The Exchanges

Q all the things we've talked about. And, you know, you charge them a platform fee and, you know, lots of platforms out there that are wonderful businesses. Versus like, okay, we have the platform, but we're also going to create the generations, the vertical application companies on top of our own raw tech capabilities. What are, what are the trade-offs of one approach versus the other? Are they mutually exclusive?

A I think in the limit, we're going to be able to explore that design space more fully, but it really depends on what's the market you're entering with that platform capability, right? So a lot of successful platforms are just, are entering markets where there's already a ton of vibrant activity, and they're helping to, you know, grease business and make that, um, happen more fluidly. There's not a lot of fragrance companies out there. Right, so there's not that many buyers. Um, The question is, like, do you become a software provider for the incumbents, or do you take your capabilities and do you compete in that market? And I think there's been examples on both sides of this. There's plenty where you are an input or a service provider. I think a recent example where folks decided to just enter and to compete would be, like, Metropolis, if you've heard of that example. I can, they're making software for managing parking lots, um, the parking lot industry just weren't ready buyers of that software, but it actually worked. It made parking lots more efficient, so they became a parking lot company, right? We went through a similar journey where we, I mean, if I could have sold the software here, and believe me, we've tried, like, yeah, we'd be selling software. We might not be talking, um, but there's not that many fragrance houses, period, and Um, I think that we have the opportunit…

AI assessment note: “it really depends on what's the market you're entering with that platform capability”

Q in government and at the White House, but it wasn't long. You were under investigation by special counsel, which is a very scary sounding term that I would not want to be on the receiving end of, about colluding with Russia. That must have been a tough and stressful episode of life, especially so early in your government career. What was that like and what did that experience teach you?

A I try to laugh about it in retrospect, but it was quite a surreal experience. And I'll never forget, I was getting interviewed in some windowless conference room. I had my two lawyers next to me. I had three FBI agents and three representatives of the special counsel sitting across the desk. And I kind of had like an outer body experience where I'm like looking down on the room and I'm like, you're being interviewed by a special counsel investigating the president of the United States who happens to be your father-in-law. How the hell did you get here? You were a kid in New Jersey. What's going on? For me, it was just a massively surreal experience because we went through the campaign and we could barely collude with our team in Iowa. And on most days I couldn't collude with Donald. And so the fact that they were saying we colluded with Russia was crazy. Initially, I didn't make much thought of it because I was like, we didn't do that. That's a crazy accusation. But then I saw all these institutions that I had respected previously, like the New York Times and CNN and the Washington Post, breathlessly covering this thing like this was the next Watergate scandal. I was just kind of surprised with how serious this was being taken by everybody, and everyone in Washington thought it was real. And so it was very stressful, and the thing that I felt very badly about was a lot of the g…

AI assessment note: “I try to laugh about it in retrospect, but it was quite a surreal experience.”

Q So you've known for, I guess, almost a year now that there's a decent chance that Donald Trump would be president again. How have you taken action to prepare yourself and the business for that potential outcome?

A It's hard to bet against Trump. So as he started doing well, I thought there was a pretty high probability he could become president. And so the things we did to prepare the business, one is basically just keep our high compliance. We spend millions of dollars a year on compliance. We had our first exam from the SEC and we got a no action letter. And that's very important to me. I wouldn't do that because we're high profile. I just do that because in life, let the investing part of it be hard. Don't let any technical things Make your life hard. Number two is we were getting to the place at the end of our fund where we were gonna have to raise more capital, so I spoke to my investors earlier this year in February and said something that we should talk about if you're thinking of re-upping in the fund, let's do it sooner as opposed to later. They all agreed, and so over the last year, we closed a billion and a half extra capital and extended the investment period of the fund to 2029. My investors really liked that we went slow in the first two years, so adding an extra two years On the end was something that was easy for them to do, and increase in the capital we did from lunating QIA. That gave us more firepower to be able to go forward. We closed all that before the election. They wanted to do that irrespective of what the outcome was, and I made very clear to them that in the …

AI assessment note: “the things we did to prepare the business, one is basically just keep our high compliance”

Q then the question becomes, do scaling and improvement laws hold forever, or for a really long period of time? And if they do, then there's value to being three to six months ahead, and that will just last as long as it lasts, and they can charge a huge premium for those tokens relative to a very cheap open source token. Is that the right way to think about it?

A I think it's possible. I don't know that the premium for being three to six months ahead is going to last that long. I mean, if you look at, like, enterprise deployments, uh, they don't move at three to six months speed. A lot of enterprises are probably still on, like, four six, Opus four six, or Opus four seven. They don't, They don't adopt the bleeding edge rapidly. There's a lot of questions that people have around rolling out any change at all. And I think we're just so early in scratching the surface that, um, I don't think there's any way to call a winner in this race. And certainly I don't even think this is a race that can be decided ever. There's always, it's a continual process. And fundamentally, I don't think open source ever goes away. If there's a vacuum because one leader steps out, a new leader will step in. There's too much incentive and too much. There's a lot of tailwinds too. It's just a It gets easier every day to treat, to train a frontier class model.

AI assessment note: “I don't know that the premium for being three to six months ahead is going to last”

Q So if you think about, uh, the theme of the day being token costs, is token costs the right way to think about this? Like, is there some other way you'd put it?

A To start with, absolutely, token cost. Today, my north stars, I want to have the lowest cost per token in the industry, and do that by a mile. I don't think tokens are the final unit of, uh, of work or intelligence, but they are what we use today, and so it's very straightforward. I think after tokens, you start to move more towards, um, more outcomes, which is like a vague direction. Uh, you can imagine, for example, today when you can zoom tokens through an agent, you don't actually control how many tokens the agent reasons for. It can reason for a certain amount of time, or it can call a certain number of tools. And increasingly, I think we will have agents do some unit of work, take as many shots on goal as they can, and however many tokens they use to get there is going to be kind of a dependent variable depending on the task. So you think about, like, agents that self-administer a token budget as opposed to a company setting a budget for how many tokens engineers can spend per month.

AI assessment note: “To start with, absolutely, token cost. Today, my north stars, I want to have”

Q And I guess the question is just like, so what? Like what, what, if you're maximally successful, dream a little bit about what that might enable.

A Yeah, absolutely. So I think for individual users, what I'm excited about most is this idea of proactive intelligent agents. Um, you can imagine a Siri that is running in the background all the time to understand what's all the emails you received in a day, all the text messages you receive in a day. And It has a much more encyclopedic view of your life and how to be helpful in that life. Right now, there's still point solutions, and so you have to, you end up doing a lot of prompting. Siri is not very proactive. It's something we can fix with abundant, abundant inference. If you trust the machine enough that it's reliable and also trustworthy as in private, um, you might even imagine the machine can understand how you interact with it and practically surface your next action. Whenever you open your phone, can we build a good model of what you're going to do next? My estimation is yes, we totally can.

AI assessment note: “what I'm excited about most is this idea of proactive intelligent agents.”

Q What did IBM do? What's the, what's the analogy?

A Well, so IBM, um, So IBM had this dominant, you know, we talked about it in the seventies. Uh, and then you fast forward to the nineties and IBM is this very sort of distressed asset. And the thought was IBM needed to break up and all these different pieces they had. So Lou Gerstner comes in and he takes it over. And I think Gerstner's real key insight to IBM is actually everything. We're pretty mediocre at everything. It's kind of like what I have with Microsoft before. And that's the price of monopoly. Once you've been a monopoly, you kind of lose your capacity to be good because you're, you didn't need to compete anymore. And I think a lot of tech incumbent companies have this problem. They, it didn't matter what they did, they were going to rake in money. And if you don't have the pressure, if you don't have the incentive, if you don't have the fear of death or the fear of God, as we talk about these model companies, then you don't do your best work. And the problem is that you, once you lose that muscle, it's gone. You're just sort of fat and flabby. And so what Gerstner realizes actually the worst thing IBM could do would be to break it up into component pieces because all those component pieces are actually not very good. Our biggest asset is that we're big. Say what? No, what does it mean we're big? We can, it's the nineties, this internet thing's coming along. There's …

AI assessment note: “That's basically what IBM did. So they built out, and this is an echo”

Q You mentioned the word differentiation before. Can you explain your law of differentiation?

A If, if anybody wants to learn sort of three, like, three laws of physics that are the most important in fundraising, law of differentiation, law of trade-offs, and law Pipeline. Let's talk about law of differentiation. This is, this is like the law, alright? This is your track record plus your differentiation, and you divide all that by the complexity of your story. So, track record. That isn't just your returns, but is, is how do you behave? So if you're an official, if you're an elected official, your track record is your voting record, right? Or it's the way you show up in the media. It's your consistency. Differentiation can be anything. Like, it can be, I can take contrarian bets. It can be, I only do one or two things, but when I do them, I'm highly operationally intensive. It can be, I access this part of the market that no one else does. It could be my GP commit is, is abnormally large. So let's take those two positive features when you're trying to build a portfolio, because almost everybody at the institutional, at the big money, not the small money, big money has a portfolio. So you have a portfolio of diversified assets, and you try to have those assets not replicate what they're each other's doing, because if they're auto correlated, then, you know, you didn't do a great job. So you kind of want people who are Differentiate it. And so, you're trying to add somethin…

AI assessment note: “This is your track record plus your differentiation, and you divide all that by the complexity”

Q If you think about your whole set of experience doing stuff like this, you were talking about this idea of inner games before we started recording, and you're, you're being interested in the inner game of interesting, exceptional people. How would you describe your own version of that? Like, what has the inner game been like for you across this period?

A The inner game of fundraising is really about putting the person in the room as the center of all conversation. That I actually don't exist but for the fact that I'm in your mind. At this moment in time. I actually am just an object in your mind, and that object in your mind is being processed by all the stuff that is Patrick O'Shaughnessy. And so, now that I'm living in your mind, What can I do to, in this case, make myself interesting, make myself compelling, make myself somebody you want to meet again, make myself somebody that you're satisfied that you actually invited onto your podcast, because I'm inside your mind, and what is going on inside of there, and when I look at you, I see such a deep curiosity. I see incredible patience as well. You're allowing me to have these long form explanations. So when I'm talking to you, I really myself don't even exist over here as much as I exist inside your head. And that's the inner game of most I think of the most or the highest level of persuasion. Mentalists do this. They get inside your head. Hypnotists do this. Psychologists do this. Anybody who is engaged in a mental discussion, if they're really good, they're not just saying, here's what I am. Who are you? And how do I address you in a way that actually is satisfying to you? And I hope I've done that in today's podcast.

AI assessment note: “The inner game of fundraising is really about putting the person in the room as the center”

Q You take a job, you show up. What, what is the impression on day one through 10 or something like this? Like, what did it feel like?

A I mean, complete chaos. Uh, the, the company, Travis, uh, had not been at the lead of the company for some period of time. Uh, and there was a group of executives, uh, a committee, if you want to call that, who had been running the company for a few months, and obviously the company was in the public sphere in many, many bad ways. Lots of things going on, both internally, externally, and, and the business itself was very dynamic, hugely competitive, so it would have been difficult enough if there weren't any of the external distractions going on coming into, uh, Uh, the business, which I think fundamentally was strong, but was going through a lot of change, a lot of chaos, both in terms of the board and what the board wanted, in terms of the stability of the management team, and then getting structure around the business as well. But, you know, with time and with a lot of work and working with my team, we were able to bring some order to the chaos, both externally and internally, and It's turned out to be a great ride for me.

AI assessment note: “I mean, complete chaos.”

Q that happened, like, did you personally believe that it was possible? Like, did that seem absurd? And has it continued like now that it's becoming consistent? And so maybe, maybe it's becoming more, uh, commonplace to you or something, but What was your, like, own view staring at this cone about, like, the odds of hitting, you know, a 10 X type of growth so many years in a row?

A Well, when I joined the business, it had about two hundred fifty million of run rate revenue, and the plan was to get to a billion, and I said, great, in what year? And that was, like, linear thinking, right? And, you know, consistently, you know, Dario has been a much better predictor of the revenue, uh, than, than I have. I think we're gonna close the gap over time as we get better at forecasting and understanding the business, but Yeah, definitely the first time I saw it, you, you have all these arguments about the laws of physics, and law of large numbers, and this can't, you know, where is the revenue coming from, and how can it be added this quickly, and how can customers move this quickly, and is this even possible in enterprise, and all of those things start to get broken down over time as you see how the business works internally, and you see how the adoption curves and the, um, exponentials that are happening, again, we have the exponential that's happening on revenue, but that's Underlies it are these many other exponentials that support that. You start to see and believe in that more. Now, that doesn't mean we're not disciplined and thoughtful about the forecast and how we think about the range of scenarios, but it does mean that like the, my thinking has at least shifted a lot more from linear and incremental towards, you know, leaning into this exponential and rea…

AI assessment note: “my thinking has at least shifted a lot more from linear and incremental towards”

Q What are you most excited about? Like you get to, you have a privileged seat. You sort of get to literally see the future because it's happening inside the business before those outside the business see it. With that perspective and in that seat, what, what, what are you most excited about in the future?

A I really think that the, the biotechnology and healthcare outcomes that can come from this technology are the things that I'm most optimistic about it. We may live in a world where you're diagnosed with a disease that is not curable, but in your lifetime that cure can be found much more rapidly and you actually might not die of that disease. And I think of this as like, you know, a lot of what we're doing today is helping to speed up The, the, the drug, uh, development process, right? A lot of the paperwork and clinical studies reports and things like that that are needed to be done. AI and our solutions in particular are helping to rapidly accelerate that. I'm really most optimistic and excited about when it goes further back into drug development and drug discovery, because, you know, our humans are, are, are incredibly capable at research, but if you think about these molecules and proteins, like they're, So complex, and such small changes have such big implications for the outcomes. AI is perfect for that. If you think about what can happen when the lab's throughput goes up 10 X or a hundred X, and we can run that many more experiments, you can probably get better results faster. And that can be something that helps, you know, people around the world, right? And it doesn't have to be limited to a small set of diseases or disorders that can really go Much further down the ch…

AI assessment note: “the biotechnology and healthcare outcomes that can come from this technology are the things that I'm most optimistic about”

Q everything was frictionless and I was a GP, I would of course have matched liabilities. Like if I could just snap as much capital as I wanted into existence. Yeah, of course. Like I want to have no problems. So what was going on? What is the, like, what's the series of events starting in 2018? What were the first examples of this? And then how is it like evolved?

A So the first signal is underwriting because investing or lending, you can invest As much money as you want , you can lend as much money. That's not the skill. The skill is investing. It's that artisanal behavior. So when you start to see, and again, it wasn't just, everyone talks about private credit, but we started to see it in every asset class. We started to see it in real estate. We started to see it in infrastructure. We started to see it in private credit. It wasn't actually Bad, but we started to see behaviors like terms that you would never do, because obviously when you lower your underwriting standards, guess what happens? Your deployment pace can go up. You have an origination engine, you're sourcing all these deals, and let's say you're an artisanal, you know, you might have a hit rate of, you know, half a percent you look at. If you lower your underwriting standards, your hit rate on deals that you might do might go to two percent or three percent. It's literally all in your control. So I think we started to see it, but it wasn't, it was just, like, something we started to notice, like, changes of behavior, but it wasn't, like, full-fledged factory model industrialization. COVID happened, and then post-COVID, it was, like, game on for the factory model, both on the liability raising side and also on the asset side. Literally, that behavior started to accelerate In …

AI assessment note: “So the first signal is underwriting... COVID happened, and then post-COVID, it was, like, game on”

Q So if five criteria companies don't outperform eight criteria companies, doesn't that imply the criteria aren't predictive? So then why have the criteria?

A Because you need to set a framework for what to focus on and what not to focus on. That's it. Like, it's just getting to a small enough. It's not predictive, but it's getting us to a small enough pool to, like, it's like knowing your strike zone. It's like, my partner is a big baseball fan. He uses a baseball analogy. Like, Ted Williams knew in the hitting zone exactly where to swing and what his probabilities were swinging the ball. Like, yes, you can hit a ball two inches above home plate, and it could be a grand slam and hit the ball the far as you've ever hit it. But if you do that over an entire career, your entire career won't be very long. Um, and so it just enables us to know, like, what pitches to swing at. Our biggest mistakes have honestly been Not swinging at the pitches when they were in our strike zone. And I think that's like what we've learned over the last 15 years to get more comfortable and like when it's in our strike zone, swing at it.

AI assessment note: “Because you need to set a framework for what to focus on”

Q I'm thinking about your mentors and your co-founders. I'd be curious if you go one each from Beto, Marcel in Georgia, what lesson stands out that each of them taught you?

A So, Georgie has this incredible ability to see very far. So he really understands the potential of a business, the potential of a person, and his vision, I think, is unique. He's unique that way. He thinks very clearly and is able to chart a path Out of any situation. Beto has this incredible ability to relate to people and lead people and get people even at the shop floor, quote unquote, of a business excited and enthusiastic and is someone that's completely fearless. And Marcel is probably, of the three of them, the one that really honed this business model that we all liked the most at the brewery when he ran it. And it was able to basically create so many good people over the years and a very clear process. Of course, what we do and what the companies do have their own different flavors that evolved from it. But he was the most involved in creating that operating model. They're all very complimentary. If you put the three things I said together.

AI assessment note: “Georgie has this incredible ability to see very far... Beto has this incredible ability”

Q Yeah, of course. Walk it off in my news catchers. What are the most misunderstood or surprising things about three G do you think from the outside?

A I think people, again, may not perceive how focused we are on business quality, first and foremostly. If you were to participate in investment discussions here, for instance, what proportion of those meetings is dedicated to determining whether a business is really good or not, and the bulk of it, versus talking about what the cost opportunity is, that's secondary to that. And secondary to the quality of the business and the growth potential. I mean, that may surprise some people, frankly, that look at us from the outside in. I think they might be surprised how lean we are as a group of people, given the size and global footprint of some of these businesses. I don't know what else then.

AI assessment note: “I think people, again, may not perceive how focused we are on business quality”

Q Why do you think 12 to 24 months is, is a timeframe worth mentioning, that some of this stuff will start to be felt more broadly?

A It's all kind of starting to take effect now and, you know, it's got to roll out, get deployed. Now, you know, deployments of technology in particular in the past have taken a long time. Um, but you know, you had to build out the infrastructure to do it. So like for cars, you needed things like roads and traffic lights and all that kind of thing. And for the internet, you needed, you know, fiber in the ground and, you know, people to have smartphones and you needed to do a lot just to get going. Um, the internet is here, so if you want to use AI, if you want to apply it to your business, you just do it. Like, there is no infrastructure that needs to be built to adopt the thing.

AI assessment note: “there is no infrastructure that needs to be built to adopt the thing.”

Q that basically this kicks off an industrial revolution for services. This is an interesting opportunity to ask about what your philosophy of product is. Um, you're such a product centric person and builder. That's, that we, that's what you've done. That's what you've invested in. As we face down this like industrial revolution for services, what, what is your like broadest possible philosophy of product as we enter this era?

A Very simple. A product Person or product manager, if you call them, their job is to balance customer needs and business needs. The product manager, there has to be somebody at the company who's the keeper of the why. Why are we building it? What customer need are we solving? Why is this a pain point? How intense is it? How deep it is? And second, how does it add value to the company? If you build this thing, solving this customer need, how does the value add to the company? And I think balancing those two It's a very delicate act. You can build something amazing that adds a tremendous amount of value to the customer, but doesn't build any value to the business. And you can do something that is awesome for the business by raising prices, but it is value detracting for the customer. So balancing customer needs and business needs at the highest level is what I think of the product and what it comes down to, in my opinion, over the last 10 or 15 years, I've really gone down to this notion of outcomes. Outcomes, I think, are what define the best product people. And outcomes have to be defined in the form of customer behavior. I strongly believe that the, because customer behaviors are leading indicators for every business outcomes. If you think about it, the simplest thing that a product does is to make somebody go from not a customer state to becoming a customer state. And from bec…

AI assessment note: “their job is to balance customer needs and business needs.”

Q and then you also said that Jack would do this across the company, not just in the product. How would you sum up the process of great design that you've observed from the people that are the best at design? What is the, what is the thing there, the method that they're going through over and over again as they apply it to different parts of the company or product?

A The number one thing I've seen is they try to minimize the number of steps. Everything should be in one page. And you need to cut down things. In fact, Jack called the product manager role product editor. Why? Because he believed rightly so that the role of the product manager is not to add more features. Any of us can look at a product and say, here's 10 things you should build. The best, the best designers, the best product people edit down things. Similarly, we have a hundred features. What are the two things that really matter that will drive the customer outcome? So the best designers really take 10 pages of design, and say, cut out all the experience. So I think it's the process of editing, and this goes to judgment. I think this is, in an AI age, humans with amazing judgment, which is really editorial capabilities, are the ones that are going to do well and thrive.

AI assessment note: “So I think it's the process of editing, and this goes to judgment.”

Q the, the job of acquiring the customer, positioning the product, marketing the way it sort of presents itself to the outside world. What's the dispatch from like the cutting edge that you're seeing of how people do this? All these things, position, brand, customer acquisition, the ways they do that. What does, like, new excellence look like to you across this, the many, many companies that you get to see?

A One of the most interesting things now, it's different between enterprise focus and consumer focus. For consumer focus companies, the big thing is how to scale influencers. I think influencers have become much, much, much, much more every year. They become much more powerful in how people, especially younger people, consume products and even choose products. Somebody said that TikTok is the best local search engine, and I think that's right. My kids have discovered crazy when you go traveling, crazy restaurants on TikTok that Google Maps would not really show or Yelp doesn't show, etc. So how do you reach influences on TikTok? And there's a set of companies that's coming out that's essentially making it easy. The problem is, Influencers on TikTok, obviously there's head influencers, but there's a long tail that go viral for different reasons, and you want to capitalize on those viral waves if possible. So there is a set of companies that is building products to see if they can help brands connect with these influencers in scalable ways. On the enterprise side, I think the most interesting thing I'm seeing, it's not really a, um, acquisition channel as much as it is a, uh, Uh, onboarding channel. It is basically presenting an outcome to a customer and saying, let's collaborate on outcomes. Palantir does that very well. Palantir goes to customers and say, what's your most importa…

AI assessment note: “For consumer focus companies, the big thing is how to scale influencers.”

Q You said fast and loose. Can you say more about loose?

A If you over manage, for example, a tight process or specific hours that you have to be in the office or a wide variety of things, you filter out, uh, performance and creativity. And the looser that you can run, the more creative that the organization will be. So we talk about it as managing on the edge of chaos. You don't actually want to fall into chaos. Okay. In chaos, the product barely gets released. It's full of bugs. People are upset. Payroll's not made. Lots of bad things happen. Okay. But it's getting us close to that edge of chaos where there's last minute saves and a lot of dynamism, uh, as you can possibly tolerate as opposed to say a semiconductor factor. which is trying to reduce variation and reduce error to get rid of variance. If you're going to be a creative organization, you want to be high variance, high creativity, and again, managing on the edge of chaos.

AI assessment note: “the looser that you can run, the more creative that the organization will be”

Q and story behind Netflix, the sort of invisible Part of the business everyone just takes for granted. They can hit a button and have this beautiful thing pop up, but I know there's quite a lot of building that happened behind the scenes. Can you tell that part of the Netflix story of what it took to the infrastructure wise and technology wise to make what we all enjoy possible?

A Well, it's always been a sort of medium barrier to entry. Um, I would say, uh, first with DVDs and we had incredible sorting and shipping machines and postal integration and, you You know, I used to spend all this time on types of polycarbonate plastics that break and don't break. And we were impressing plants. And the biggest issue we had was that the DVD would get to you without cracking or shipping or being damaged. It was on time. The postal carriers didn't steal it. So there was like, you know, a huge amount of machinery to shipping a million red envelopes a day consistently, you know, kind of FedEx style. Right. And then certainly, uh, streaming the mechanics of getting the bits to people. Uh, you know, was challenging. We first, uh, launched in 2007. And for probably 15 years, the internet was underpowered and you had to do a lot of clever engineering things. But for the most part, the, you know, there's a hundred companies that stream now, uh, consumers can't particularly tell a difference between them. So I would say that's now just become part of the base, um, uh, systems. And commoditized. What's unique is still being able to do the AI recommendations, uh, all the deep learning on what do you, you know, there's a thousand things on Netflix you would enjoy. Which one would you enjoy most at what time? Um, you know, that's still a big area of, uh, tech innovation. Um, …

AI assessment note: “first with DVDs and we had incredible sorting and shipping machines and postal integration”

Q If you think about the reasons why you do this versus something else, what are the most important ones? Like why, why don't you, why aren't you a founder? Why don't you work in some other industry? Why don't you have your own firm? Like there's other things that you could do. What are the most important reasons why this is the thing you do?

A So my wife would say that I have a low attention span. What she means by that is I'm interested in a lot of different things, and this is a really cool way Of getting to learn about tons of new stuff. I suspect that this is the same reason that you like to invest is how lucky are we? We get to sit and spend time with the entrepreneurs who are building the most interesting companies in the world right now. We get to learn about the most cutting edge technology stuff that if you were in the public markets or just in a job, you would never get a chance to learn about. So I love to learn and I love to be around You know, kind of great founders as they're exploring really interesting things. So that, that part of it is really, really attractive. There's another part that plays to a totally different side of me, which is this business is a scoreboard business. And like, I convey this to our team all the time. There's a scoreboard in this business. And our expectation is that we win. Now it's a very long dated scoreboard, you know, especially in the venture side, but on the growth side, even it's a pretty long dated scoreboard. But at the end of the day, like we have to put up returns, like our customers are our founders and our LPs. And on the founder side, we need to make sure we do a great job with them. And there's sort of a virtuous flywheel if we do. On the LP side, like, it's p…

AI assessment note: “I'm interested in a lot of different things, and this is a really cool way”

Q What are the most interesting strategies or things that upstarts do to beat incumbents? Like what are your favorite? Ways that company companies beat incumbent.

A Business model shift is a super powerful thing. That's very hard for incumbents to react to. That's part of what is so exciting about the customer support industry and Decagon. It's like the odds are so stacked in their favor because the business models can be very hard for incumbents to react to. And it's on the customer side, better, faster, cheaper by an order of magnitude in, you know, in each case. So business model shift is one. The two simple components that I'm looking for, which generally we're not really seeing yet, is completely re reimagined UI and then completely new sources of data. So we're large investors in Databricks. We're very optimistic about the data layer. I think they'll have some success in, you know, enabling applications built on top. But the UI UX thing and the data thing, I think are what paired with a business model shift, I think are what are going to give the startups the best chance against the incumbents. The more dramatic the shift in those, the harder it's going to be for the incumbents. So take salesforce.com. Like I use this as an example, like it's a good company. I never thought it would be as big as it is. It's a, it's a good company, so maybe they'll be one of the incumbents that survives and, you know, reacts. What do people do in salesforce.com? It's basically like a sophisticated form checker with some, with, with some analysis, and …

AI assessment note: “Business model shift is a super powerful thing. That's very hard for incumbents to react to.”

Q Do you basically just not care? Like if a company has zero percent gross margin, for example, but the revenue growth and the customer love and all this kind of stuff, the poll is all there. Does it round to we don't care?

A So there's a big difference between having 30% gross margins and 70% gross margins. So we do, we do care. Our expectation is if you're producing a lot of customer value and if the models get a lot better over time, you're going to increasingly produce customer value that The cost is going to go down. There's not going to be so much market power of the model providers that it's going to settle out where these businesses are probably higher margin businesses. I think there'll be lower margin businesses than SAS businesses. You know, maybe they end up as 50% margin companies as opposed to 80. But the size of the impact and the usage and the amount that they'll be able to capture to our point on business model earlier is probably so high that it's fine.

AI assessment note: “So we do, we do care. Our expectation is if you're producing a lot”

Q It was actually Daniel Eck, At Spotify who introduced us originally, and obviously he's built one of the great, you know, European origin businesses. Why do you think there are not more of them? Like you, obviously you're seeking to change this, but, but there's not that many. What do you think the, the deep reasons are?

A I think the main reason is a matter of default. Why does California have so much, has had so much success over the decades? One of the reason is you have seen incredible companies being created and grown in California. You just assume that's where you go and do it. Especially as a founder, you don't know much, like you are ultimately, at least I was, I think that's true of many founders, you are a passionate, determined, maybe talented idiot, essentially. And so you don't have, you don't know the world enough to, to actually determine what, what the ideal location will be. If you even think about it, because typically, how many times have you heard of founders doing a kind of a location study, where should I start my company? It tends to be kind of, Momentum. Like, I happen to study here. I know people are there. I should probably just do it. And so, a lot of talented Europeans, many of the most talented Europeans who have an entrepreneurial streak, I think they just default to building in the U.S., which is being, it's been fantastic for the U.S., of course, and, ah, but, but there is a gap, I think, and if we had more virtuous examples of people who have built incredible businesses, again, with a seed in Europe, I think more people will, will not, would not default Uh, to that and think, oh, I could actually build such a business from France or Portugal or Italy.

AI assessment note: “many of the most talented Europeans... just default to building in the U.S.”

Q So in those early days, what were the key lessons that you were learning? Like, what did you start to realize were the right attributes of an app, a piece of software, a company that you might acquire? What were the, what were the things that you were after?

A It's always been the same things on a high level, and that would be Number one, so far we've always focused on digital technology. We haven't bought supermarket chains, nor do we plan to, because you've got to be, I feel you want to stay reasonably within your circle of competence. Ideally, here and there, you want to take a step kind of half outside of it. You need to kind of keep pushing the boundaries, because that will keep your TAM expanding as you expand within the TAM, but I don't think it would be wise, especially as long as the model is Um, works well. It's efficient to take massive leaps outside of the circle of competence just because. So, digital technology, scale. Scale is relative, but we, because our approach is so hands-on, so time-consuming, I mentioned we can, sometimes we radically rethink a business, or at least several components of it. We will do maybe five acquisitions a year. Could be one, could be 10 max. I mean, if it's really a stretch. And each Some more than others will, will really go super deep and rethink the details. And the time investment and the effort does not scale linearly with revenue. So for us to do an acquisition that will bring in half a billion in revenue is not five times as time consuming as one that will bring a hundred million. Maybe it's on average a little bit more time consuming because it tends to be more complicated, but now…

AI assessment note: “It's always been the same things on a high level, and that would be”

Q And so between, um, Evertail and Bending Spoons, what was the, where did the insight come from to be sort of an M&A driven acquirer of businesses rather than building them?

A Evertail was your startup by the book, meaning this idea that probably won't work, but if it works, could be huge and very innovative, like nobody had attempted, as far as we could tell, anything like that before, and so we worked super hard on that project for about three years, and naturally, as you are a start-upper, or when you do something, I think you tend to network with people in a similar situation for a bunch of reasons, and so over time, we got to observe Probably a couple dozen teams go through similar journeys as we did, and through that observation, we saw that, of course, most failed, which you would expect, and maybe three or four had levels of success, and we saw almost no correlation between the teams we considered more talented and more hardworking and those who came out on top, and so we concluded probably, I mean, it's not a huge sample, but probably to go from zero to one, Luck plays a huge role. There are so many factors and variables that you, you know, even if you're a genius and you work your ass off, you're still likely not to, you know, the stars will probably not align for you, um, anyway. Whereas at the same time, we also found that our skills at all the functional things like software engineering, Well, AI at the time, for what it's worth, product design, product management, marketing, although they were still pretty, ah, crude three years later, …

AI assessment note: “we concluded probably... to go from zero to one, Luck plays a huge role.”

Q one. We talked about Evernote, so maybe pick a different one. I'm curious for another acquisition, whether it's Retransfer or Kamud or anything else. Uh, what's the, what's the AI photo sharing one? Remini, which I was just looking at out before we started this morning. Are there other acquisitions that have taught you personally the most about your own process about doing this well? That stand out in memory?

A There was one time where, let's say we bought a product at the peak of, uh, let's call it a viral moment. This is really not applicable to the type of businesses we buy today, but at the time, it was a thing. It's many years ago now. Um, and, uh, and then as soon as we bought it, they basically, that viral wave was reaching and had reached the peak, and, and that completely changed. We thought we had been conservative, but it completely changed our assumptions and led to Drastically inferior returns, uh, versus what we expected. And that taught us to be absolutely paranoid when it comes to the sources of user acquisition. So basically, either we buy businesses where almost all the value lies in existing customers or users, like people, ok, these have been acquired, it's just about now managing them as well as possible. Or if a lot of the value, uh, is predicated on substantial additional user acquisition or customer acquisition, then we need to really clearly understand The drivers of that expected acquisition and make sure that these drivers are things we can predict. For example, we can make pretty accurate predictions of word of mouth rates under normal circumstances, but not under sudden viral moments. We don't feel very confident making predictions of the future rates of user acquisitions through paid advertising, for example. So that was a big lesson learned. Um, another …

AI assessment note: “that taught us to be absolutely paranoid when it comes to the sources”

Q Do you have a favorite example of that so far that you've actually deployed in this work?

A Yeah, of course. I'd say the most obvious one is, um, Um, what we, we call internally our, our, what we call the policy agents. So most companies have a travel, uh, an expense, uh, management policy. Uh, it's generally a document. Sometimes it's well written and clear, and sometimes it's not, but that's, that's, it's essentially a document that you write to drive the behaviors of, uh, people in the company and how they, they, how they manage their, their expenses. Right. In the old world, people make transactions. Before they make the transaction, sometimes they'll go and check the expense report. Sometimes they try to remember it from memory. They'll make a transaction. They'll file an expense report and some manager will try to make sure that some manager or someone on the finance team will try to make sure that the expense that was made by the employee actually abides by the rules of the expense policy. It's very manual. Uh, generally the transactions are missing context. So there's a lot of back and forth between the people enforcing the policy and the people who made the transaction. It takes a lot of time. So we've built our policy agents. Uh, in a way that it has more context about the transaction than most people reviewing those transactions today. So it's integrated with your calendar. It's integrated with your email and knows your expense policy, et cetera, and more c…

AI assessment note: “I'd say the most obvious one is... what we call internally our policy agents.”