding, not us, but some other parties, pull out funding because COVID happens and regulators are worried about runs on the bank because hopefully COVID is a fear-inducing moment and a loss-making bank in a fear-inducing moment is not necessarily a recipe for success.

56:44There was a huge set of demands on that company. from capital raising to the lease structures that we had for the business in the market to regulator kind of management and comms around this moment in time. And many people from Thrive stepped in, and we probably had three calls a day with different people on the Monzo leadership team, from the CEO to the CFO, helping out. Another one, when GitHub had interest for acquisition, one of our colleagues spent a lot of time with the finance team there because they wanted to keep it small, working through model questions and structuring and everything else with them.

57:38So good times and hard times are here to be able to step in. And the only reason we can do that is because we don't make that many investments so that we can be concentrated in that support. As we start to wind down here, Thrive has this really elegant, I think, mantra of being concentrated in both people and ideas. If you could distill down the few ideas in work and life that you take the most seriously, what would those be? It's a fun question. one I think I've I talk about with my siblings actually which is be passionate about what you're working on so you'll work really hard and I've loved the the really multi-faceted challenge of trying to build amazing companies from the product work of that to the team building work of that and getting to do that with the founders that we work with.

58:42I've worked a day in my life, but I've never really worked a day in my life in that sense. I can fill all of my time with it because it is just so great. And so be passionate about you once you work really hard. I think to the ethos of team, Johnny Ive did a great talk with Patrick Collison recently. And one of the things he said really struck with me and I think about with my partners here at Thrive is, you know, really care for the people you work with because you'll be vulnerable to really listening to them. and when to our point of being a being that biologist mindset being curious for that spark of greatness um really caring about our team our partners such that you can listen to everyone's unique perspective to get to the very best decisions together i think it's just a magical a magical thing and so i really i really feel that and i think um as an ethos uh without johnny's words we've thought about in the culture and the team all along.

59:48I'd like to lead in to this next question with an excerpt from a section in the old Nomad Partnership Investment Letters talking about Amazon. You shared this like a long time back and it was kind of a nice little excerpt, so bear with me for a second. It goes, my mistake in not recognizing that these businesses share similar roots might be termed by psychologists as a framing error. When looking for an explanation to a situation, the brain tends to latch on to what can be easily found to frame the situation. And if what is easily found is also vivid, then the brain stops looking for another explanation.

1:00:21I had gone looking for what I thought ought to be there, a vivid smoking gun, such as a brand name, a location, a clever reinsurance contract, or a patent. However, there is no a priori reason why a comparative advantage should be one big thing any more than many smaller things. Indeed, an interlocking self-reinforcing network of small actions may be more successful than one big thing. What about venture, right? Like in an environment with, let's be honest, so many dollars and so relatively few category defining businesses is sustainably out competing as an investor or as a firm more about doing a few big things or a million little things.

1:01:00It's a great excerpt from from the Nomad letters describing Amazon, I think, in large part.

1:01:12What's powerful in that, that's sort of not quite said there in many ways, is Amazon was all about breadth of selection, price for the customer, and delivery speed. I think were the core three, I think. and the letter there talks about taking light bulbs out of the vending machine to save on electricity costs at the distribution centers which is going to be able to drive down price which is going to be more value to the customer and so a thousand tens of thousands I think even in one of Bezos's letters I think he talks about the idea of like a thousand things they've done in that past year that have laddered up into selection, price, delivery, which is ultimately value.

1:02:04And I think thinking about that adventure, for us, it's to be concentrated in the very few, we believe, category-defined companies. And we're always thinking about all the many little things. We don't have as many little things as Amazon does, obviously, but all our little things, laddering up to making that possible. That's everything from the impact team, being able to have the bandwidth to support the companies in recruiting and marketing and finance and everything else. That's on the investment team, making sure that we keep our calendars free enough that when that two-person, that five-person team gets in touch with us or we get in touch with on a dime, we can be there working with them.

1:02:54The number of times this year that we as an investment team have had Sunday afternoon calls talking about an investment or talking about a portfolio company and working through something together is plentiful. That's a cultural thing. It's a small thing that ladders up into being able to be really fast, being able to be a really great partner to those teams. The ability, I mentioned Monzo back in the day to be able to step in in a tough moment for firms. You know, we anchored when some prospective investors walked in COVID, like we stepped in and we supported them. And now they're the 10%, 15 % of the UK population bank.

1:03:34And so all of the things that we think about are constantly refining or in service of that singular goal for us in a way in which I think is true for many companies in a way in which I think the nomad letters in some sense brilliantly articulated of Amazon in that letter. People who excel at their crafts, I think you often find are students of its history, but also have some method of sharpening the blade. I think it's Tyler Cowen who has this great question he likes to ask, which goes, what is it you do to train that is comparable to a pianist practicing scales? What does that look like for you?

1:04:16I wish I had more scales, cheap music to practice with along the journey at Thrive. We've just learned together. In many ways, there's this great clip of Swedish House Mafia being in the recording studio producing. I think the song was One. and you just watch it evolve. Like it starts and you're riffing and you're trading thoughts and you're tweaking something. And it's so beautiful. And then you get the ultimate creation. And I think it's not solo practicing scales, but the group practicing scales, I think is a really powerful way to improve and hone one's craft and get better. And I'm fortunate to get to do that with an amazing group here at Thrive over 15 years now almost, where we didn't necessarily have sheet music from anyone else at the time.

1:05:25We were all building this together. We didn't have careers that we'd come from from anywhere else. But we were in that studio. This room we were recording in was the entirety of the office in 2013. and we'd often be here at 10 p.m. at night all together riffing, trading ideas. Our investment in GitHub came from a late night, 10 p.m., being in our studio, so to speak, together where we'd had the fortune. I'd been really excited to partner with some of these developer services. So Segment was one of them and Mapbox. And I said, why not GitHub? It's the ultimate developer tool. and off we went pursuing that.

1:06:10And so I think as a group, that's a really powerful feeling and a way I love to practice and get better, so to speak. Miles, I have one more for you. It's the same one I ask at the end of every interview. So inside or outside the scope of anything we've talked about today, what should more people be thinking about? Whatever they like thinking about. it's i don't know it um there's the where's the world going versions of those questions i think um uh there's it's a time of such amazing change from um to novo protein design which is crazy that we have to the best tutor you could ever imagine in chat gpt to um coding agents to GLP ones to stalling.

1:07:09So you can get internet connectivity wherever you want in the world to a car that can drive you. Uh, my son, I have a almost four year old son. And sometimes when he gets in the, when we get in the car, he says, data, come sit with me. Cause he doesn't get that. I have to sometimes be in the front of the car cause we've taken Waymos together. And so it's just an amazing time. And I think with this technology across really many industries, which I think is what's powerful about this moment, because it's so fundamental of a change. I think in many ways it's bigger than what the internet was, which allowed us to wire up more and disrupt middlemen.

1:07:49It's almost more, I think, like what electricity electricity is in terms of its impact to be able to change. It's fundamental in that way. And I think whatever someone is passionate about, they should be thinking more about that with a how do I use intelligent computing as part of it? How does intelligent computing change what I can do there, what we can think of and dream of doing, et cetera? Because I think such a wise swath can be vulnerable and open to change. This was so much fun. Thanks for having me. Thank you for coming. Really great to do.