E27: Compound’s Michael Dempsey on Thesis-Driven vs Founder-Driven Investing

26:57Yeah. But you do have this belief that 2019 to 2021 rewarded scope creep and 2022 to 2025 will reward focus. Is that largely based on the macro? If the macro were to change in a certain way, would that change your sentiment or say more about why that's the case? I think there's like a bunch of different things. So I think one of the things, some of it was macro, which was that companies basically were allowed to come up and say, okay, I need to raise money in six months. How should I think about spinning around my story so that I can pop back up in three months and then have a new narrative that works?

27:37And investors get really excited. And so what that meant is that you needed to say, okay, I can do this thing. And then I can also do four other things that are going to be valuable. And the canonical example of this was Uber, which was originally Anand, former boss of Mayan who runs CV Insights, called it a Ponzi scheme of ambition. And Uber was really interesting because Uber basically out-fundraised Lyft by using this strategy of saying, we have rideshare, and then we have food delivery, and then we have scooters, then we have Uber Elevate, if you remember, which was like flying vehicles for delivery.

28:13and then it was we have ATG, it was self-driving cars. And then Uber reached a point where it was about to go public and it realized it wasn't profitable and a bunch of these things looked horrible to more traditional investors that weren't only narrative-driven and started divesting them and divesting them. And Uber still went public, still cratered in stock price after the COVID peak. And now it's gotten to become its first profitable quarter, I think a quarter or two ago, and it's at around all-time highs now. And so I think you can quantitatively see that you no longer are going to be able to play this continual scope creep game without having some fundamentally strong scaled product market fit on the initial idea.

28:55I still think you have people like Sam Altman who can build these kingdoms of products around them. But even Sam, interestingly enough, is now doing it across multiple companies. Right. Like we recently in the news has been him. He's raising money to do like chip manufacturing. But there's a world in which like if this was five or six years ago, Sam might be doing that within OpenAI and saying like, hey, OpenAI is like a frontier model developer and also makes the chips. And so I think that we're all just starting to see that from like a higher level tech landscape. And I think if you kind of filter that down, a lot of us were kind of coming up in our careers over this past decade.

29:35and we were doing it either as like younger people in tech or we were doing it as um other people older than you and i who were making all of their money in their careers over this past period of time and i think that it became really hard to take away any lessons besides like i need to be continually pushing the envelope of everything i do i can't have like a core competency or core focus um and i think we see we saw that in founders who you know were like angel investing, raising, raising, uh, really large rounds, um, launching different products, like doing a bunch of different things. And it all kind of came crashing down.

30:11And I think that like that culturally, I'm not sure we've fully gotten away from, but maybe that's like the last domino to fall, which is as more and more people start to kind of, um, shut down their companies in the next 12 to 24 months, which I think we've all been waiting for. Um, we might feel some sort of like recognition of like, okay, it turns out like this thing is really hard and you can only focus on so many things and do so many things well i don't know so that's like my high level view on scope creep and and more to that you have this belief that venture isn't broken contrary to you know what some people are saying people have just been acting you know acting impatient yeah yeah i think like again sam like sam lesson's a great example he's like very thoughtful guy but like all you can really complain about like venture so much like what we do is we back people taking the most risk of anyone in any part of the capital stack in human history.

31:03And we do it over and over again. And we should be right two out of 30 times every three to four years. And if you do that well, then it's not broken. It works. It keeps working. I think people just perpetually want to do the bigger and next thing without proving they're good or great at the initial thing um and so i i yeah i'm totally i look at that and i look at that narrative and i look at this conversation we're having around like creativity and like not really being technologically constrained ever and i just think it's like insane that these that these two conversations are having at the same time um because it's it's more of just like a a healthy reset or a healthy kind of like disinflation of a bubble that we've had it's not like some structural thing is broken.

31:52Like just don't raise$4 billion and like yeet a bunch of money in 18 months. Like it's not like, this isn't complex. None of this is complex. We all know how to not do these things. We just can't help ourselves. Yeah. It's, it's funny because going back to crypto and AI, both of those technologies, it was rumored or it was, it was claimed would disrupt something about venture, right? Like crypto member with the ICOs and there was this, you know, kind of novel way of raising money. And, you know, it was like crypto would just be different. Like it was it was the end of venture capital or something.

32:27And similarly, with with AI, there's, you know, like, maybe the technology itself would be used to replace humans in the process of evaluating companies. And I have a friend who raised a bunch of money to to try to do exactly that. And so it seems like, you know, rumors of the death the VC have been greatly exaggerated, at least from those technologies. And it just feels like it's going to be what it used to be. Even though people have been talking about, hey, venture capital hasn't changed for decades, and yet it invests in industries that entrepreneurs fundamentally disrupt, but VC itself hasn't been fundamentally disrupted.

33:05Is that what you think just is the case? There's just something about venture that makes it a cottage industry that is unlikely to have the same change as the industries it invested? I just think it progresses, right? It progresses from a hyper-cottage industry to a slightly more institutionalized one to now like a bifurcating one where you're either a Goldman Sachs of venture, a la Andreessen, or you're a small boutique firm, like a compound or like a USV or something like that. And the middle definitely like that's like the progression is those middle firms potentially get destroyed in some way or they die out or whatever.

33:49I also think that like it also progresses in maybe for the first time ever starting to progress more and more in terms of like the market participants, which is we kind of have like 20 years of the same people largely controlling venture. And I think now we're starting to have like a lot of new people. and so you'll have a lot of new ideas but i kind of think of it like uh when you join a company like we all have friends where we they join a startup and the first two things that they say are like this company is a total disaster because all startups kind of are in the middle and then the second thing they say is i have a bunch of ideas how i can fix this thing and i think when you have a huge inflow of new investors who are um maybe less students of like the history of the asset class and more aficionados of the asset class today, you start to get a lot of really low signal, high noise ideas floating around.

34:48And when a ton of capital flows in, unlike in a startup where the CEO will say, cool, this is an idea. We've thought about this. You're going to have a bunch of people who say like, cool, try it and like give you money. And they give a bunch of people money to do a bunch of things. And so I think like some good ideas will come. I think the idea of talent investing and pairing founders and incubation, those things are interesting ideas with experiments that will take iterations to get better. And I think quantitative-driven only investing and whatever Tiger was doing for a few years, those were interesting ideas that were not good.

35:25And they were done at such a scale that normally that wouldn't happen in the same way that if you joined as you know an ic at a startup they wouldn't say cool here's like a new product go launch it um so yeah i think i think it's about progression i think people are too quickly to want to disrupt this industry that is like very weird and like not optimized it feels kind of stupid when you look at it um and then as you do it more and more you're just like it's like that's the beauty of it and that's like the that's the the magic of of kind of what tech as an industry has, which is like, it makes no sense that people are able to raise a lot of money off of an idea.

36:04And that we value people who like drop out of college more than we value ones who finished college, like all these types of like nuances, but it's also like the awesome part of like, it's an industry built on optimism and taking tons of risk. And we all are like, okay with people losing money. So yeah, I think, I think it's just a progression. We shouldn't think about it as like full disruption. Yeah. And one also interesting incentive problem that you wrote about is this, and you sort of alluded to it when you talked about, hey, you know, these companies are going to go out of business, sort of zombie companies that exist, is one challenge that they have is that their VCs are relying on those markups.

36:42And they don't want to mark them down because that will affect their numbers. And so it's this, and the LPs are sometimes also incentivized by the markup. So it's this weird like chain where they're all incentivized to keep up these zombie companies. Yeah, it's tough. And it makes it really hard too, because if you're a founder, you're getting advice that you expect to be done as a fiduciary to your company, and it's not. And so I think that is really difficult. And I think right now, a lot of founders are looking around and understandably are pissed at VCs. And they're annoyed because they got bad advice.

37:17They're annoyed because they are being abandoned. And I think lastly, the worst part is they've been told for two years now, they've been told, hey, if you do X, I will behave in form of Y. Which is like, if you hit this revenue, you can get funded. If you do this, you can do this. And then they're coming back and they're like, hey, I did X. And the VCs are like, whoops, never mind. Sorry, I don't believe that anymore. And so it's a really complex time. And honestly, there's not a great answer for it. And I think like, it's probably going to be a really large, uh, just like a good learning moment for a bunch of people.

37:53And it sucks because that like this fun learning moment we're all having results in a lot of people losing their jobs, a lot of people losing people, money, et cetera. But, um, yeah, I just, I always go back to like, investing was, is like not supposed to be easy and like never was. And it's like an incredible privilege to manage large pools of capital for people. And so like somewhere in 2020, 2021, it felt like we all felt it was easy. and I'm glad people are now kind of like looking up and they're like catching their breath a little bit and like feeling it's very difficult. Yeah, totally. Speaking of difficult, I want to segue into bio.

38:29You know, we talked about AI, we talked about crypto. You've also thought about bio for quite a long time. Why don't you briefly trace your evolution as to how you guys have approached the space as a fund and some ideas of where you're excited today? Yeah. So I think on bio, one of the things that we originally thought about was a lot of people in 2014, 15, 16, was just the digitization of healthcare. And I think what that showed us was that there was a lot of really interesting ways in which technology was giving scale to traditional providers and also maybe bringing in novel care models. And so we backed a company called Tia, which is kind of a healthcare for money, a woman company, and now kind of all clinical healthcare for women.

39:17And what we saw there was just like a novel care model, which is technology enabled, enabled people to get better quality care. I think what we saw over the next few years really was that many other digital only services didn't really result in better care. and for us i think we are largely interested in like um non-incremental improvements and that shifted a lot of our focus towards what was happening at the early stages of biology and what is now called tech bio and why it was interesting to us is because it took all of the um kind of technologies that underpinned the things we liked like robotics and machine learning and brought a company called Juvenet on the therapeutic side very early on.

40:04We backed a few others as well. And we kind of just saw how these started to develop and how more and more these became product platforms and all the things that now people talk about, which is you have more shots on goal, you have more sophisticated and tighter feedback loops, and theoretically you should have much more success. And that's kind of the next generation that we're getting to. And so now when we look at bio we have this framing of like there are a large number of n of one companies and there are a large number of copycat companies and um there will be crowded spaces copycats not the right word and so we kind of are looking at what are like novel creative ways in which people can build new types of biological products and that's like some pharma some consumer products so an area we're really obsessed with um which is led by shelby on our team has been like science-driven consumer and a lot of that is built around the idea that there is new novel science going into consumer products for maybe like real science maybe the first time ever and that enables a bunch of business dynamics that we're really excited about um from retention to margin to a bunch of other things and then all the way on the other end we have more kind of just out there people that we think are building incredible things that maybe are less consensus but we've started to understand more from like a robotics angle which is like bio not where it's a nanomedicine the robot, you inject it, you use a magnetic system to steer it to deliver drugs very precisely.

41:28And we think that can enable people to get significantly better outcomes from lower dosages of drugs and just any sort of drugs in general. And for that, the core understanding of that was like robotics. And so I think for us, we feel more and more comfortable with taking pure play of regulatory risk. But where we think we are uniquely suited as a firm is that intersection of novel technology entering its way into biological systems and a bunch of different kind of random areas that we've talked about publicly there. And I recommend anyone follow Shelby's newsletter. She writes a ton about this stuff on our team and really helps kind of push through a bunch of those ideas.

42:04One horizontal thesis you've had for a while is the future of family planning. Can you say more about what's excited you there or what opportunities could emerge in terms of what you might back? Yeah. I think for that, that started in 2015, 15 maybe and um tia was an example of investment just because women's health is like a natural feed into their people don't build relationships with their doctors often until they start to think about having a child on the women's health side we backed um conception which is macrosoft's company which is doing uh embryosynthesis using stem cells um and we've backed a few other companies that have uh been acquired and or are still stealth today i think for us like the main thing is today we think it's very clear that there are a bunch of uh structural uh care models that could be innovated on the family planning side and that's on the fertility side the IVF side etc those are a little harder to invest in now and then maybe a few years from now there will be some of these areas that um fundamentally shift how people think about family planning whether that's on how people think about male fertility and some of the stuff like Contraline is a very well-owned company in this space we're not investors in but like a male birth control company all the way to some of the people who are doing research around artificial wombs and kind of removing some of the physical burden for people who either aren't able to have carry a child or for whatever reason elect to not to so i think most of these most of that space likely from us will come from really science-driven innovations but i think it's been really incredible to see when we back to women's health and family planning wasn't a thing like wasn't an investable category really and over the past whatever eight years it's become um maybe one of the most important and at least like hilariously enough like nerd sniped categories in silicon valley health care because some people i think are starting to like want kids who like kind of don't want to make some of the trade-offs of uh on the biological side so it's interesting um it's also super delicate and so we're yeah we're continually looking at a space we haven't made investment in the past couple of years.

44:13Yeah. Another example of, uh, of sort of overcoming the health stuff is the delaying menopause. Yeah. I mean, I think, I think all, yeah, all of these things are, we think a lot about longevity, um, as like an industry, right? We think a lot about longevity in two forms. We think about it as living longer. And we think a lot about it as having more years to your life or life to your years rather than yours to your life. We don't really think about it, the implications of those things. Um, and so if you have various like biological clocks across a bunch of things, sometimes having children, sometimes, you know, movement, et cetera, mental biological clocks as well.

44:49You actually have to start thinking about how do you make sure that you're optimizing the life in your years, in our view, more than the years in your life. So it's great that people are trying to figure out how do we live forever. And I think Brian Johnson is like one of the best things that has ever happened to the world, because this guy is spending millions of dollars trying to run a bunch of experiments that probably people would forbid him from doing. But a lot of these other areas are built around that life in your years thing. And I think people will fundamentally shift how they think about what stages of their life they are in as we start to have cohorts, people that on average live to whatever, 110, 120, 130.

45:25But I think these things are becoming a little bit more consensus, which is good because more people will start to build in them. How do you think about technical risk more broadly? We've talked about some deeply technical spaces. You guys, you know, have been in investing in them for a long time. How do you think about trying to avoid not being too early? I think about kind of what level of technical risk are you comfortable taking? How do you approach that? Yeah, I so a core belief we have is like you in 2024 and venture and technology in which we have the most innovative and there's a there's a tweet the other day about how five big tech companies spend on average the same amount as the Manhattan Project in R &D per year or something.

46:08I don't remember what it was. And so we have the most innovative companies in maybe history in terms of incumbents. And so we believe you have to be too early in order to get to any category and have proper value capture. And so we're pretty okay thinking through what are the time arbitrage between what people think is properly early or proper timing versus what is too early. I think some of that comes down to the fact that we believe kind of like the canonical, like people underestimate, decade, overestimate a year type shifts. But also I think if you start to look at the rate at which areas are going from academia or pure play research into commercialization, it's so much faster than even a decade ago.

46:49And that's across like every modality, right? It's across, the only thing again is bio and that's because it's regulatorily limited. it. And even then there's like a really interesting thing that I would urge everyone who's interested in this to look at, which is Montana has something called the right to try law, I think is what it's called. Don't quote me on that, but which just passed, which basically means that any drug that is in clinical trial at any phase, you now as a normal citizen can go and do and take in Montana on your own, or kind of, you can go and do it. And so like, we're even starting to break down the barriers of some of these rate limiting factors in biology.

47:21And so in general, we just think that you have to be early. You have to have a misunderstanding of time horizon based off of some form of optimism and some form of understanding when something has crossed the chasm. And I think we have this incredible system in the United States and broadly in big tech where like so much of research just gets pushed out for free and then people can adapt it. The biggest concern I would say that anyone should have is as we now have a lot of large for-profit or not-for-profit AI labs, we might see a closing down of research, which could slow some of that progress in adjacent areas.

47:58Meta, thank God for Zuck, is not doing that. So we're very comfortable taking that risk. And I think where we often feel strongest is feeling that people are being overly pessimistic about time horizon. Yeah, that's really interesting. Zooming back out in terms of how you think about venture, do you still think there's this big bifurcation between post-product market fit investing and pre-product market fit investing and everything kind of segmenting accordingly? Or how do you think about that? I think it's like pre-single point of proof and post-single point of proof investing. That's maybe how I'd frame it.

48:34The moment someone can point to any sort of data point that can be extrapolated a hundred times, they have crossed the chasm into some form of universe of investors that can then understand that business better. And that's because they are able to build a frame of reference for themselves, the investors, and for their partnership, where ahead of that, it becomes a leap of faith that a lot of investors don't want to make because that's how you look stupid is you make a leap of faith and you're very wrong, which is to be clear, unfair, but that's just like the dynamics of large firms with very top-heavy firms, a lot of people, etc.

49:11So I think there's that bifurcation. I will say my views on venture change every six months right now for anything post-Series A because it feels like there is such an immense lack of consensus and lack of conviction and confidence in the entire venture capital investment base. And I think you feel that very, very precisely in the mid-to-late-stage investors. And so I've never seen more people that are less confident in their views or are holding their views weaker than ever before. And I think it's why you see such insane flows of capital in any direction from climate to AI to crypto, back to AI, all over.

49:58Because it's kind of like we've entered into the like, you don't get fired for picking IBM phase of venture. It's like you don't get fired for like eating$30 million into a 2 million ARR AI business. and I think that's like incredibly negative for the industry but maybe that'll shake out over the next few years as people start to have a little bit more of a baseline feeling of like okay we made it through this this is what happened this is what didn't happen etc yeah that's uh that's really interesting what's a seed firm to do when you're advising your your your seed peers your fund managers um you know kind of in that sort of in between or you know not a you know multi-stage firm, what advice are you giving them or emerging fund managers, et cetera?

50:41My main advice is like stay small to the point where it makes sense and like figure out what is the thing that you are. I said, I tell us the founders too, which is make the decisions that if you fail, you can feel good about the decisions you made. And so I think for venture, there's a long period of time where, um, even for compound, we didn't know if this whole thesis driven research centric thing was at all right um it just was the way we knew we could do this job best and we were willing to go down with the proverbial ship doing that um and so i think it's trying to figure out what that is and i think it's like being really honest with yourself about what that is and i think people want to they want to be certain things in venture um when they're not and like i think to the point that we've talked about you can make money a million different ways in this industry and so i kind of say try to figure that out and then focus on being really good at it like i say like i don't i think compound you could ask rlp is like we've proven we're good at this i don't think we've we've proven we're great and maybe someday we will but like i have very minimal interest meaningfully expanding um or having new hypotheses to test in firm building and venture until we have that one kind of like fully in marker checked um Um, so I think that's probably the main advice I give.

52:01And then obviously like, I don't know, right on the internet is probably the other one. To that point, let me segue off a different question, which is you have, um, worked in the research business, um, you know, with CB Insights, it's kind of how you got your start. Um, but then you also think of your investment firm as a research organization to tie back to how we started this call and you monetize via investing, but there are businesses like Like Tegas, let's just say it's a different kind of research organization that monetizes via selling to investors, such that some investors who don't, certainly later stage or private equity or whatever, who don't subscribe feel like they're at a disadvantage.

52:39And that's something I'm exploring right now with this business, Turpentine. And I'm curious for your advice. If you were less focused on investing and more wanted to, you know, build a research organization that would sell to investors, how would you think about it? Or where do you think is the opportunity or the white space there? I think I would probably go back to one of the things I said before on the financial analysis side, which is how do you help people figure out what their novel frameworks are? So I think you can do two things. You have what are the novel frameworks that people should be thinking about as they look at businesses?

53:16And frameworks is a general term that can mean a bunch of different things. That's like metrics. That's from a quantitative perspective. It can mean a bunch of other things like data signals that people should understand that you create. um i think that and then i think the other thing is uh kind of to the other part like giving people ideas like the idea factory is actually the best part about research and when you're entirely unconstrained by anything but research ideas are like the that's the best flow state you can be in because it is entirely about figuring out um what are unencumbered what are where would ideas come from and what are the most interesting things that people should be spending time on they're not And kind of to my prior point, I would argue this is probably the best time in venture since maybe CV Insight started, which was I think a decade ago, to be selling to people that are looking to figure out a bunch of stuff.

54:14Because there's a very minimal amount of confidence and a lot of change happening. And I think those two things just enable you to build a brand moat. And once you build a brand moat, you have so many other areas that you can start to experiment in. And I think for CB Insights, the thing that we realized is the brand moat came from the tone at which we wrote, the quality of the data. If we said 32.2 million instead of 31.2 million on a chart, and someone emailed on end randomly because they're like, I've summed up these columns and it's actually 31.2, we would immediately hear about it. And it was like all hell had broken loose and we should never make that mistake again.

54:58And so we got this kind of brand around being very precise. And I think that's why media started to cover us more and talk to us more for their quotes and more kind of serious like private equity side investors in M &A group started to come to us. And then that enabled us to start to do analysis where prior to that, we were just some kind of random quote unquote people doing research on startups. but then people started to pay a lot of money for our analysis that was more opinionated and qualitative because they knew that like the rigor at which we had gone through on the other side. And so, yeah, I don't know.

55:31I think like, it's like an incredible time to build that business. I appreciate it. And you're the poster child. Like I'm trying to hire a bunch of analysts and I want to tell them, Hey, this is Mike Dempsey path. You know, Mike Dempsey did it for a few years, you know, earned his, cut his chops or, you know, earned his stripes, built a knowledge base, made that legible to the outside world. And, you know, VC firms want to hire people who have expertise in certain domains or networks in those domains, or maybe you can even be a emerging manager. So you're a, you're a success case for the type of people I'm trying to, I'm trying to hire you, you like, you know, 15 years ago, 10 years ago.

56:07Yeah. I'll happily, happily, happily tell them the story to sell them on your company anytime. Amazing. Well, getting towards closing here, you've given us quite a bit about where you're excited, how you've traced your evolution. Say more about Compound today, 2024, for people who are looking to learn more, whether it's founders or other people just looking to understand the model and any upcoming plugs, et cetera. Yeah, I'd say for us, follow us on the internet. We write a ton. What we're most excited about is people who are taking science and engineering risk often or who have really novel ideas around areas.

56:46And we back companies at pre-seed, seed stage as their first investor, leading coding in their rounds. And I think the main thing that we continue to kind of focus on is as a partner to founders, we care about being really high contacts on their space so they don't have to waste a bunch of time educating us and being really along for the emotional journey. I always joked like part of venture is being equity compensated therapists. And that's kind of what we do, you know, 70, 80 % of the time. And so, um, you know, always down to nerd out on everything, on anything. And we have a thesis database.

57:19We post a bunch of stuff there of ideas we want to see exist in the world and come tell us why we're idiots and why we're wrong. And, um, yeah, just excited to talk about all these types of things always. Awesome. And if you as a listener want to hear more of Mike, I, uh, I'm trying to convince him to start a podcast because he loves talking to smart people. uh, about, uh, deep technical things. So, uh, I encourage you to peer pressure him as well. Uh, Mike, thank you so much for joining us. It's been a great episode. Really appreciate it. Thank you.