The OG Defense Check, with Ross Fubini (Managing Partner of XYZ VC)

The secret of starting early stage venture funds — the one thing you have to do to convince people to give you money — is that you know something about a network that nobody else knows about. Palantir is lots of things to lots of people. What it actually worked at is Karp, the CEO, would open the door at some large company, and what you do is you chuck in a bunch of technologists and basically be like, 'Hey, go find a $50 million, hundred million dollar problem worth solving.' And so it was this unbelievable crucible that was unknown at the time. It was super vivid if you were on the inside. It just ended up being correct because those people were exceptional and they went on to choose a bunch of really interesting businesses to build.

Thanks, Mo. Thanks for having me. I really appreciate it.

Thanks. Thanks. You know, it's funny, Mo. So I was on — I was on the list a few years... I was on the list last year, and I sent a note to my mom about it, like, you know, and my mom immediately replied. And she basically said like, 'Yeah, you should think about contacting everybody ahead of you on the list and finding out what they do that made them more successful so you could figure out how to improve,' which is good pressure. We improved a little bit this year, but the funny phenomenon of being on again is really you end up just realizing how exciting it is to keep working with the founders you got, and then, you know, the long journey of venture is just, you know, venture and pressure.

Yeah. So really it's two points of dumb luck that made XYZ, as a firm, and then as successful as it is. The first part of dumb luck is, as you said, so I'm an engineer by training. I worked for Netscape and then ran a division of Symantec on the product side. So I started some companies and sold them. The first lucky thing I did was a guy named Mitch Kapor, and Mitch was the founder of Lotus 1-2-3, the visual spreadsheet company, gave me my first venture job. And being able to go do early stage venture in companies like, you know, we did Twilio and Optimizely and Clever and Formlabs — that was just incredibly fortunate. As an entrepreneur, an engineer that moves over into venture, just getting that opportunity and getting to learn, you know, in the last, you know, 15-plus years, all stems from that moment and that person and learning in the mentorship structure there.

The second thing, ironically, at about the same time, I introduced a friend of mine to a guy that used to work for me. I introduced Shyam Sankar to Alex. These were two of the people that ended up being the founding executive team along with Bob McGrew at Palantir. This matters because I'll tell you, then, years later in 2017, when I was starting XYZ, Mo, the secret of starting early stage venture funds — the one thing you have to do to convince people, foolishly, to give you money — is that you know something about a network that nobody else knows about.

Yeah. So Palantir is lots of things to lots of people — a lot of lens in the press or in the stock, all these things. But Palantir originally really was a technology-powered consulting business at the earliest stages. We've wrapped that accurately in forward-deployed motions, working on the world's hardest problems, doing custom solutions on top of Foundry — all that is true. But when it started, this is the thing that I knew, was two things. Number one, incredible at hiring nerds. It is the place that CMU, MIT, Stanford — you know, great people wanted to be at, and they were really good at getting those people in for a bunch of funny reasons.

And then secondly, what it actually worked at is Karp, the CEO, would open the door at some large company and what you do is you chuck in a bunch of technologists and basically be like, 'Hey, go find a $50 million, hundred million dollar problem worth solving.' And so it was this unbelievable crucible of entrepreneurial development, mostly for technologists. And that really — that was a secret that was unknown at the time. It was super vivid if you were on the inside. So it was such a strong observation, and then it just ended up being correct because those people were exceptional and they went on to choose a bunch of really interesting businesses to build.

Yeah, a couple of things. So first, you'll start at the end and then go back to the beginning. So in the end, XYZ — you know, we're a little over a billion and a half under management, as I said, over 130 companies. Really fortunate because of the folks we got to work with. The firm's been very successful, which is exciting. We do mostly stuff in the government — that's Anduril, ILK, and Cape and Turbine One. In the enterprise space, businesses like Sardine, which is an anti-fraud platform, Rootly, which is an SRE company, Scribe. And then we do quite a bit in a whole other part of industries that haven't been transformed fully by technology, and that's everything from tax to healthcare, etc. We think there's like real businesses that are really being accelerated, many of them even more so by AI. So that's us — billion and a half. We get involved very early and then we aggressively dollar-concentrate into those partnerships. That's us — small, tight team, mostly in the Bay Area and into New York. And that's us in a snapshot.

The interesting part about the first fundraise was, oh my god, it was so hard. To answer your core question, it really was one of these rare — mostly in venture, arguably in life — moments where people were flat wrong. Usually you're telling a story that is backwards-looking; you kind of refit the narrative, right? In this case, at XYZ, it really was basically: we are going to go very deep with a small number of networks, we're going to be really involved with those companies, and we're going to dollar-concentrate into them. That was the actual pitch, and that ended up being true, and really fortunate. In fact, all the mistakes we make were not concentrating more money earlier. But those are the right ideas. The hard part on the venture side is this insight of: first funds are all about unique networks. I can only see that in retrospect. When I was raising the first fund, Mo, I would be like, if you got 100K, I would pitch you as hard to get those dollars as I would pitch Yale to get their dollars, and everyone in between. So it was very difficult to do because the network was so unknown, and it's really fortunate now that those investments ended up being correct, and that made every subsequent venture round easier and easier, or funding round easier and easier.

He was a mathematician, Guido Fubini. He worked with Einstein.

Yeah. So first, the history part. You know, all of us are standing, you know, in the — whether the shadows, the ripples behind our family. My family is full of these incredibly smart, talented people. My great-grandfather, as you said, very famous mathematician. My grandfather was assistant secretary of defense under Kennedy, spoke like a dozen languages. He was truly a brilliant guy. You know, folks in the audience might even have taken classes from my uncle who teaches at Harvard, my mom, you know, my aunt is the dean of veterinary medicine at Cornell. But the problem is in math there's this idea of regression of the mean. So like a next instance goes closely, so we're trending low. So first, like, expectations — so all the work is to try and, you know, offset this great history.

That said, the part that I really do have and hold dear in relevant to the investing is firstly, just fundamentally a technologist, and I really believe in technology's opportunity to change industry, society, all those things. That would go all the way back to my earliest days at Netscape and being passionate about the internet being a thing before that was obvious. Obviously the investments that we've done are really about technology transforming government, transforming societies. It just happens you can make an absolute truck ton of money in the process because the government is the largest spender in the world. The Department of War is the largest spender within there. Health and Human Services is arguably even larger depending on how you cut it. So that's all things where we spend a lot of time.

The second thing that's sort of interesting — so we're fundamentally a technologist. Mo, something that's changed for me on the investing side, the 'what to invest in' has shifted, where the main thing that I'm now looking for is where a market is changing to the benefit of a company.

So an example of this — and you've written about this — would be, say, Apex Space. So, Apex Space, Ian Cinnamon and Max founded a company, and the idea of it was there's going to be a lot more vehicles, payloads up in space. That was even arguably a few years ago when they started it. The further argument was that there was an opportunity to standardize those systems — that basically, you know, Mo, you're putting your satellite up, you'd rather have a standard satellite bus so that you knew your product was going to go up on time, on budget. And so that was the underpinning there. So it wasn't a technology belief. It really was a belief that there was an opportunity for standardization to come into the market. And if that was true and you owned the standard and you provided the standard satellite bus, you could build a really big business. And so that's an example. When we're doing the investment, it's no longer just these people, this technology that's necessary. We can talk about Apex's strengths and weaknesses, but this macro shift that's moving the market to their benefit is a must-have.

Yeah. So it actually is a good example. So we were Anduril's first investor, along with Founders Fund. I got the very first pitch ever for the business, and I actually originally got it from two of the four founders — so it was from Trae and from Grimm. And the pitch basically fit this thesis. The pitch at the time was: there is an opportunity to build products for the government. Well, actually that's it. Everything within Anduril is built into that idea. Because, first of all, products meant repeatably sold versus services, high margin, and then the opportunity was that nobody was creating those products. Right now, you can get Lockheed or Boeing or Leidos to go build you a one-off thing and pay you a 7% margin, but you couldn't build products the government wanted to buy from a startup. That just did not exist.

The second part that was interesting was just the team itself. So, shortly after this, Brian Schimpf — I've been involved as Brian Schimpf's exec coach for a long time. Brian was coming from Palantir. He joined as CEO. The most notable thing about Anduril's trajectory — this isn't the original pitch, but it was required — was these immediate ingredients were the combination of skills to both take something to market as well as build that product.

Two answers. One is I knew it was going to win. I just knew it. And I knew that because the team was that good. I'd seen the Palantir pathing, and I had already believed, for a bunch of familial reasons we already talked about, in the size of the market opportunity. That said, when I did the investment, I thought winning was going to be like a $3 billion company, a $5 billion company, right? Like even that was like an impossibly large idea. Some of that is because of the amount of transformation that Anduril has brought to the industry. Some of that has been how much excitement came into that market because of the Ukrainian war, because of the further acceleration of procurement under the Trump administration, and now I think because of technology plus those motions, I now think we're building a company worth hundreds of billions of dollars.

So first: reliable and reasonable — this is not how I would describe the government buying. If anybody here has done a capture activity, they are like, 'No, no, no, no, no, ever.' Because not only is the budget wonky — so that's going to interact. That thing that was going to close last year might close next year. The technology assessment that you've been really successful at might get reconstituted in a whole new set of requirements. And the reason it's called capture and not sales is because, Mo, I got to get that money before you and you bastards at Leidos steal it from me. So I think, first of all, it's truly a war of ideas and a war for dollars to get done.

Yeah, 100%. And which — just to pick a thousand issues with that — is like, the government already has that capability. They don't buy that other capability. Like, if you've already got that thing sorted — I've got my autonomous vehicle from Forterra that does this, you don't — the government does not buy another one. They're like, 'Yep, I checked that box. I'm building a whole maintenance system behind that.' But it just goes to the challenge of it, and not naively, but that there's just so much to understand that government — you know, technology chasm crossing.

To go back to your comment about Anduril's early days — what then the issue was that nobody, I mean, damn near nobody, believed you could build a company that sold products to the government from Silicon Valley. It just — I mean, the number of pedigreed people and Sequoias and Greylocks that would turn down even a meeting because they all had experience. The number of LPs I pitched — 'Give me more money. We were a small firm then. I want to put more money into this business.' The smartest LPs called former SecDef chiefs directly, and those people told them not to invest. That is how unsexy, uncompelling, uninteresting that world is. You know, if there had been even a noise about it on Twitter, we would have had those awesome Keith Rabois, 'Boy, don't buy Anthropic at four billion' things. We don't even have those because people don't even bother tweeting about it. They just quietly hung up the phone, right?

Everybody's all the money. It's like ghost pirates.