--:-- Erik Torenberg: We met with one firm, Ben, you might remember. One of the partners said, "Venture capital is like being at the sushi boat restaurant. A thousand startups come through and you just meet with them. And then every once in a while, you kind of reach down and you just pluck a startup out of the sushi boat and you invest in it." And I was like, "Oh my God. The sushi there is typically not great."
--:-- [Intro music fades in and out]
--:-- Erik Torenberg: What does it take to build a venture firm from scratch? And how should it evolve as the world changes around it? I recently sat down with Marc and Ben for a wide-ranging conversation on the origins of a16z, the evolution of the venture capital industry, and the structural choices that have shaped the firm—from platform to federation and beyond. We talk about everything from how a16z got started in 2009 to how we think about platform, media, governance, and why venture is more barbell-shaped than ever.
--:-- As it happens, this conversation took place during my first week, so it was the perfect moment to reflect on where the firm has been and where it's headed. This episode originally aired on *The Ben & Marc Show*, which you can follow for more candid conversations from inside the firm. Let's get into it.
--:-- The content here is for informational purposes only, should not be taken as legal, business, tax, or investment advice, or be used to evaluate any investment or security, and is not directed at any investor or potential investors in any a16z fund. Please note that a16z and its affiliates may maintain investments in the companies discussed in this podcast. For more details, including a link to our investments, please see a16z.com/disclosures.
--:-- Erik Torenberg: Hey, everybody. Welcome to another episode of *The Ben & Marc Show*. I'm Erik Torenberg. I'm Andreessen Horowitz's newest general partner, and this is my first week. Lots of exciting things planned for the future of the firm, but we thought that this would be a great opportunity to talk about a bit of the history, about the conversations that led you guys to start Andreessen Horowitz. When did you know that you guys had to do this?
--:-- Ben Horowitz: I think it was a conversation on AOL Instant Messenger, if I recall it correctly. We had sold Opsware to HP and had been out of it for a little while. We had started doing some angel investing and that kind of stuff. We had an angel fund called Horowitz Andreessen. That's not even a joke, too. So we were doing that, and we were talking about what might each one of us do next. And as I recall it—and Marc might recall it differently—he said, "You know, venture capital is so underwhelming. It's a great product for investors, for LPs, but it's a very mediocre product for entrepreneurs because you get almost nothing. You get some money and then a smart person. But that smart person, they see you once a quarter. They don't know much about what you're doing, so their value kind of diminishes to zero in about three or four months."
--:-- We thought, it's so hard to build a company. Somebody ought to be able to do something better than that. And I said to Marc, "We could start a firm and we could call it Benmark," which was a pun on Benchmark.
--:-- Marc Andreessen: I don't think I ever got that until just now.
--:-- Ben Horowitz: Yeah. So that was the beginning of the conversation. And I was surprised because Marc was into the idea. I think he had the idea separately. So it was just one of those things.
--:-- Marc Andreessen: I'm absolutely amazed and flabbergasted that venture capital even exists. My first 22 years of life, I had no idea. I had never heard about it growing up. I never heard about it even in college at Illinois. And I came to Silicon Valley and my first business partner, Jim Clark, was like, "Yeah, we start a company, we go raise money from these venture capitalists." And I was like, "What's that?" It completely blew my mind that there were these people literally scouting for crazy startup entrepreneurs to start these things, and they would give you money when you had nothing. I was like, "Wow, that's amazing."
--:-- The fact that the field exists is amazing. VC in its modern form started in the 1960s. I look back at the history of these guys—Don Valentine, Tom Perkins, Pitch Johnson, Bill Draper, Arthur Rock. These guys for me are legends because the fact that they could go out and source a Bob Noyce to start Intel, or Steve Jobs to start Apple, is just amazing.
--:-- I had quite good experiences. Ben mentioned the issues with the field, but Ben and I had the chance to work with two VCs at the very top of their game when their firms were on top of the world. That was John Doerr when he was at Kleiner Perkins in the 90s, and later Andy Rachleff, who was a founding partner of Benchmark when they were on top of the world. In general, we got a lot of value from those guys and considered them partners. They helped us build good businesses.
--:-- But over the years, what happened was we learned from those experiences. And also, Ben and I had become active angel investors and advisors and mentors to a new generation of founders. This was in the 2000s. Just so people understand the setting: after the dot-com crash in 2000, there had been an angel and venture boom in the late 90s. Then after the crash, almost all angel investing went away, and a large amount of VC went away. It was a full-on depression for early-stage tech.
--:-- By about 2004, when Ben and I ramped up our angel investing, the whole industry was down to maybe a half dozen angel investors. It was small. We knew all of them. Angel investors had so much power that there was this scandal, "Angelgate"—angels meeting together and fixing prices.
--:-- Ben Horowitz: The Angelgate, yeah.
--:-- Marc Andreessen: So this is a famous—
--:-- Ben Horowitz: Well, we all talked to each other. So there was something to that.
--:-- Marc Andreessen: Then tech started to take off again in the late 2000s. TechCrunch at the time, Michael Arrington had built it up to be the main online news source for startups and venture investing. One of Michael's most remarkable things was he was at some dinner and he somehow cracked the code that there was a back room where all of the prominent angel investors of that era were basically sitting around this round table. Ben and I were not there. Michael Arrington kind of walks in, and he's a journalist. At least he described it as these massively guilty looks on everybody's faces. The accusation was that they were colluding, teaming up to see if they could keep valuations low, which is a no-no in business. But a significant thing about that is that it meant there were enough angels to actually fill a table, right? That was when they were like a dozen of them as opposed to just a half dozen.
--:-- Ben Horowitz: Yeah.
--:-- Marc Andreessen: Anyway, Ben and I just started working with what turned into dozens of founders. We ended up being very involved because we had raised venture before and run companies. We helped a lot of startups in that era raise venture. We helped them meet the venture firms and understand how to negotiate the deals. The other thing that happened was we would get called in when they would get sideways with their VCs.
--:-- Ben Horowitz: It happened a lot.
--:-- Marc Andreessen: A lot. This became one of the main things people were calling us on: "I'm in some big fight with my VC. He wants the money back, or he freaked out at the board meeting, or I hear rumors this firm is shutting down." Or the VCs would call us up and say, "This founder is nuts. Could you please talk to him and try to get him to do the right thing?" So we ended up in this sort of arbitrator, coach, judge mode, helping patch these things up. I think, Ben, because you and I were angel investing at that point, a big part of it was, well, if we're going to end up doing that anyway, if we just showed up with the checkbook, we could short-circuit the—
--:-- Ben Horowitz: Yeah, yeah, yeah. Right. We wouldn't have to fix all these problems.
--:-- Marc Andreessen: That was a little bit of it as well. There were just very few people at that point in the venture capital world who had built any kind of company that had gone to any kind of complexity or was worth anything. It was just a different kind of background of people. So the ability to relate to founders—really relate to founders—was a little bit missing.
--:-- Erik Torenberg: Okay, so you decided to start Andreessen Horowitz in 2009. You come up with a $300 million fund. Talk about what the strategy was going into it, or how you were planning to differentiate from the outside in. You guys were very loud. You had this platform approach. Talk about behind the scenes, how you thought about differentiating.
--:-- Marc Andreessen: By the way, the big VCs at that time seemed overwhelmingly invincible. They were giant, long-lasting businesses. The whole industry had been around for 50 years. Some of these guys had invested in every good company—if you look at some of the things Sequoia had invested in along the way, it's a spectacular set of companies. So we were trying to figure out how to challenge the status quo.
--:-- One of the ideas we had was we would do a lot of angel investments in addition to venture investments, which was unheard of at the time. We would start out in this complementary way, and then eventually we'd build enough reputation where we could start doing the A's ourselves—that was how we pitched it. We took it around. We visited a lot of our VC friends, and many of them said it was a really dumbass idea and we should definitely not pursue it. It's been tried before and it didn't work, and so forth.
--:-- The other idea we had was what I alluded to earlier. We got a lot of this from our friend Michael Ovitz from CAA. What if the firm wasn't just a collection of partners? What if it was more than that? What if, rather than paying the partners a lot of money—or in our case, we didn't pay ourselves any money—what if we took all that money? Because it was a lot of money. Even on a $300 million fund, it was a lot of money. What if we took that money and we built a platform? The purpose of that platform was to give an entrepreneur the confidence and power of a big-time CEO like a Bob Iger or a Jamie Dimon, who could literally pick up the phone and call anybody at any time. Why should I be CEO of this little company, even though I've never managed anything myself? All I did was invent the product. Well, because like Bob Iger, I can call anybody from the president of the United States to the CEO of FedEx, and I can get them on the phone. So we were like, what if we built that capability for our companies?
--:-- People said—I think the criticism was—it's been tried, it'll never work, you guys are stupid. That was the polite version of it. Those two things were really the original idea that we pitched. In fact, if you look at our original deck, that's how we pitched LPs.
--:-- Erik Torenberg: And the first fund works, right? I believe you put $50 million into Skype, and there's a big markup there in that acquisition. There's Instagram.
--:-- Ben Horowitz: It's actually $65 million. But $15 million of it was generously given to us by Silver Lake for participating in the deal. But we invested $65 million, yeah.
--:-- Erik Torenberg: Yeah, so Instagram, Tiny Speck was turned into Slack. So the first fund is a winner.
--:-- Ben Horowitz: Yeah, Okta.
--:-- Erik Torenberg: Okta. Did you know, fund one, what your long-term vision was? Would you say, "Hey, we're going to start with $300 million and then we're going to scale and be one of the biggest firms in venture"? Or what were you thinking was the future at the time?
--:-- Ben Horowitz: The thing that Marc and I knew from our experience in starting companies is it is just as hard to start a small boutique thing that means nothing in the world as it is to build the world-dominating monster. It's no more amount of work to do the latter. So we were never interested in anything but the latter. We had zero interest at all in building a little venture capital firm that was like a beta to the big boys. We always thought this was the start of doing something much bigger and much more important. How we got there, we certainly didn't have it all mapped out from the beginning, but the ambition was always there.
--:-- Marc Andreessen: A lot of this comes from the fact that we'd been running companies. If you're running a company, it's a product company in full competitive battle with other companies, and you're going through the wars that companies go through. You naturally think in terms of strategy, ambition, industry structure, the economics of the business, the competitive position, evolution over time, marketing strategy, unique selling proposition, differentiation. All these things that any business operator thinks about.
--:-- To their massive credit, a lot of the original VCs who built the firms back in the 50s, 60s, 70s were actually operators. Tom Perkins and Gene Kleiner had been operators. Don Valentine, Pierre Lamond had been operators. The founders of Greylock had been operators. It was very natural for the first generation to think in those terms. But by the time we entered the field, their successors, for the most part, had not run businesses. They had grown up as professional investors, inheriting businesses that other people had built. There was just this fundamental difference in mindset.
--:-- By the way, the formula was working very well for them. They were very happy to sit on Sand Hill Road and wait for the deals. I've got countless stories in this. We met with one firm, Ben, you might remember, a very prominent firm. One of the partners said, "Oh, venture capital is so much fun as a business. Venture capital is like being at the sushi boat restaurant."
--:-- Ben Horowitz: The sushi there is typically not great. That was the first thing that jumped out at me. I wasn't sure if we went to the same sushi restaurants.
--:-- Marc Andreessen: If you haven't been to a sushi boat restaurant, you sit there and these little sushi boats go by on the conveyor belt, and you just pick up pieces of sushi. He said, "That's just what it's like. You just sit here on Sand Hill Road, a thousand startups come through, and you just meet with them. Every once in a while, you kind of reach down and you just pluck a startup out of the sushi boat and you invest in it." I was like, "Oh my God." Complacency. Entitlement. The hair on the back of my neck went up. I was like, "All right. Soft target."
--:-- We were so oriented with a startup. All you do is work, and you're focused on the work. If you aren't doing enough work, you think of other work that you could do that might improve things. It was such a foreign idea that you would be oriented around doing no work, literally sitting at a sushi boat restaurant and having a great life and playing golf or whatever they did. That was inspiring. Like, okay, we can do it better. Building a company was so hard, and any additional help would be so appreciated. That was how we always thought about it.
--:-- Ben Horowitz: Yeah.
--:-- Marc Andreessen: By the way, one more story. Around the same time, we met with another VC. Like I said, we'd been running companies, dealing with investors, and then we'd been running public companies. Ben was the CEO of a public company. When you're running a public company, you've got investors, you've got your investor relations team, but you've got hedge funds invested in your company. When you meet with them, you don't even know if they're long or short your stock. Often short. Half the time, you're giving them ammunition they're going to use to try to smear you and drive your stock price down. It's absolute bedlam when you're running a company with shareholders.
--:-- In venture, it's different—or at least we thought it was different—because your investors, the funds called limited partners, are institutions like endowments, foundations, and sovereigns. They invest on these lockups—10 or 15 year lockups. They put the capital in and sit patiently and let you do your thing. We were like, "Wow, dealing with an investor who is locked in to be with your company for 15 years sounds like the best thing in the world. This sounds amazing. These are super smart people running these endowments, people like David Swensen and Ann Martin." We were like, "This is going to be great."
--:-- Then we met with a very famous, prominent, longtime VC. He said, "Boys, the part of the job you're going to hate the most is dealing with the LPs." He said, "These people are not smart." He said, "The way that you do it is you treat your LPs like they're mushrooms. You put them in a cardboard box, you put the lid on the cardboard box, and you put the box under the bed and you don't take it out for two years." He literally said that. It was such an insane thing to hear because we treated hedge funds that were shorting us better than that.
--:-- But then, when we went out to visit the potential LPs and pitch them, that was actually the thing that made me know we had made a good choice by starting the firm.
--:-- Ben Horowitz: They had very interesting things to say. They knew a lot about the industry and a lot about investing in general. Dave Swensen, who Marc mentioned, wrote the definitive book on how endowments invest. When they invested, they were so interested in us. It's a nice thing in life when anybody takes any interest in you. The LPs did 30 or 35 reference calls on both me and Marc. Every single one of them did. They learned a lot about us. They really got deep on it.
--:-- Funny enough, one of the funny things about the firm is I think we're the only firm in Silicon Valley who has this. We have a two-person key man thing. Normally, as long as one person is intact, there's no vote on whether the fund continues. But in fund one, both of us had to be there because they were like, "You guys are both flawed. But when you're together, the flaws go away." They had gotten that deep on us. It was cool.
--:-- Erik Torenberg: Say more about that. How do you guys complement each other? What were they getting at?
--:-- Ben Horowitz: There was Netscape and LoudCloud. With Netscape, Marc started that company when he was 21 or 22 years old. He was literally a kid. So he had some things in his reputation. The stuff that I couldn't do when I was 22, I can do now. There was some of that. And then there was the same thing on me. LoudCloud got into absolutely horrible trouble. We burned through a stupid amount of cash. There were definitely negative things. But it was interesting because both companies had very good outcomes. I think how the legend went was somehow, between us, we could figure it out. I don't know if the criticisms were right. Maybe they were. And I don't know if the solution was correct. But it was just a fun thing that they had gotten so deep into our backgrounds that they insisted that be in the LPA.
--:-- Erik Torenberg: Talk about how you guys have made it work—divided and conquered, or your working style, whatever you could share about that.
--:-- Ben Horowitz: We're co-founders and we work very closely together on the strategy and the direction of the firm. But the CEO position is a chain of command position. That's me. I'm the CEO of the firm in that sense. Marc doesn't try to override chain of command decisions. It's not his thing. And then Marc does things that I can't do. He's a much bigger celebrity. I always say he's a little bit of a magic trick. People in the firm call him Marc GPT because he knows everything about everything. He has very unique things that he does.
--:-- Marc initially recruited me and then said, "Hey, Ben, Erik, you guys figure out the details."
--:-- Marc Andreessen: Yeah. So that's a good example. Marc had been on this thing that the world has moved, and the way we market the firm and think about the media hasn't changed nearly as much as the world has moved. We need to bring somebody in. And he had you in mind. So I was like, "Okay, good." I listened. I had been on the show, so I knew who you were. I went back and listened to a lot of the Turpentine stuff. I was like, "Yep, that seems like a good idea." And then it was on me, in my CEO job, to put the thing together.
--:-- Erik Torenberg: Marc, maybe just give us a couple minutes on how the world had changed. We'll do a whole separate episode deep dive on it. But maybe just preview what was the main change you identified—like, "Hey, the world has changed from a media perspective."
--:-- Marc Andreessen: A lot of my thinking on this is from a book from our friend Martin Gurri that he wrote back in 2015. I think it's called *The Revolt of the Public and the Crisis of Authority in the New Millennium*. The world really did change. How information flows through the world really did change, not just with the arrival of the internet, but specifically with the arrival of social media.
--:-- All of us who grew up over the last 70 or 80 years grew up in an environment of primarily top-down media. There were these major forces in broadcast TV, cable TV, newspapers, and magazines where editors, publishers, and reporters wrote all the stories, and it was everybody else's job to read them and keep up. Now, we've moved to a world that looks completely different, where everything is peer to peer. Hierarchy to network. Centralized institution to decentralized network.
--:-- That is happening throughout the economy and throughout society. There are good and bad things to it, and tons of arguments to be had about it. But it is happening. In the new world, if you're running a business or a movement, you're not going to do it through the traditional method. You may still participate to some extent, but you're going to primarily tell your own story and go direct. You're going to have your own relationship with your constituents, your fans, or your customers.
--:-- In some sense, that sounds like a truism and a cliché. But I think there were a couple of tipping points where it really started to happen. One was around 2015, because social networking hit mainstream and smartphones hit mainstream around that time, which is when Martin wrote his book. But I think it's really only been in the last five years when almost everybody—let's say everybody under the age of 70 and a very large number of people over the age of 70—has shifted from top-down media as their main source of information to social media as their main source of information. It is relatively new to live in this world where information really does flow differently. It's a fundamental shift.
--:-- As a firm, we always had a big focus on marketing and telling a story. We did that primarily through the old centralized channels from 2009 to probably 2017 or something. But really since then, it's been at least as effective, or more effective, to do it the new way. It's the old William Gibson thing: the future is already here, it just isn't evenly distributed yet. Everything I'm saying, people can nod out. But most companies have still not adjusted to this. Most politicians have still not adjusted to this. Most entertainers have still not adjusted to this. Most sports leagues have still not adjusted to this. It's very important that we continue to do it, and I think it's also important that we set an example for our portfolio companies.
--:-- Ben Horowitz: A lot of it has to do with the apparatus. From a company standpoint, you know what you do. We know how to help entrepreneurs. A product company knows how to build their product. And then it's like, okay, now you've got to get your message out. How do you do that?
--:-- Ben Horowitz: And then the apparatus that gets your message out, all the people, all the kind of tools, all the channels are oriented at least partially in the old world. And so, it actually is in some ways much longer to adjust than it is for the individual consumer who goes, "Oh, there's just better stuff over here."
--:-- Erik Torenberg: Yeah. And Mark, also talk about the shift from corporations to individuals in terms of where brands went and who people want to hear from. Not to say there isn't a role for the corporation, but talk a little bit about that for the corporate brand.
--:-- Marc Andreessen: Yeah. So rewind history a little bit. This sort of decentralized media environment that we're entering is not new. It's actually very old. And correspondingly, the centralized media environment we all grew up in is not the historical norm.
--:-- The way we think about centralized top-down media today is basically an artifact of the period of the 1940s through essentially the 1970s. Before the 1940s, you didn't have top-down media in the same way. If you go back to the 1930s or before, you had a much larger number of newspapers, a much larger number of radio stations, a much larger number of fly-by-night publishing operations, pamphlets and so forth.
--:-- If you go back even further, one of the most interesting things to study is the American Revolution. The time of the colonies in the 1760s through the 1790s. I've been rereading some of this stuff recently. Basically, the media environment of the colonial American era was a lot like today's social media environment. You would have 15, 20, 30 little newspapers per city. They would occupy every micro-sliced niche. You talk about echo chambers or whatever. They each have their own little echo chamber.
--:-- The founding fathers would write all these columns and essays. They would fight things out by writing essays, and then they would publish them under pseudonyms. You have characters like Benjamin Franklin or Alexander Hamilton that would literally have a dozen or two dozen pseudonyms at a time. They'd actually get into fights with themselves. They'd have their pseudonyms actually fight with each other. Ben Franklin used to set off arguments against his different pseudonyms to really litigate out an issue and to drive newspaper sales.
--:-- But really serious stuff also, like the Federalist Papers, which were the explanation of the new constitution in 1789. Hamilton and Madison wrote those under pseudonyms. This idea of the internet anon is an old idea. The idea of a pseudonym is an old idea. The idea of self-publishing is an old idea. The idea of pitch smash mouth battles with very little centralized control over what people say.
--:-- If you read about how Hamilton and Jefferson, and then also Jefferson and Adams, had these absolutely extreme battles. They had their own pet newspapers and it was just an absolute level of smash mouth politics. I would say even more extreme and deranged than what we have today, which people can't believe. If you read about the election of 1800, it was maybe more extreme than any election in my lifetime in terms of slandering. So polarization is the norm on every conceivable front.
--:-- Unfettered conversations were the norm. Anonymity was the norm. Rumor, scurrilous accusations were the norm. The Overton window being wide open was the norm. For people who want to read about this, the best book on this is called *Infamous Scribblers*, which was the name for journalists in those days.
--:-- The point is, this sort of centralized media thing that we've been living in, that we grew up in, is a historical aberration off the norm. It's a consequence of technology changes, a consequence of mass publishing, mass media, mass radio, mass television, mass newspaper, that only started in the 1940s.
--:-- Correspondingly, everything that we think of as corporate branding is an artifact of a specific point in time, the 1940s through the 1980s. All of brand marketing, corporate brands, corporate messaging, corporate crisis management, all these playbooks they teach at business school were very specific to a time and place that had a very small number of centralized media outlets with tremendous influence and control.
--:-- Why does the corporate brand exist? Why does a Procter & Gamble brand or any of these brands exist? It exists because if you have centralized media, information is going through a very narrow straw. There's very little bandwidth to get something on a TV, or very little bandwidth to get something in the newspaper to get it to consumers' attention. You kind of had to wrap up everything about a company into a single word and a single image. Then through advertising, you would just pound that over and over again, trying to get people to remember it. But that's because that's all you could do.
--:-- If you open everything up, and everybody can publish, and everybody can debate, and everybody can be present, and then you have individual influencers with 200 million followers and all this stuff, all of a sudden you have a completely different method of communicating with an audience that can be much more based on personality. Authenticity, transparency, and then personality. There can actually be a human being.
--:-- It just happens that because your audience always consists of people, people relate much more to other people than they relate to a corporation. As an individual, am I going to feel a stronger emotional affinity to a person who I follow, or to some disembodied corporation with an office tower in New York City? If the communication bandwidth is there where I can interact with both of them, of course I'm going to have a lot more affinity for the people.
--:-- I'm sort of a radical. I think the whole idea of corporate brands is kind of on their way out. As a concept, it just doesn't make sense in the new media environment. Correspondingly, the terms people use these days like influencer marketing and so forth, but the parasocial relationship is actually a really interesting one. Sort of one-to-many personal relationships. I think so much of how this is going to work in the future is based on relationships with individuals.
--:-- Obviously, this is happening. Everything I'm describing is what's happening in the entertainment industry and it's happening with consumer brands. You've got Kim Kardashian with multi-billion dollar businesses being direct marketed online. Many people are doing this, many politicians are now adapting to this. This is happening, but I still feel like it's underestimated. If we project forward 10 years, most people are going to think about the people they relate to as opposed to the companies they relate to.
--:-- Erik Torenberg: It's very interesting how you bring that up, Mark, about there being no centralized medias and no corporate brands pre-1940s. As a kid, I always was surprised that I knew more entrepreneurs from pre-1940. I knew Thomas Edison and Henry Ford and J.P. Morgan. But who were the entrepreneurs after that? They weren't. They were just corporations. You didn't know actually who ran any of those things. It would leak out slowly, but it wasn't really a thing. Now we're getting all these celebrity CEOs again. That idea is reemerging, which is fascinating.
--:-- Marc Andreessen: Yeah, that's right. People can't believe it. Before 1930, you would just go to the store, or it was just the corner store. You would buy a pan and they weren't branded. Maybe it was Joe's store, but it wasn't like... The consumer brands didn't exist in the modern sense. To Ben's point, to the extent you knew any business at scale prior to 1930, they were almost all named after their founders. It was the Ford Motor Company.
--:-- There's this whole school of psychology. I think it was Freud's son-in-law, Edward Bernays, who was the father of public relations, a new field in the 1920s when radio and newspapers took off and centralized media started to take off. They created this whole psychological theory of creating abstract brands for the reasons I described. Which is very linked to the methods of political propaganda that became very successful in those days.
--:-- It is amazing to me. There were hundreds of years of industrialization before that and modern economic activity before that where those things essentially didn't exist. That's why I'm so confident in pegging all this to technology shifts. The thing that shifted how we think about companies happened as a consequence of the shifts in communication technology. Correspondingly, if the communication technology unwinds, which is what's happening, then you're actually going to go back to the future. There are more data points to support that every day.
--:-- Erik Torenberg: Is this something you guys had internalized in 2009? Is that why you called the firm Andreessen Horowitz when every other firm was going with something else?
--:-- Ben Horowitz: No. That was a different thing. When we were raising the money, it was 2009. It was a difficult year to raise venture capital. I think there were only two new funds raised that year. The number one objection we got on the fund was, "You guys are very successful entrepreneurs. What's going to stop you from going out and quitting and just starting a company? We're going to be left holding the bag." We had no plan to do that. So we got the idea, one easy way around that is just name the firm after ourselves. Then they'll know that we're going to be tied to it forever. We did that. Then I had the idea that since nobody could spell Andreessen Horowitz, we should have this A16Z thing. That was the name of the firm. Immediately all the competitors said that we were egomaniacs and narcissistically insane because we named the firm after ourselves, which we just ignored. Maybe they have a point.
--:-- Marc Andreessen: Yeah, it's kind of true.
--:-- Erik Torenberg: Well, the irony is that you're still running the firm 16 years later, still as active as you were beforehand, whereas a lot of other folks have retired.
--:-- Marc Andreessen: That is also true.
--:-- Ben Horowitz: It worked. It did tie us to the firm.
--:-- Erik Torenberg: Is it as simple as, you guys have had billions in distributions, you don't need to be doing this anymore. Is it as simple as, "Hey, this is your baby, this is where you have the most fun"? What's kept you going far after you needed to, going at this pace?
--:-- Ben Horowitz: The firm always had a mission. The mission of Andreessen Horowitz was never, "Let's make a lot of money." Both of us had enough money for a normal person to be happy in life before we started the firm. That was never the thing. It was always, could we make it much easier and better to build great companies? And with those, could we make those companies better? What possible activity could either of us have that would be more important than that?
--:-- One thing Mark and I both share is that maybe the single best thing that you can do to improve the world is to build a company that delivers some product that improves the world. That is actually the thing. It has a better impact than any kind of activism or political activity or anything else. It's just literally making things that make the world better. Then doing something larger than yourself where you bring a lot of people together to do that, and they all grow and improve their lives through it. What could be better than helping people do this single best human endeavor possible? Neither of us ever thought there was anything we wanted to do with our time that was better than that. There's no reason to stop because we don't have any better ideas. This is the best idea.
--:-- Marc Andreessen: Mark told me he got the story about Larry Page. I think if I understand correctly, Larry Page says, "I see no better use of my money than giving it all to Elon Musk to build more tech companies."
--:-- Ben Horowitz: As a philanthropic idea, for sure.
--:-- Erik Torenberg: One other idea I wanted to bring back to the idea of people as corporations. It's not only the CEOs. It's like building a cinematic universe. It's the CEOs, but it's also the surrounding characters. It's Chris Dixon, it's Catherine Boyle, it's Martine Cassato, it's Alex Rampell. You guys have done a phenomenal job of building stars and a collection of people.
--:-- Marc Andreessen: We're not really a company, we're a firm. Those hyper talented people we were able to recruit, it's just a platform for those people. We're two of them, but it's not that hierarchical. Everybody is kind of doing their thing, but in a common context with a common culture and a mostly common set of investors. It functions more like a team than a normal hierarchy.
--:-- It's great because if you look at the top people, Martine Cassato, Chris Dixon, Alex Rampell, David Ulovich, that team is better IQ-wise, capability-wise than the executive teams of Meta or Google or Apple. In a way, they're all the boss and they all act like the boss, and that works. That's been a nice outcome of the platform.
--:-- Erik Torenberg: Talk about how you guys developed this idea of a platform, because most firms don't have that, didn't have that, or you guys moved to this sort of federated model. Talk a little bit about the evolution of the firm and how you guys figured that out.
--:-- Ben Horowitz: It came in two pieces. The first thing was when we started the firm, the history of venture capital, if you had done a back test on it, you would find there were never more than 15 companies in a year that would ever make it to $100 million in revenue. The technology industry, that was the general size of it. That's the amount of new technology the world could absorb in those days.
--:-- But Mark had an idea which he wrote up in 2011 called "Software is Eating the World." The idea was that every company that was going to be worth anything was going to be a technology company because software could make anything so much better. So there were going to be not 15 companies, but 150 or 200 companies.
--:-- The result of 15 companies meant the optimal venture capital firm was like six or eight people going after those 15 companies. Each one gets two and you've got a monopoly. There was no need for it to ever be bigger. As a result, they set up their organizations with what I'd call shared economics, but also shared control. That shared control made sense if you're only going to be six or 10 partners, because if you're never getting bigger than that, you don't have to reorganize. You don't have to make difficult management decisions that people are going to disagree with.
--:-- We knew software was going to eat the world and we were going to need to be way bigger than six or 10 partners. We were going to have to be able to reorganize, decompose the problem, set up the organization in a way where very smart teams of people could work independently and address the different facets of the industry. You can see it with American Dynamism and infrastructure and apps and crypto and bio and so forth.
--:-- We never had shared control. We always had centralized control. We got that advice from Herb Allen, who was super helpful in us understanding why that would be important. Also, Mark's father-in-law, the late John Arriaga, was very clear: if you're going to run something, eventually there's going to be conflict. There are going to be these issues and you've got to have control. It's not important until it is important, and then it's the only thing that matters.
--:-- With that control, we've been able to reorganize, reimagine the firm, and address every single vertical. The people who know American Dynamism need to know everything from rare earth minerals to rockets. There's no way those same six people are going to know everything about crypto. It's not even possible. These fields are too deep, not only the technology, but the whole entrepreneurial ecosystem. You need separate teams to address these separate very large markets. Before, you just needed a person on that. You could have a person on crypto would be fine, or a person on AI would be fine. That's never going to work again.
--:-- Our ability to field a whole team against that and restructure things, to say, "Okay, you were doing consumer internet, that's not going to be relevant in the next 10 years," these kinds of things are very hard to do if you don't have control.
--:-- Erik Torenberg: Those two examples, crypto and American Dynamism, are also interesting because these are examples where you guys helped create the categories. I believe you're the first big venture firm to have dedicated crypto and AD practices. You created a firm that can be adaptive to new theses, new ideas, new trends, and build firms against them.
--:-- Ben Horowitz: Mark kind of identified that early on. One of the things he used to say when we started the firm is, "Venture capital is a young man's game." What he was really saying is the technology is always changing. The people who know the new technology best turn out to be often new people. In many venture capital firms, once whatever they exploited runs out—they did network effects and consumer internet, and they were amazing at that—but when that stopped being the thing, they didn't get to the next thing. We were able to get to the next thing.
--:-- We're always watching for the next thing. As soon as we see it, we have such brilliant people. Chris Dixon saw crypto. We're like, "Chris, go get it." David Ulovich actually saw Catherine Boyle, who saw American Dynamism. Catherine was like, "This is a very important thing." So we just go do it. We can do it because we don't have to repurpose our old people. We can build a whole new team. We can change the organizational structure.
--:-- Erik Torenberg: In an offline conversation, we were talking about how some firms look the same as they did 30 years ago from a structure perspective. The world has changed. Firms need to change to meet the evolving needs. The stuff has gotten so much more complex. There's been this great complexification. A generalist firm could have been able to cover the entire landscape, but now no individual can have deep knowledge on all the fields: bio, crypto, all the fields we cover. That's one great example of how the world has changed and that leads to a need in venture firms to change as well. Are there other examples that come to mind around how the asset class has evolved or should have evolved to meet the needs of the world changing?
--:-- Ben Horowitz: It's probably changed more since we started the firm than it did in the whole history before then. There are so many ways. As Mark said earlier, angel investing became a real category. The public markets have become very difficult and dysfunctional to the extent that OpenAI just did a giant raise in the private markets, which I don't think they could have done in the public markets. The fact that you can raise more money in the private markets than the public markets in one shot speaks to the expansion of the private markets to deal with the fact that the public markets are just not a great environment anymore for companies. That changes venture capital because we're the private market. Our market just got much more enormous.
--:-- As you said, media change, how you go to market. We were actually the first ones to market a firm in venture capital. Margaret Wittenmark did an amazing job of creating a brand for a firm that popped up out of nowhere. That had never happened before. But then the way you market it changed entirely, as we just discussed. The world is changing really fast just in general. I think AI, the way we work, the way we operate as a firm is changing very fast due to AI. What we can automate, how big a reach, how many entrepreneurs we can know. All these things are very different now.
--:-- Erik Torenberg: Let's double click on the brand point. You guys in 2009 came out and you made a lot of noise. Some people had different opinions, but everyone had an opinion. You thought deliberately about building a brand in a new way. You guys and market and team, you crushed it. Talk about what that strategy was and what perception it was. What was it like as you were building out the brand?
--:-- Ben Horowitz: It was a conversation Mark and I had. Mark is kind of one of the ways we understand each other. Mark asked me, he said, "I've been studying the history of venture capital. I've been trying to figure out why they don't do any marketing. It turns out the industrialists and venture capitalists, the Rothschilds, J.P. Morgans, were sometimes funding both sides of a war. Any kind of publicity might get them killed. That just kind of carried through to modern venture capital so that the original rationale for not doing it was no longer really valid. They told themselves other things like, 'We're very humble, so we don't market,' which is always a rationalization for laziness." He said, "What do you think? Should we market it?" Sometimes when Mark asks a question like that and I already know what he thinks and I haven't thought about it that much, I just go, "Yeah, of course, let's market it." That was kind of that conversation.
--:-- He had worked with Margit prior at Ning and thought super highly of her. He said, "Let's talk to Margit, let's see what we can do." We spoke to her. This was hilarious. You have to remember this is the days when magazines were a big deal, which they're not so much anymore. When we launched the firm, she said to us, "Do you want to be on the cover of Fortune or Forbes?" We were like, "Fortune, of course." That's exactly what happened. That was the beginning of it.
--:-- Erik Torenberg: You guys were able to accrue amazing people early on. Talk about what it was like to get one of the first big partners. You've got some partners like Chris who've been here over 12 years. What was it like in terms of how you thought about recruiting the early partners and landing them?
--:-- Erik Torenberg: I would say that's one of the things we got wrong in our thinking. Well, we got right and we got wrong. One of the things we got very right was the first person we hired was Scott Cooper, who we knew super well and had worked with for years. It was brilliant and really fundamental to building the firm. He's recently joined the presidential administration in the White House, but he was invaluable and fantastic. He didn't want to join when we started the firm because he was worried we wouldn't be able to raise the fund. So we raised the fund and then we hired him. He was employee number one.
--:-- Then the second idea we had was that only founders or CEOs were allowed to be general partners. The reason for that was a little bit what Mark said earlier—we were counter-programming what had happened in the industry where you had a lot of people who were smart but didn't understand founders. So we wanted everybody in the firm to understand founders. But that profile turned out to be not perfect in many ways. We hired some really great people. One of the early people is Peter Levine, who still works with us now.
--:-- Then we kind of started to relax. The first thing we relaxed was, okay, maybe the company you had to found or be CEO of didn't have to be that great. If the company did okay, that was fine. That gave us permission, which was controversial at the time, to hire Chris Dixon. One of the things Mark and I recognized early was that Chris Dixon was a far better investor than either of us. That was an indication that maybe we were too rigid in our criteria. That started to open it up quite a bit.
--:-- Erik Torenberg: I want to go back to one of the unique insights you had. Going back to Mark's "Software is Eating the World" piece, the idea was that there were going to be more winners and those winners were going to be much bigger. There are a lot of implications from that—you'll raise bigger funds, you'll have this decentralized or federated model, you'll be able to invest at higher valuations if these companies are going to get bigger and bigger. It feels like that was something you guys saw relatively early that other firms, even later-stage firms, then got on board with.
--:-- Marc Andreessen: Yeah. The big thing on that is this really important transformation that's happened in tech that went kind of unremarked on as a pattern, although you started to see it in the early 2010s. Up until roughly 2010, if you make a list of all the big winners in tech over the preceding 60 years, they were basically all a form of a tool company—a technology tool. They would build personal computers or microchips or operating systems or databases or routers or web browsers. But fundamentally, they were building components of a computer system. They would sell those tools to either consumers or businesses, and the consumer or business would figure out what to do with them.
--:-- Ben will recall when he first started the firm, one of my early investing things was Nova Verticals. You look at that list and you're like, basically, the big winners have all been these big horizontal tech companies building general purpose tools that many downstream industries pick up and use. But historically, if you had a tech startup that was focused on a vertical, it just meant you were a small tools company. The classic example: I am a tech company and I want to be in the boutique hospitality industry, so I start a software company that makes booking software for bed and breakfast hotels. Such things existed, and they were just very tiny companies.
--:-- Fast forward to 2010. The internet really started to work. Broadband really kicked in, a bunch of things catalyzed. What you started to see was that the tech companies that went into vertical started to get huge. Probably the first two that made this clear for me were Uber and Airbnb. Airbnb said, how about we not only build the booking software for the bed and breakfast, but how about we run the entire service? The entire booking engine, the entire search engine, all the transactions, all the customer service—the entire end-to-end experience. Same thing for Uber and Lyft. You could have a small boutique software company doing taxi dispatch software, or you could build Uber or Lyft and actually build a giant transportation network with drivers and riders and money flowing through.
--:-- More recently, a company like Anduril. Our companies for many years have sold many parts of computers and software into the defense department and defense contractors. But Palmer Luckey came along and said, let's just build a defense contractor. Let's build a direct competitor to the big defense primes and actually build defense systems. Tesla is another one. Instead of building embedded power management software for cars, how about just build the car? SpaceX. We keep going.
--:-- In the last 15 years, if you do that same list again, many or most of the big winners have been companies that have gone into a vertical. But what they've done is they've tried to eat the entire vertical. They've provided an end-to-end experience with everything required to service that vertical, often in direct competition with the incumbents. Anduril competes head-to-head with existing defense primes. Uber competed head-to-head with taxi and limo operators. Airbnb famously competed head-to-head with hotels. We got extremely angry about that. Netflix competed directly with cable channels and movie theaters.
--:-- So basically, you're going to have more and more of these companies that use technology to insert into an end market and then try to take that end market. Good news: those companies can get to be gigantic. If you crack the mother lode, like Netflix has with entertainment, or Tesla has with cars, you can build a company that's maybe multiples in size of that entire industry as it existed before. The challenge is that those kinds of companies are much different than historical tech companies. They're full service. They're much more complex. They have a lot more moving parts on the operating side. They require a different kind of discipline from the management team. In a lot of cases, they're operating in regulated industries with a completely different political dynamic. And they're going up against entrenched competitors who certainly have no intention of just turning the business over. I think that's been the defining theme of the last 15 years in the Valley—the evolution from just tools companies to what we used to call full stack, just do the whole thing.
--:-- Erik Torenberg: The one knock against a16z that I've heard over the years is, "They think of their firm as a product, or there's like a machine," as if that's not a great thing. If a startup said, "We have no moat, I'm just a smart guy," you'd say that doesn't feel super defensible. It doesn't feel like you've really built something of power. Yet when people think about their venture firms, they sort of run them the opposite of how they want their startups to run. They really think about structural advantages, durable advantages, network effects—all the things they want their startups to have. It's been an interesting contrast.
--:-- Ben Horowitz: There's a kernel of truth in the critique. The kernel of truth is that at the end of the day, as an entrepreneur, your VC is the only personal touch. The reason for that is you're going to have somebody on your board. You're going to have somebody you call at 4am when the world is caving in. You're going to be dealing with that person in high tension situations. You want to really rely on them. You want them to really know what they're talking about. You want them to have throw weight in the industry. That personal relationship really does matter.
--:-- I don't think what would work is just trying to not provide that and instead just provide a machine. But what I think works incredibly well is to provide that *and* provide the machine. And the team. The thing that really distinguishes what we do from what we experienced is we always had a person. But when we tried to reach through that person to the rest of the team, they were like, "Not my company. I'm not making that introduction. I'm not doing that." Whereas almost on a daily basis, we'll have a company run into something and they'll go, "You should talk to Joe Morrissey. He dealt with that sales issue over here. You should talk to Ben. He knows how to deal with a crisis like this."
--:-- In fact, we just had one this week. One of our partners said, "This seems like a bad crisis. Bring in the guy who lived through all the crises." That's me. I can really help because I understand what to do, but I also understand what it feels like. So much of that kind of thing is having a deep understanding. In the firm, we understand almost every situation you would be in, and there's somebody who's a great expert who will be there in a flash, even if that's not the person on your board. That is probably the thing I'm most proud of in the organization. People always get their money's worth from that perspective.
--:-- Marc Andreessen: This is the big industry structural transformation that we think has taken place. This is one where we did predict it. We have been talking about it for a long time, and I think it's really happened. It's played out over the last 15 years and it's still playing out. There's this pattern in industries as they mature. They often start with what you would call a strategy of being in the middle.
--:-- A classic example is retail. Once upon a time there were department stores—names like Sears and JCPenney. You would go to the department store. It would have a pretty good selection of products at a pretty good price. Growing up, that's where we would always go shopping. By the time you hit the 80s and 90s, the department stores stopped working. For the most part, they've gone under at this point. They got replaced by competitors that were not in the middle, but were on the far end of one side or the other of the spectrum of strategies. That's why we call the outcome the barbell.
--:-- The department store was replaced by two sets of companies. First, high scale: Amazon, Walmart. What you get is incredible selection at absolutely fantastic prices. But it's a very machine experience. You go to Walmart, the shelves up to the ceiling. It's a specific thing. That wiped out a huge part of the department stores. Then the other thing on the other side was specialist boutiques. For the thing you care about the most—whether that's fashion or jewelry or consumer electronics or candles—you go to the boutique. You go to the Gucci store to buy your scarf. You go to the Apple store to buy your iPhone. What the boutique offers is a very narrow selection at a very high price, but what you're getting is a very specialized experience. You're often getting the personal touch. You go into a wristwatch boutique and it's just great. "Would you like some champagne? Let me get you a comfortable chair. Here's all the espresso. You want to stay late? Great, we'll lock the doors." You get this very personal touch experience.
--:-- What happened was the department stores just died because they didn't offer either one. They didn't offer scale and they didn't offer the boutique personal touch experience. If you look at the history of business, as industries professionalize and mature, many of them go through this. This is a big theme of the TV show Mad Men because they were right in the middle of that. There's a certain point in the show where they sell their midsize ad agency to McCann, which was one of the big scale players. They got frustrated there because it was just a big machine, so they went and started their own boutique.
--:-- It's happened to law firms. It happened with Hollywood talent agencies—Michael Ovitz catalyzed this in the 70s and 80s. It happened in banks, investment banking, commercial banking. It happened in hedge funds. It happened in private equity. We've seen this pattern happen over and over again, but it hadn't happened in venture capital.
--:-- When we entered the field, what we observed was that they were all department stores. The venture capital version of the department store is six to eight general partners with a 300, 400, 500 million dollar fund, doing the sushi boat strategy. Sitting and waiting. No website. God forbid you ever tell your story to anybody or make yourself visible. You sit on Sand Hill Road and wait for the deals to come through. It had run that way for a long time. It was this cartel, self-referential thing.
--:-- Our bet when we went for scale and built out the kind of teams Ben described—the machine that results from it—was that the death of the middle was going to happen. The barbell was going to play out. There was going to be an opportunity for a handful of firms to go for high scale, but only a handful. You don't get 50 at high scale. You get a bunch, but not that many, because scale economics kick in. Then what would happen on the other side is the rise of the angel investor and the seed investor. We had been part of that. We had been on that side of the barbell.
--:-- This was part of the transformation that had happened in venture. The original venture firms in the 50s, 60s, 70s were first money in. They were the first check for a company like Intel or Apple. By the time the 80s and 90s rolled around, they were no longer the first money in. They were off in the second or third check after the angels and the seed investors. We put two and two together and said, "What's going to happen is this field is going to bifurcate just like every other field. We're going to go for scale, and we're going to encourage the seed investors." We've been very actively trying to invest in seed investors and help them. We've been trying to be very friendly with them.
--:-- The structural question posed is: what's the point of having a department store? Of having a midsize firm? The answer is there's no point. For the same reason there's no point to a department store, there's no point to the midsize firm. For the reason Ben described: they're not the first money in, they're not at scale, and they don't have any depth. At the end of the day, there's really no value proposition if you have access to seed investors on the one side and the scale platforms on the other side.
--:-- Ten or fifteen years ago, we would say this and everybody would get mad because it sounds like we're predicting everybody's going to die. But sitting here today, this has really played out. Many of the midsize firms that I grew up with are gone. In some cases, they're gone because they failed. But in a lot of cases, they're actually gone because they succeeded. The partners made a lot of money. Then at some point, the rationale for being in business started to fade away. Maybe they had to start working a little bit harder and that wasn't fun. So they just folded up shop.
--:-- The LPs have correspondingly adapted to this. If you talk to the LPs now, increasingly they are focusing capital either on the scale platforms or into this very specific kind of early stage seed angel strategy. Their interest in funding the department store equivalent of the VCs has really faded. I view this as a very natural evolution. This was destined to happen. It'll happen in many other industries in the future. It's a process that plays out in response to customer demand. The customers of venture firms are the entrepreneurs on the one hand and the LPs on the other. If they both want this change to happen, then it's going to happen. It's disconcerting to be on the wrong side of this, and it's an adaptation process for people to figure out that this is happening. But I think now it's pretty clear.
--:-- Erik Torenberg: That's well said. That's one example of how the asset class has evolved. Let's get into other ones. One is that as your thesis has played true—software has eaten the world—there's been more demand on the LP side to get into the space. Much more money has flooded the space, which means more venture capital firms, which means more competition. When supply is constrained, VCs have the power and founders are clamoring to get on the conveyor belt, pretending not to care by not having websites. But when there's an explosion of venture firms, founders are the ones picking. VC firms have to change their tune. You guys were early on to it, but it also changes the types of LPs that want to be involved. Talk more about how the asset class has evolved from more capital flooding into the space, or any other changes that emerged from it.
--:-- Ben Horowitz: Yeah, I'll say a couple of things. First of all, we used to have this discussion with our friend Andy Rachleff, who's kind of the master of venture. He's a co-founder of Benchmark and actually taught venture later at Stanford. Very analytical on the topic. Extremely thoughtful. We had this discussion with him of like, "Wow, money comes whipping in and out of venture. These dynamics really change who has the upper hand. Is it the founders or the VCs?" Andy made this very interesting observation.
--:-- He said that for as long as he had been in the field, going back decades, venture has always been overfunded as an asset class. There's really never been a time in which venture has been underfunded, maybe a little bit in the extreme crises like 2009. But generally, venture is overfunded. His rough back of the envelope math at the time was that it's always overfunded by like a factor of four. These days it's like a factor of 40 or 400. The Sequoia guys are always famous for complaining. Anytime they give an interview, they talk about how there's way too much money in venture. They're always trying to discourage people from doing any venture. They're trying to talk the LPs into stopping the money flow because of more competition.
--:-- The question is: why is it always overfunded? Andy said it's a consequence of the broader financial landscape. What are LPs? They're large pools of institutional capital being invested for many reasons, but a lot of it ultimately is retirement funds. They're large pools of capital that need to generate a certain level of return over the next 50 or 100 years to pay for people's retirement. To do that, they need to hit a certain level of return. The nature of the modern economy is population decline. You have a lot more older people, a lot fewer younger people. So there's this fundamental issue: as a steward of institutional capital, how do you generate the long-term returns you need in an environment where that's not so easy?
--:-- You invest in stocks and bonds, and you often still can't get the math to pencil out. You're not going to hit your return target. Then there's this asset class called venture capital where sometimes it works and sometimes it doesn't. But when it works, it blows the lights out. When venture capital works, it's the top performing asset class. There are individual venture capital funds that have been absolutely spectacular, driving a lot of the return for an entire institutional portfolio.
--:-- The way I describe it is: venture capital is never the majority of the money in an institutional pool. But it's like the cherry on top of the sundae. It's the small position. But if it works, it might make the entire formula work. Then you look at how many pools of capital there are like that out there. How many LPs are there? The answer is a lot. Basically, all the LPs read the Swensen book, which describes how to run these institutional capital pools. They say, "Dave Swensen says you put X percent in venture capital. But Dave Swensen says the key is you only invest in the top venture capital firms because venture is a feast or famine business. You only want to be in the top 10 percentile of firms."
--:-- Then what they do is they go out and talk to the firms. They find out they basically can't get into most of the firms they want to invest in. So they develop a theory of how these other firms are actually in the top 10 percent. If you ask LPs who they think are the top 10 percent firms, they often have very different lists. Part of it is a function of maybe they've sniffed something out. Part of it is just because they have to allocate the money. So they convince themselves that there are undiscovered gems out there. As a result, they just overfund the asset class.
--:-- Too many LPs managing too much money leads to too many VCs, leads to too many startups getting funded. That leads to the phenomenon that founders experience: I start a company, and not only do I have three venture-backed competitors, I often have 30. Andy's point is that's just an artifact of the world. We are the tail on a much larger dog. The dog is large scale institutional money flows. Venture is a rounding error in the global financial system. But it's one that's just prone to be overfunded for very long periods of time. Andy said until there's a new approach to investing these large pools of capital, we should assume this process persists over a long period of time.
--:-- Marc Andreessen: Would it be better if the amount of money was equalized to what it should be relative to the opportunity set? For people like us, yes. For the world, it would be worse.
--:-- Ben Horowitz: That's the other thing. If you had less money in the space, would entrepreneurs be able to take as many swings? No. Should I have the arrogance to sit here and say that we're going to invest in all the great companies and that we're not going to say no to people we ought to be funding? We make that mistake all the time. If you're going to have an asset class that is overfunded, this probably is the one to overfund. There's a societal surplus of all the swings that entrepreneurs get to take that they wouldn't get to take if the sector wasn't overfunded. Some of those work. Founders come out of nowhere, raise money from no-name VCs, and end up building huge successful companies. On a societal basis, I think it's a form of dysfunction that maybe is not optimal financially, but on a societal basis, I think it's probably positive.
--:-- Marc Andreessen: Yeah. What could be better in terms of wasting money than taking money from people who have too much and giving it to people who want to change the world and make it a better place?
--:-- Marc Andreessen: I mean, it seems like... and are building a company to do so. That seems like a pretty good idea.
--:-- Ben Horowitz: Yeah.
--:-- Marc Andreessen: The other thing I'd add is that venture capital is a little bit unique from our point of view, in that it's the only asset class where the top managers tend to persist for decades. If you look at stocks or bonds or anything else, the pickers—because they're all kind of picking against the same thing and they all have equal rights to invest in everything—there's some amount of randomness that puts somebody on top, and then they're no longer on top the next decade. But in venture capital, the top firms often remain the top firms for a very, very long time. The reason is the best entrepreneurs will only take money from the best venture capital firms.
--:-- So if this was the NFL draft, which I think is today, we'd have the number one draft pick every single year, despite already having the best team. That doesn't matter if there's too much money. If you always get to pick first, you still can win very consistently. That's sort of what happens. It's a great system from our perspective. Good for the world. Good for us. We love it.
--:-- Ben Horowitz: Some people will say things like, "Oh, there's too many founders or too many people want to be founders," as if it's an already efficient market and there aren't people out there in the world who... That's a bit. It's the best thing in the world for people to try to do something larger than yourself and try to make the world a better place, and get people along the ride with you. Everybody's got a great purpose and they're all working hard, and maybe there's a great outcome for them in the world. Why wouldn't you want to fund as much of that as you can? I never understood the argument that there's too much venture capital. It's crazy. It can never be too much.
--:-- Erik Torenberg: When did you guys realize that you were entering the pandemic? Like, when did you realize, "Hey, this is really working"? What was the biggest inflection point in a16z history when you felt you reached that point?
--:-- Marc Andreessen: Very early on, we realized we could win what we thought were very high quality A rounds from top tier VCs. As soon as we could do that, we were like, "Oh, we could be top tier. We could definitely be top tier." In our original world domination plan, we thought that was going to take 10 years or whatever, but it happened really early on, right in Fund One. By the time we got to Fund Three, it was in full effect. It just happened much faster. We're in a whole other world now than we were then. But we knew it—as soon as we could be in those days, Kleiner, Benchmark, Sequoia in a deal, that was a very clear indication that we could be top tier.
--:-- Ben Horowitz: Look, I think this is the advice I'd give people—not on how to compete in venture, but how to compete in other spaces that are potentially right for transformation. It's really two things we were able to do. One is just having been a customer—you have a perspective on these things. There is a real knowledge advantage if you've been a customer to really understanding the shortfalls and the opportunities. That's one lens. But you actually have to do that.
--:-- Marc Andreessen: Yeah, but that was a hell of a hard lesson that we had to learn that way.
--:-- Ben Horowitz: It was. Building a company is a lot of knowledge gathering. Yes, 15 years of pain and glory. The other thing is, we've been talking about this the whole discussion—to take a structural view of the industry. These industries are not permanent and timeless. Just because things work a certain way today doesn't mean that's how they've always worked. Almost certainly that's not the case. Almost certainly the structure of any industry has changed a lot over time as circumstances have changed. Therefore, the structure of whatever industry is today is not going to be the same structure it's going to have in 10 or 20 or 30 years. But incumbents, especially incumbents that no longer have their founders, are highly likely to underestimate the amount of structural change and they're going to have a hard time adapting to it.
--:-- If you adopt a structural approach, you can get a little bit of a crystal ball. Combine that with the customer mindset. You can look at a little bit of a crystal ball and say, "Okay, it's going to change this way." Then it's the gap between the way that the incumbents are currently doing it and the future way that it ought to work. That's where you have the insertion opportunity.
--:-- Erik Torenberg: There's a related quote to this, Marc. In a New Yorker profile on you many years ago, there's this quote that says, "Marc Andreessen sometimes wonders if Naval Ravikant is onto something, the founder of AngelList. He asks Horowitz, 'What if we're the most evolved dinosaur and Naval is a bird?'" Impact that.
--:-- Marc Andreessen: This was in the middle of the call. This was in the heyday of AngelList. That was a good question. Well, first of all, the question is totally ruined because we now know that the dinosaurs were birds. So the T-Rex is running around with feathers and a beak, which my 12-year-old self is deeply disappointed by. Jurassic Park—the next Jurassic Park reboot is going to be very sad and depressing.
--:-- The specific point when I said that, whatever, a decade ago—Ben will recall it was when AngelList was right in the middle of aspiring to structurally replace venture the way that we were doing it by having it be an online marketplace approach. That was one kind of disruptive opportunity. By the way, crowdfunding—there are cases where that's worked really well. That's one form of structural change.
--:-- The other form of structural change, of course, is AI, which I didn't have in mind a decade ago applying to venture. But today you certainly ask that question: "All right, smart guys, you're sitting around doing all this analysis and you have all these smart people doing all this modeling and research. Why can't you just plug this into Claude or ChatGPT or Gemini and have it tell you what to invest in?" That's the new version of the question.
--:-- Ben Horowitz: There was also crypto a few years ago—ICOs or DAOs. Was that going to be disrupted?
--:-- Marc Andreessen: Had ICOs stayed legal—they were outlawed basically—but had ICOs stayed legal, you'd have a totally different way things happen. By the way, it turns out to Ben's point, the main thing that actually happened was the private market screw-up. What actually happened played to the benefit of VCs just through happenstance in this particular case. Firms like ours raised much larger growth funds and play an even bigger and important role. But there's absolutely no guarantee in life that the next structural change like that will work on our behalf. Ben will tell you, I'm always a little bit of an obsessive paranoid about what happens when the next change happens.
--:-- Ben Horowitz: Yeah. It's interesting. I would just say AI—I think it might eventually be better at us than picking. But the great thing about venture capital is picking is a small part of the game. Who gets to pick is as important. How much of that can be done with AI? So much of what a venture capital firm is—its relationships with the world—do you get that benefit? To build a company, you end up needing a lot of relationships. That's why I say 90% of the activity at the firm is that.
--:-- Marc Andreessen: Eric, you may know Tyler Cowen has talked about this long-term pattern that actually goes back 400 or 500 years of what he calls "project selectors" or "project pickers." The story has been told many times, but the origin of the concept of carried interest in the venture capital and private equity world—how we get paid—actually goes all the way back 400 years to the whaling industry. "How much whale can you carry?"
--:-- What would happen is you would have these project pickers—basically angel investors in whaling expeditions. A whaling expedition, like in Moby Dick, is literally a ship and a captain and a crew going out to try to get a whale and bring it back. In the early days of whaling, it was like two-thirds of the time the ship comes back. The other third of the time the ship doesn't come back. High-risk, high-return occupation.
--:-- There were these guys who were the capital suppliers, and they would sit in coffee houses or pubs. The captains would come in and pitch the project. They'd say, "I'm going to buy this ship, go to this spot, this is my approach, here's how I'm going to staff my crew." The project pickers, the financiers, had to decide whether to back the captain. If they did, they'd give the captain the money to buy the ship and hire the crew. If the ship didn't come back, they'd lose all their money. If the ship came back with a whale, the carried interest was the 20% of the whale that the captain and the crew got to keep. That was how they got paid.
--:-- But like venture capital, they're literally doing venture capital. Queen Isabella did venture capital when she financed Christopher Columbus—exact same thing. The Puritan founders of America—that paid off massively. It had some negative consequences or side effects, but it was a good investment. It was a very good venture bet. The original Puritan colonists of Plymouth Rock actually spent 20 years exiled in the Netherlands, essentially raising venture capital to buy land and come to the U.S. to create the new colonies.
--:-- We're also describing the process of A&R people at record labels who pick new music. We're also describing book publishers who pick new novelists. We're also describing movie studio executives who decide what movies get made. What Tyler says is basically anytime you have a part of the economy in which you have an entrepreneur going on a high-risk, high-return endeavor where it is far from clear what's going to work, and there are many more aspirants than there is money to fund them, and it's this multifaceted skill set that's required to do it.
--:-- By the way, funding them to Ben's point—you're not just funding them, you have to actually work with them to help them execute the entire project. That's art, not science. We would like it to be science, but it's art. How do we know it's art, not science? Every great venture capitalist in the last 70 years has missed most of the great companies of his generation. The great VCs have a success record of getting, I don't know, two out of 10 of the great companies of the decade. That was true of all these legends I mentioned earlier. If it was a science, you could eventually have somebody who just dials it in and gets eight out of 10. But in the real world, it's not like that. You're in the fluke business.
--:-- There's an intangibility to it. There's a taste aspect, the human relationship aspect, the psychology. A lot of it is psychological analysis. Who are these people? How do they react under pressure? How do you keep them from falling apart? How do you keep them from going crazy? How do you keep yourself from going crazy? You end up being a psychologist half the time. It is possible—I don't want to be definitive—but it's possible that this is quite literally timeless. When the AIs are doing everything else, maybe this is one of the last remaining fields that people are still doing.
--:-- Erik Torenberg: Ever since I co-founded a firm in 2016—and I'm sure before that too—people were talking about how software was going to disrupt venture completely. Whether it was crypto or AI or something else, the asset class has changed in a bunch of the ways we described. It hasn't been fundamentally disrupted in the same way we think about disruptive innovation or the Clayton Christensen term, perhaps, as in other industries. Not yet.
--:-- Marc Andreessen: Yeah. But it could. Again, we could be doing a podcast next year and be like, "Oh, oh, hell."
--:-- Erik Torenberg: Well, this is great to get some of the history of the firm. In future episodes, we'll talk about where we're going among other topics. The Ben and Marc Show is back. Mark, Ben, thanks so much.
--:-- Ben Horowitz: Yes. Okay, thank you.
--:-- Marc Andreessen: Yep.
--:-- Ben Horowitz: And welcome, Eric.
--:-- Marc Andreessen: Yeah, welcome, Eric.
