Inside a16z Growth: Waymo, Stripe, ElevenLabs, Revolut, Coinbase, Kalshi

There are fewer than five companies growing north of 30% this year. We didn't invest in a company last year growing 30%. There's just a major disconnect in those growth rates. OpenAI and Anthropic alone added as much revenue last year as half of the public cloud universe. The most underappreciated company in America right now is Flock Safety. Garrett Langley should be a national hero. The most common question right now is like, do gross margins matter? A lot of these AI companies are showing up with zero to 50% gross margins. And we're used to looking at companies that have 70% gross margins.

Well, if that market existed today, I think it would probably be like 90% plus odds.

Thanks for having me. Excited to be here.

Coinbase, Elise AI, Hebia, Sardian, Anderl.

Yeah, I mean we're very fortunate to work with what I would think of as the best early stage team in the business. And so we get tremendous leverage and opportunity from them. Of the companies you named and of our portfolio at large, roughly 50% of our investments are follow-ons to our early stage winners, companies breaking out. Those are companies where we're on the board, we have access to information, they're excited to partner with us, that provides tremendous opportunity for us to invest in them. And then the ones that are outside the portfolio, like a Cali or Revolut, we've met with them many times over the years. 90% of the companies we're investing in out of our growth fund, our early stage team has met. And so we get to take advantage of those opportunities even though they didn't pursue them or it didn't work out for one reason or another.

The most underappreciated company in America right now is Flock Safety. Garrett Langley should be a national hero. They got a lot of positive press attention in December around the Brown and MIT case. Their camera helped solve and find the criminal. When we look at it though, that's just like a normal day in Flock's life. Flock solves over 2,800 cases every day. That's 15% of reported crime in America. And so like there's a lot of, oh, like is tech doing good in the world? AI, is it good for humanity? Like Flock is an incredible example. Like one I'm so proud of. And like we need Flock. If we want to eliminate crime in America, we need to deter criminals from doing bad things, right? Like we don't want bad guys to think they can get away with it. 50% of murders are unsolved in America. We need Flock to be deployed and eliminate all crime.

Look, there's a big disconnect between public markets and private markets in terms of growth rates they're seeing. So, my partners David and Santi put out a report calling private markets the new high growth public markets. If you were to scan public internet software fintech, there are fewer than five companies growing north of 30% this year. We didn't invest in a company last year growing 30%. There's just a major disconnect in those growth rates. If you look at OpenAI and Anthropic alone, those two companies added as much revenue last year as half of the public cloud universe excluding the Mag 7. I mean just all the growth if you're a growth investor is in the private markets. I mean I think the most common question right now is like do gross margins matter? A lot of these AI companies are showing up with 0 to 50% gross margins and we're used to looking at companies that have 70% gross margins. And like do they matter? Of course they matter. They always have mattered, always will matter. I think how we make sense of it are the lower margins because of LLM costs. If they are, like we're cool that we can make sense of that. The real question is as we look out, are you a company that's value is primarily related to someone else's model or are you building something meaningful on top of it? And so through the metrics can we see differentiation, can we see defensibility in the workflows, the integrations and the data? And so that's how we've been thinking about gross margins. And like is the topic du jour today, I think engagement today has really come into the right around us. I mean if you think about engagement, it's like a leading indicator of retention. A lot of these companies, they sign annual contracts, they're growing so fast. All of their revenue base hasn't come up for renewal. And so like are people actually using the product? We're really excited about Harvey. We led their last round. One of the most exciting parts of that company is if you look at their engagement cohorts, they're increasing over and over and over. And so you can just see that becoming the lifeblood of a law firm or lawyers inside corporates.

The meme has definitely slowed down. I think it's going to become a billion. Hopefully, you know, Cursor shared that they reached that in record time. I expect other companies are going to, you know, try and get there.

We can start with Waymo.

Waymo raised $16 billion at $126 billion valuation. It was their Series D. We're really excited to be a major participant there. We've participated in all four of their external rounds dating back to the Series A. They're growing phenomenally quickly. In San Francisco, it got to 25% market share really quickly with very few cars. You know, surpassing Lyft and being a formidable competitor to Uber. That is without having like full coverage of the Bay Area, without having highway access or airports until recently. So like that market share was really constrained based on those dimensions and it was also constrained by the number of cars. Like Waymo today still is constrained by the number of cars they put on the road. If they put more cars on the road in San Francisco, that market share would be, you know, much closer to Uber, but they're opting to, you know, grow across the country. They're scaling to over 20 new cities this year. And so the growth is very much limited by just getting cars on the road. They have incredible utilization. Look, ride sharing is a great market. Today, it's like a $125 billion market growing low double digits, which is a healthy growth rate for a relatively mature industry. 40% of that is in the US, where Waymo plays today. Waymo, you know, as I alluded to, like you put cars on the road, they're taking share. High utilization. There's only like a thousand-ish cars in the Bay Area and you like can't go down the street without seeing one on each block. It's truly tremendous. But as far as like ride share goes, like thinking about the ride share market today is kind of like looking at the early days of Uber and looking at the taxi market. It's just the tip of the iceberg. Ride hailing in that aggregate large number I just shared is less than 1% of vehicle miles in the US. So really like massive, massive way to go already. You can look at a Waymo and say, wow, there's a bunch of new use cases opening up. You might be like talking to someone about confidential information that you wouldn't have felt comfortable. I got a text this morning about a parent who got injured and was so excited because they were able to put their child in a Waymo and they wouldn't have felt comfortable doing that in an Uber. But that's before the price goes down. And today Waymo is pricing similar to Uber and Lyft. If you look across the country in San Francisco, it's a premium product. You could think of it like Uber Comfort. It's still like growing like wildfire when the price goes down. And it's meaningfully cheaper than human-led cars. People are going to think about the trade-off between driving themselves or getting in a human car or getting in a Waymo and it's going to be cheaper to get in a Waymo. And so the ride share market is going to grow by, you know, at least an order of magnitude in the coming years.

The value, look, I think the value is going to accrue to the driver and whoever is making the driver. There's been a lot of investment. It's a really hard technical challenge and so I think the driver is where the value is going to accrue. That would be Waymo in this case. I think where margins land over time is an interesting question. Like on one hand Waymo could go the route that they are today and own and operate cities like in San Francisco and Los Angeles. They own the experience with Waymo One. They are managing the operations and the depots. As an alternative, they could outsource some of that as they are in Phoenix, Atlanta and Austin where third parties like an Uber or a Move are managing the operations. And so when you outsource that and those are like, you know, real costs, you could lead to a higher margin product. This is never going to be like high margin software, but it'll be better than, you know, operating the fleet themselves.

Yeah. So this is one where you have to think about a variety of scenarios off into the future. And so to make it simple, I would frame it as there are three like base scenarios. The first is Waymo stays the only game in town. I certainly like hope that is the case. Their dominant market share of a market that grows quite significantly. The second case is the market still grows significantly because it's such a great value proposition, self-driving. However, there are multiple players. That might be a Tesla, it could be a Zoox, it could be a WeRide. There are Chinese players as well. But because of the technical challenge, I think it's a productive oligopoly. And then the third scenario is where things don't work out as well for Waymo and, you know, maybe there's impairment, maybe one of those other providers has an advantage over them in the fullness of time. I think the most likely scenario of those is that it is a productive oligopoly, that there are many players, but we expect Waymo to be, you know, a meaningful share of that overall market.

So, we love Elon. We're investors in SpaceX, XAI, which you may have heard are a single company or soon to be a single company. So we're huge fans of his. We have tremendous respect. Tesla from a hardware stack perspective, they have taken a camera-only approach. Critics would say that lidar provides more data and is better performant in adverse weather. So like fog, snow, rain. And so we believe like the Waymo approach today is clearly like safer. That said, like if you look at FSD, like the publicly available data, they have shown significant improvement. The slope is promising. FSD 14.1 I think showed an order of magnitude improvement in terms of disengagements. There's a lot of puts and takes with disengagement metrics, but like they are definitely improving. But when you take a step back and compare the data, whether it's disengagements, reported accidents in Austin, Waymo is still significantly ahead today and we feel really good about their position.

I have not been in every single one of their markets, so I have not taken one, but I've taken one here. I've taken one in Los Angeles.

I love to hear that. I love to hear that. They're growing really nicely there. They're growing nicely across all their markets. And each subsequent market is growing faster and faster, which is a reflection of how their brand is growing across the country. Which I think will be a powerful force for them as you think about their position long term. Like nobody wants to have five ride share apps in their phone, right? So I expect that there could be several players at the end state, but there's not going to be endless. So they've announced partnerships with Uber and Lyft. How those relationships evolve over time is like TBD. In Austin and Atlanta, you can order a Waymo through the Uber app.

Autonomous.

It's magic. It's the best manifestation of physical AI today.

Yeah. The grandparents are like more likely to be like fearful before they get in their first one and then they realize how amazing it is.

Grandparents are probably going to need to be the first to get their driver's license taken away.

Yeah.

Look, I mean, you brought up Elon. He announced last week that Tesla is shifting production from some of their older vehicles to Optimus. I think it would be really cool to see their robots, their humanoids be more involved in our life. China has done a really good job on that front and I think like the US like needs Elon, needs to continue to push that forward.

No, I mean I think it's really exciting. I think that like whole phenomena this weekend and last week is like a great representation of like how fast distribution is today. Like we kind of take for granted OpenAI got to a billion users in lightning fast speed and it's because like anyone can get access to these things immediately and immediately deploy. We saw a bunch of our founders messaging back and forth about how they were using it. One of them set up a personal assistant where they like wake up each morning. It reminds them of birthdays. Like gives them an email summary, a great dashboard of what's been done by their team. Like that happened in a day by one personal agent. We're all super excited about that. I think like CISOs are really concerned about the security what these agents are going to do in corporate systems, but like agents are here. It's really promising.

I do not think they're going to take over.

Software is eating the world. Mark was pretty prescient in 2011.

Yeah.

Huge round.

Yeah, it's a great representation of what we were talking about before. Jennifer, Brian, and Justine and our early stage team got to know Maddie, swooped in, led the Series A, involved a flight over to London to get the term sheet signed, all those fun things. But they maintained a really special relationship with Maddie. We led or co-led the A, B, and C. The A and B, our early stage team did. We co-led the C last year and are big participants in this round. It's one where we got to know Maddie through the B. Stayed close to him throughout the year, watched his progress, tried to be helpful. We engaged our operating team a lot there too. Our revenue operations team plugged in. This is a company that has grown in record time. And what's unique about it relative to a lot of companies is they have this beautiful organic consumer business that like people look at ElevenLabs and they're like, oh, how many consumers, prosumers, influencers are going to use this? Like it's going to run out of room on the consumer side. It continues to hum. It's amazing and highly organic. The enterprise business has also been beautiful and it's very rare to have a great consumer business and a great enterprise business. Fast forward to today, the enterprise business has customers like Meta and Salesforce and Deutsche Telekom. They announced Revolut last week. These are massive companies that are building agents on top of their platform.

Yeah, I mean like the answer that I keep coming back to is like Maddie is an exceptionally fast executor. He is really curious. He's really hungry. He's hired well around him. I think like the starting place for a lot of these AI companies is they have the market leading model. If you look at OpenAI with ChatGPT, like their lead came with having the leading LLM. Midjourney's origins came from having the best image model. I think what like Maddie has done really well is like a lot of people in the early days would look at him and say like, yeah, you do have the best model, but is that going to commoditize over time? Is open source going to come? Audio models aren't that expensive. It should be really competitive. And I think like that chip on the shoulder, like the critics, like maybe motivate someone like him to just keep moving faster and faster and build more defensibility and ship product. And so they've shipped more models in terms of breadth. The consumer product has evolved and then now on the enterprise side they have a huge platform that addresses needs across sectors, across use cases. And so it's been pretty phenomenal to watch.

So there's sales, there's customer support, there's interviewing, there's healthcare, there's consumer, there's financial services that they really have a broad array. Which is like when you study the company, it's actually really impressive how diversified the base is. It's a platform already with all these different modules where their customers can pick and choose and buy multiple things. Like audio, like another one of the criticisms in the early days is like, is audio just a feature? No, like audio is a platform and like the main way today or main way in the future that humans may interact with computers.

Yeah. So this is another one where Alex Rampell and Anish on our early stage team were close with Tarek from the early rounds. For one reason or another, it didn't work out. And so dating back two years ago, I was just intellectually interested in prediction markets. I thought it was pretty cool. I was in New York. I met up with Tarek. This is like before they had legal approval. He's like walking me through how they're like fighting and I'm like, I'm a growth investor. I don't invest in companies like before they have regulatory market fit, let alone product market fit. And so I got to know him. I stayed close. He obviously got those approvals. And the election was like a boon for them and the growth has been phenomenal since. We stayed close until it started to really grow and we believed that it was more than just the elections and like boy has it become way more than just the 2024 elections. We built conviction in them and the company.

I am familiar with both companies. It would be hard to study the space and not be. When you look at Kalshi and their origins with Tarek and Luana, it's hard to find two more resilient founders. They were getting punched in the face for four years and told no. And they took the difficult but let's say more responsible path towards getting licenses, being the first CFTC regulated prediction market. And as we reflect on why that was important to us, I think a lot of what we saw in them stem from our investment in Coinbase and them taking a regulatory first approach. Taking a regulatory first approach builds trust with consumers, regulators, partners. And so if you fast forward for Kalshi's journey, they have the most consumers and users today. That wasn't always the case, and I can talk about that in a minute. They were first to get regulatory approval. They have Robinhood and Coinbase as partners, CNN, CNBC as partners. These are like real mature organizations that are entrusting Kalshi. You can't get those types of relationships unless you play by the rules. When we made our investment, Kalshi was number two in reported volume. And you know, we take great pride in backing market leaders. And so when someone's in a number two position, we have to have a really strong thesis. A lot of that stems to what I just said about Tarek and Luana and the regulatory first approach. But if you looked at their slope at the time, the company was absolutely ripping. The space was growing, but their share gains suggested that in a few months, particularly with sports and the products that they were shipping around sports, that we had conviction that they were going to surpass and become number one. And today they're number one across users, revenue, volume, any metric that anyone would care about.

I mean, you just referenced Robinhood. Is trading stocks and options on Robinhood gambling? Those behaviors received similar criticism, but I think are now pretty accepted. Back in the day, like grain futures were considered gambling. As ridiculous as that sounds, so much of the financial market at one point in time has been considered gambling. I think in each of these there is speculative behavior. Speculation is important because it enables liquidity and like liquidity is important for hedging. What I would say though is like a lot of the association with gambling stems from sportsbooks. And today a lot of Kalshi's volume does come from sports. There's no hiding that. If you look at sportsbooks, they set the odds and they profit when you lose. Kalshi is not like that. They match you with other market participants and they have a take whether you win or you lose. So there's not the misalignment that you find in traditional gambling.