You know, when we started the firm, a big idea that we had was that venture capital was disappointing as a product for an entrepreneur. We always thought, wow, a much better product would be give me the network to be confident and the advice I need to run this thing.
Ben, I'm really happy to be back here doing this. I got to be in the same room with Mark earlier in the year, and I'm really happy that you're doing this with me right now.
I'm glad to be here. It should be some fun.
Can we start with your relationship with Mark? Because I think it's a super unique thing where you guys obviously work together, running companies, you've built this firm together. You have a really unique relationship. I can't think of that many examples where I feel like I've seen it in that sort of equal way for so long. Can you talk about it a little bit?
We've been working together 30 years. I would say we're both different and complementary and the same, and not too complimentary, so that helps. He's kind of like we're a little more like relatives than anything else at this point, working together 30 years and so forth.
Are you friends outside of the work context? Like, what's the interaction like in a non-work situation?
Yeah, we're friends, but not like drinking buddies or something like that. We both work so much we mostly talk about work anyway, because we're working together. But it's kind of like the way I would describe it... I want to say this without sounding like I'm making a level comparison, but the relationship comparison that I think is most similar that I know about is kind of the Michael Jackson-Quincy Jones relationship. Mark's more Michael Jackson-like; he's a star of talent that nobody else has. As a firm, we can just put him out there and it's like a magic trick. For me, my relationship with him is like Quincy Jones. I'm certainly not Michael Jackson, but I know enough. Quincy Jones knew so much about music and how to get the most out of somebody that talented. That makes it work together. So I can surround Mark with the kinds of people and ideas that maximize him. And he makes me much better because he's Michael Jackson. You're never going to make Thriller if you don't have Michael Jackson. All the great albums Quincy Jones made, nothing was as big as Thriller because you need that. So that's the kind of relationship. We're very complementary. We're in the same field. I'm an investor, he's an investor. We built companies together. I'm an engineer, he's an engineer by training. But he's different than me, very different, and I'm different than him. That's what makes it work. We're enough the same and enough different. But I'm not saying that I'm Quincy Jones. The great Quincy Jones, rest in peace. I love him. I don't want people mad at me.
Do you think about the firm similarly? Are you on different sides of certain ideas about the firm, and that's what has led to it working the way it has? Or is it like you believe the same things about what it's going to take to win, and then you just have a different daily skill set and set of work?
No, we've had a lot of arguments both on which direction it should go and in what kind of time frame. He has more ideas about things we should do than I do, just because in running it, I try to keep it a little more contained. Although sometimes I'll push us in a direction that he wouldn't have normally gone otherwise, like some of the international things we're doing. It's more common that he's generating and I'm editing. Although I definitely generate some ideas and he definitely will edit some of them. We talk so much that by the time we get to the idea, we've worked through it. But it's a good back and forth. I would say I'm more decisive as a personality type than him. He's more open-ended because he's more of an idea generator, and that helps where I can go, okay, we're committing the flag here but not there.
It's kind of rare in venture that there's a CEO model. To have somebody who's operating as a CEO is kind of rare. At this point, Andreessen Horowitz is big, like 500, 600 people, right?
And your day is management to some extent, or half of it?
600 is relatively small on the scale of things I run. A lot of it is still like I have 25 one-on-ones with entrepreneurs every month, helping them with CEO stuff, helping win deals, doing our international stuff, and working with investors. Maybe management is like a third of my time, but that's probably more than most VCs put on it.
I guess doing that stuff is probably important for you to be able to run the firm anyway. As much as generating returns, you probably also need to do the deals to stay close to the work and know what's happening.
100%. I think you don't really understand the VC business if you're not on board to make investments and understand that hiring has changed a lot in the last 5 years. Offers didn't used to be like this, and if you're running a venture capital firm and you're like, how in the world did they get that many RSUs? When I was a boy, you only got four. You can get out of the loop really fast.
When you think about the way you've decided to structure the firm, you basically have a certain caliber of GP. I got to speak with Martin, who's amazing and could obviously run and lead his own firm. How have you thought about what needs to be true to have people like Martin and Chris Dixon at the company?
Mike Morris had this great quote years ago: the key to running a venture capital firm is to keep the principles from killing each other. There's a lot of truth to that because you have very high-powered, super high-IQ, disagreeable people who are the best VCs. You can find a VC who you don't like, but it's pretty hard to find one who you go, that guy's not smart. Can you be a good agreeable VC? I don't know many. Some are more agreeable than others. To be a long-term, all-time great VC, most of those end up being disagreeable because you really have to think about everything for yourself. Wanting to be liked can be a problem. However, there's a phase of VC that's heat-seeking. You could be a good agreeable heatseeker. In fact, it probably helps because you just want to be liked by all your friends. The best heat seekers probably are agreeable. But the truffle hunters, they're all disagreeable.
Have you ever seen a very good set of results from somebody who is just doing heat-seeking work?
In a period, yeah. Usually what happens is they show up during a boom and then they go away after the boom is over. Partly because they were interested in the chase more than the actual breakthrough new technology. The true heat seekers don't care about that at all. They're just like, wow, that's hot. I'm going to go see if I can get that deal. I know everybody wants that deal. That's a real thing in VC. We kind of came out of that period. We may be going into that period again.
There's a good version of this where some people who would get framed as heat seekers can tell that it's going to be hot. They don't know if it's going to be a good long-term business, but they're like, this is going to be hot by the next round. I can just see the setup. I think there's something interesting there. I've seen seed investors who just seem to have a nose for this is going to get a hot Series A.
Yeah, that's a real thing. I agree with you. It's not so much our business, but it's a real thing.
I talked with Mark about how there's this issue once you're really big: you actually have to be really careful with those because of conflicts. It's kind of not worth it if you're able to win stuff later. There becomes a prisoner's dilemma where you should basically wait to do the earliest round where you're sure it's the winner in a category.
There's some of that in the calculus. For us, we have to believe in it technologically and believe in the entrepreneur. We view ourselves as more mission-oriented. Our mission isn't to get higher returns than the S&P 500, although that's a good side effect. Our mission is to help the best entrepreneurs build the best companies and make us, the country, and the West strong technologically. When we think about it through that lens, we care much more about what it is than what the next round might think. Although, of course, it's a consideration: will anybody fund these guys after us, or do we have to fund every round?
Going back to the type of VCs that you think can be good, you're managing these really good investors. Can you talk about the difference between managing accomplished GPs versus managing execs in an operating company?
It's quite different. Good execs understand the importance of chain of command. They're managers, execution people, process people, as opposed to people generating a lot of ideas. Every good investor is a massive idea generator. They don't necessarily like rules or be willing to follow them. The burden on everything making sense every step of the way is much higher. From an organizational design standpoint, you've got to minimize conflict or it's going to be complete chaos. In a company, there's always cross-functional dependency, and you can live with it through rules, process, or telling people to shut up. In a VC firm, you can tolerate much less of that. If the org doesn't solve the conflict issue, the ability to solve major conflicts one-off will cause a lot of problems.
You're saying interpersonal conflict is worse in a venture firm.
Much worse, both because of the personality types and because you can really wreck each other's businesses. Let's say I'm Martin and I know everything about AI and foundation models. I've gone through every single video model and understand all the nuances. Then someone else who hasn't done any of that work finds an entrepreneur they like and invests in that model, conflicting Martin out. He's going to be beyond mad. He's going to want to murder the person. It's such a bad problem to have conflict like that in a venture capital firm. You have to have respect for the work people are doing and design the firm so that all that hard work gets them to the end they expected, and their own people don't undermine them. The conflicts are extremely intense.
When you see a real conflict, are you like, I'm going to let them figure it out, or are you going to mediate and get to a resolution?
I stand up and figure it. My biggest thing is sometimes they're too respectful of me on that. I'll just resolve it. I need to resolve it.
What do you mean respectful? They're like, Benjamin.
The worst conflicts we've had are ones that nobody brings to me. It could be something very stupid and trivial, but they get so hot.
Is it between GPs that you worry about, or is it inside a team lead sort of org? Do you work out their conflicts for them, or do you say you need to run your own division?
Sometimes there's something inside a fund, but those are pretty cohesive because we design them. Every fund we have runs like a little VC. There's not more than five GPs. It's pretty cohesive, so there's not that much contention. You get contention when things scale and everybody's not talking to each other every day. Then that becomes a problem. Mostly it's cross-fund or cross-functional. It's very important to squash them. They're all kimchi problems. The deeper you bury them, the hotter they get. There is no problem that gets better over time in terms of conflict in a VC. It always gets worse.
What are the things that you need to give guidance on? You and Mark, at a firmwide level, is it about deployment pace, structure, approach to the market? What is the broad intersection point across all the different funds?
Some of it is general principles we believe in. One is we have to make sure we're taking enough risk. That usually comes in the form of evaluating the entrepreneur and the company on the magnitude of their strength. How good are they at what they're good at? Are they world-class? Is this the best person in the world at doing this thing? Not on, oh, the monetization model doesn't make sense, or they don't know anything about go-to-market, or they don't understand accounting. Those things you can always rule out a deal on weaknesses. It's always a mistake to rule out somebody who's truly world-class on a weakness, and it's always a mistake to invest in somebody who's not truly world-class on a lack of weakness. A lot of the guidance is around ideas like that. We're always like, okay, what can they do? Let's focus on what they can do, not what they can't do, and see if that's worth investing in. It's an important psychological thing because we have so many brilliantly analytical people who can find what's wrong with anybody. I always remind them there's something wrong with everybody. You just may not have been able to find it yet.
When you think about how you're deploying capital, it seems to me that you and Sequoia are in one shape: there are many deals, and you've scaled up the number of GPs because a GP can only do so many deals. Your answer to that has been many funds, and you can sort of keep it together. That's one way to scale. Then there's another version with way fewer people and way more concentration, like Thrive, Founders Fund, and maybe Greenoaks. Directionally, you know what I'm getting at.
They do some of both, and we do some of both, but directionally, for us and what we want to accomplish as a firm—making America the strongest country in the world technologically—the concentrated approach just doesn't quite work with that mission.
Why is that? Because we just need more shots to build these companies?
If you look at Founders Fund or Thrive, both great firms, there are whole sectors they're not really in. There are entrepreneurs who could build something great that they're happy to miss, as long as they get the very biggest ones. That's not what we're about. If crypto is going to be important to the financial success of the United States, or how AI interacts with the economy, then we have to be in it. We need to help that succeed. In a sector like that, it's not just funding entrepreneurs; it's helping change the law and get the right policies. Our model wouldn't let us ignore something that important. But if what you're trying to do is say, of the five best companies of the decade, do I have big positions in enough of them? That's a good financial strategy. There's nothing wrong with that. We work with them on a lot of deals together, but it's just not who we are. You have to be who you are.
Is it the kind of thing where over time, there's no reason for somebody doing the broad base of lots of Series A's and B's to do billion-dollar checks into those companies when they're really running far? If you look at it historically, sometimes that strategy of waiting for things to get to a billion-dollar valuation is the best strategy. But in other eras, that set of companies just all suck, and the good investments are in the A and seed round because something new is happening. If you piled all your money in like Tiger Global did in 2021, it didn't work out that well. You were at the end of the cycle. That may work better now. TBD. Obviously, you were raising big funds, and it's worked. On some level, when you start any company or venture firm, you need to have something that you believe that not everybody believes, and that needs to be right. One of yours that you got to early was that venture is going to scale. Can you talk about the thinking behind why scale is so important and why you and Mark have believed it's the dominant strategy?
There are a couple of different dimensions. One is the market of technology companies. Mark wrote a piece in 2011 called 'Software Is Eating the World.' The conventional wisdom in VC at the time was that there are 15 companies in any given year that ever get to $100 million in revenue, and the whole game is getting into as many of those 15 as you can. Why have a big firm if there's only 15 companies to invest in? That doesn't make sense. But what he thought then, and the bet we made, was that if software eats the world, it's not going to be 15. It's going to be 150 or 200. That's what it's become. To get into that many great companies every year, you clearly need to be of a bigger size. There's no way to address the market if you're five guys. The more important part, which was a big rationale for starting the firm, is if you're an entrepreneur, what do you need? You need to be important in the world. How do I talk to big-time CEOs? How do I get in front of big customers? How do I go international? How do I deal with the US government if I'm in AI or crypto? How do I have employees take me seriously? There's a lot of capability you need. If we can bring that, that's a much better product than 'I'm a smart guy, I'll take you to coffee, I can give you some ideas about your product for two months until I'm irrelevant because you've heard all my ideas.' The product of a VC that has a platform and capabilities is so important. Speaking as a former entrepreneur, I needed all that. That part of the idea was really good. The challenge is mechanically how do you build a big VC that's still very good at every part of it?
I want to pull open a couple of those. One is the big brand and the idea that you lend your brand to the company until they're big enough, and then they get bigger and it improves back to the firm. That seems like a big one. These relationships you're talking about are very hard to get. If you're a first-time founder, you don't have the relationships with the government. 'Hey, I just started a company. Jamie Dimon, can you have lunch with me?' Good luck. One of the ones I'm less sure on is the platform stuff. I'm curious to pull that open. What are the platform services that you're most confident work, and what are the ones you're least confident work?
I actually think they all work because we've moved away from the ones that didn't work.
Can you talk about some of the ones that didn't work?
Early on, we did some things where we were publishing general research ideas for the entire startup community. It turns out that works really well if you do it for crypto or AI. We try out all the models and talk to the companies about which model is best at which, and we host models and try stuff out. From a tool evaluation standpoint, we can accelerate you and tell you exactly what's what. Doing that in general across domains ended up being not as interesting. That's getting true for talent as well. A CFO is a CFO for the most part, but a hardware CFO is different from a software CFO. That's easy. But an AI researcher is pretty different from a full-stack engineer. That's a different talent pool. They run in different circles, their comp structures are totally different, so you have to specialize.
The recruiting stuff is probably the number one thing that'll help me recruit. Have you found that you can effectively aim recruiting teams at companies for specific periods of time?
It helps to get the seed corn. A lot of what you're trying to do is get the first three to five people in who are stellar and then have networks to draw from. Who your VC is actually matters a lot. An engineer cares about whether this person is going away, if there's more money behind this, have I heard of them? That's an important feature. Every company has to ultimately get hyper-proficient at recruiting, closing, interviewing, onboarding, and training, or they're never going to be a good company.
You're saying to some degree you can't do it all for them, or they'll never build the muscle.
At some point, you'll end up retarding their growth if you put in their entire team. We've probably helped through our network hire over 100 people for Databricks over time, but they're really good at recruiting. Now it's just that people from our talent network are interested in working there. That's different from the first few people you hire. Those are relationships we have that will introduce you to somebody, and maybe they'll fit, maybe they won't.
How about your view on board seats and board membership? How important is that? You've got some firms that say their selling point is we don't take a board seat, we'll leave you alone. You've got some that talk about a spiritual connection between the board member and the founder. Where do you fall in this? How do you think about what the board member should be?
I probably don't believe in the spiritual connection. First of all, boards are important for founders. The idea that you're going to run without a board after you've given equity to employees and sold equity to people who are not you is the most dangerous idea in the world. If you know anything about securities laws, the only protection you have as CEO from going to jail or getting personally sued is that you run material ideas through the board. That's a massive protection. If I want to give a 2% grant to someone in the company, you have to realize your fiduciary duty to the rest of the company that you're diluting is in question. If you run that by the board, it's all good. You're completely protected. If you just make that on your own, somebody wants to sue you, they're going to win. You have very little defense. The idea that you're not going to have a board is a bad idea. Once you don't own the company 100%, you got to have a board. That's just how it goes. We went back and forth on how much the board actually helps the company. There are a couple of interesting things. Y Combinator...
People actually said, it must have been around like 2015, 2016. They came to us and they were like, we did an analysis of all our companies who had boards and didn't, and the ones that didn't have boards all failed. They didn't do well as a cohort. The ones with boards did much better. Now, there are confounding factors, right? But in their view, the other thing was just in observing the companies, the rhythm of having to tell somebody outside of the company what you're doing every three months or every two months creates internal pressure. It's very valuable. It's just a good organizing principle to keep yourself on track. So I believe in that. And then I do think certain kinds of board members can be very, very impactful at different points in the company. Just with myself, right now if Ali kicked me off the board at Databricks, I'm not sure that they would do much worse. I honestly don't. However, there were times, for example, the Series C I led, the Series A NEA led, the Series B nobody would do the Series C of that company. Pete Sinini and I led that one too, and us led the C in addition to the B and the A on the board. I don't think there's any way that's happening. So that's an existential issue. Second thing, they had an offer to buy the company for, I think it would have netted out to about $4 billion in 2018, 2019, something like that. And had I not been on the board, I think they would have sold.
And I think Ali would say they would have sold. And so they're worth, you know, they're raising now over a hundred billion dollars. So that's real value. And to say that, oh, we're going to do you a favor and not be on your board, because look, when you're in a company, all you think about all day is the company.
You don't have perspective.
So if there's somebody who can help you think through things and has perspective,
That can be worth the whole thing, right? So it's not a nothing. I think it's not a... I've seen board members that had no value the whole time they're on the board. That happens all the time, for sure. So I don't want to overstate.
Well, there's also two types of ways to be impactful. There's these high-value moments that are discreet, crisp situations where you can say X was going to happen and I made Y happen and Y was better. And then there's the daily engagement type of board member who's talking all the time, interviewing candidates, the CEO is upset about something, it's somebody to talk to. Do you think that kind is actually less valuable in some ways, or do you think about those similarly differently?
So there's the board and the work of the board, which is governance number one, as I alluded to in the beginning. And then there's certain points and certain things where board members can really matter. I think most of the work that I do is more... I'll have a monthly call with a CEO just to talk through the things that they get stuck on, the things that are causing them to hesitate.
Does this work now only at this point, or earlier in your career did you need to do more of the daily, higher engagement model stuff, and only now can you do this version?
Part of what I'm getting at is I think there's this idea in venture that you can only be on like eight boards before you collapse. And then I was talking to Martin, who's on all these incredible companies. The founders say he's awesome, and he's on way more than eight or ten boards.
