Invest Like the Best with Patrick O'Shaughnessy
· David George, Patrick O'Shaughnessy
We're in this funny environment where late-stage SaaS cloud, we would look at a company and it's like, "Oh man, if you're not 70% plus gross margin, you're not really a SaaS business or cloud business, whatever, and that's going to be a knock and people will trade you differently. And that's when you get valued as revenue versus gross profit or whatever." Now, it's like a badge of honor to have low gross margins, because we're like, "Oh, at least people are using your AI products." We get these pitches and they're like, "I have an AI thing, and I got 75% gross margins." I'm like, "Well, no one's using the AI stuff then. That doesn't really seem like an AI product to me." So we give a little bit of a pass on that. The expectation is the cost is going to continue to go down. [00:00:00] S0: Just inference cost. [00:00:00] S1: Inference cost is going to go down over time. I mean, there's so many existential questions about market structure that will predict inference cost. But the history of technology would suggest that it's going to go down over time. The cost of inference has gone down at the same time that reasoning happened. And so token usage has gone way up. So you haven't yet seen any improvement in gross margins, but I think, over time, that's likely to happen. [00:00:00] S0: Do you basically just not care? If a company has 0% gross margin, for example, but the revenue growth and the customer love and all this stuff, the pull is all there. Does it round to, "We don't care."?
· Patrick O'Shaughnessy, David George
[00:00:00] S0: Do you basically just not care? If a company has 0% gross margin, for example, but the revenue growth and the customer love and all this stuff, the pull is all there. Does it round to, "We don't care."? [00:00:00] S1: There's a big difference between having 30% gross margins and 70% gross margins. So 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, the model providers, that it's going to settle out where these businesses are probably higher margin businesses. I think they'll be lower margin businesses than SaaS businesses, 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 the business model earlier, is probably so high that it's fine. [00:00:00] S0: How much do you care that the way the product behaves and the way it's distributed is truly singular and different than competitors versus just the best of a class of company? [00:00:00] S1: There's so many ingredients to the best of. There's a foundational point, which is every great company either has unique product or unique distribution. The best companies in the world have both. The best companies in the world have such unique product that it leads to unique distribution, but if you don't have either of those... [00:00:00] S0: What's your favorite example of that?
· David George
[00:00:00] S1: I'll use a recent one, that the product is so good that people just naturally have gravitated to it is Cursor. And again, maybe in the fullness of time, that'll get harder. But GitHub is a great example of this. I'll tell a funny story about GitHub too. So GitHub was so special of a company that for a long period of time, they never actually talked to customers. So the first time I ever met GitHub, they were like, "We've got to tell you this, this is so awesome, we sold to Walmart and they're paying us 400,000 bucks and no one ever talked to them on the phone." We were like, "Wow, this is an incredibly magical product and an incredibly magical market. Just imagine if you had talked to them on the phone, what would they have paid you if you'd just called them on the phone? They'd probably have paid you $4 million." So unique product that leads to unique distribution with a founder that wants to optimize the situation. So the AI founders, I'm not the one involved with Cursor, but Michael is a very special founder and his team, they recognize what they have, and then they are aggressively pursuing the enterprise at the same time. So that's a really good combination, where you have unique product, you have great product people love, that leads to some uniqueness of distribution. And then, you can build on that advantage by saying, "Hey, we have all this bottoms-up use, we're going to go sell enterprises." And so, a big part of what we do as a firm is we help to facilitate customer introductions, new business, we call it our go to-market function, they're referred to as EBCs sometimes. And we get notes after every one. And this is the most fun thing in the world of AI, because we get these notes, and in the case of Cursor, every single time, it's like, "Immediately to POC, immediately to POC," proof of concept, whatever, "Immediately to POC," and, "Immediately to full sale deal." And you can see that, that's actually incremental data for us in making decisions, but you can see it, it is magic when it happens. So Martin led the A of Cursor, one of my partners who leads our infrastructure fund, and after one of these emails, he chimed in, and it's a big list, it's like a hundred people on the list or something, he wrote, "Product fucking market fit." So now, we're like, PMF is now PFMF.
· David George, Patrick O'Shaughnessy
So when you see that, that's unique product, that's unique distribution, and you have a founder, founding team or full set of employees who really wants to optimize it, that's magic. [00:00:00] S0: What are the trade-offs of the way that Andreessen is structured? No firm is perfect. There's choices for how you have structured and nested the team, lots of different groups, leaders of groups, like you. What are the negative parts of the trade-offs for how Andreessen is structured versus a more monolithic structure or something that was just different?
· David George
[00:00:00] S1: Our strategy for scaling is pretty well-covered. But effectively, we think scale allows us to bring more power to the entrepreneurs and give them a greater chance to be successful in the market. That's the fundamental thesis behind the scaling for us. And with more resources, you can bring more resources to bear for the entrepreneur. So for us, when I joined, every single Monday and every single Friday, we used to sit in a room together, all of us, and we'd hear all the pitches and then we'd have long meetings to talk about each of them, all as a group. So Dixon was leading our crypto fund and we'd have bio fund pitches, and we'd all listen to all of them and then we'd all debate. And then, we realized, at a point, that was not the optimal use of time, Dixon weighing in on a bio investment and vice versa probably doesn't make sense, and you could extrapolate that out to a bunch of our investment processes. Ben and Marc decided to decentralize the firm and put more power down into the investing teams that ran each investment fund, and the reason behind that is twofold. One, we thought it would allow us to have better expertise around the table. If you're only just fully deep in infrastructure or applications or American Dynamism or crypto or bio, that's an advantage. It's both an advantage in making decisions, but also an advantage in go-to-market with the entrepreneurs. And then, secondly, if we are going to scale, you can't scale an organization with 25 or 30 decision-makers around the table, it's too hard. You can't make a trade-off between, should we put an incremental dollar into a bio fund investment or a crypto investment, or how should we think about reserving this versus that, it's too hard. So we shrank the size of decision-makers by doing this to a smaller group who's in charge of their own funds, and so far, that's working really well, and I think that's mostly a function of the fact that our early-stage folks, they're really good and we're all really collaborative. The only trade-off that we have at the growth fund is selfishly, that process that I described where we all sit around the table, it's valuable for me. It's good for us to have access to all information at all times, because we sit across all of our early-stage funds, the way we operate is we invest across all of our sectors.
· Patrick O'Shaughnessy, David George
[00:00:00] S0: What percent of the investments you make did the firm have a prior investment in? [00:00:00] S1: It's a little over half. And then, if you take the number of investments... So if you just do it by dollars, it's a little over half that are preexisting venture investments. And then, if you add the dollars that we're investing in preexisting investments that were preexisting growth fund originated investments, follow-ons, it's something like 70%. So 70% of the dollars that we're investing, we get deep knowledge on the companies. I call it game film. I talk about game film all the time, it's so important. When assessing an investment, when assessing a founder, game film is not just numbers, it's how has the founder done this. [00:00:00] S0: How do you do reserving in the growth fund, is it materially different than elsewhere? [00:00:00] S1: When we first started the growth fund, I was like, "Scott, zero reserves, let's do it. Every single dollar is going to have to be scrutinized, literally every dollar." It turns out that's not really practical, you need to reserve a little bit. So we reserve a tiny amount, and this is for small follow-ons where our participation is important, but we're not a lead. We do zero reserving for large investment amounts that we think we're going to make in a company, because I think that would lead to lazy decision-making. We'd say, "Oh, well, we reserved for it, let's do it." [00:00:00] S0: So you just treat it as a new investment next time?