So far this year, the amount of IPO activity has been twice what it was in 2023, and with the Fed continuing to pivot to lower rates, we certainly think that the new issue market will continue to be very vibrant in the coming year. We're also so proud that over 80% of the companies that have gone public in the U.S. this year have chosen to list on Nasdaq. I'd like to now kick off today's program by turning it over to Morgan Stanley's Matt Cost and the team from AppLovin. Thanks.
Awesome. Yeah, thank you so much. Thank you, everyone, for being here. My name is Matt Cost, the Morgan Stanley U.S. Internet team. Thrilled this morning to be joined by Adam Foroughi, the CEO of AppLovin, and Matt Stumpf, the CFO. Guys, thanks so much for being here.
Yeah, good morning.
Thanks for having us.
Awesome. So I'm going to let you guys carve up the answers to these questions as you see fit. So don't feel excluded, either one of you, as I go through them. So let's start high level here. So we've seen an extraordinary inflection in the growth of your business over the past two years. And I want to pose a question that I hear from investors quite often as they're ramping on your company. What have you built at AppLovin that's driven this outperformance? And what differentiates it from your competitors?
So I talk a lot about culture. And I think this is one of the main differentiators. And our culture seems to me to be different from a lot of other technology companies that have scaled. And it's predicated on the concept that I've built multiple companies. And I want to work with builders. And people who build companies tend to want to work at small, early-stage businesses. And I never understood why these same people who have exceptional ability end up churning out of companies when those companies scale. Why do companies become less efficient when they scale? Why is there more process when companies scale? So we wanted to build a culture that inspired that same entrepreneurial spirit as it scaled.
That is predicated on building a team of doers, getting process out of their way, and allowing them to execute as if the company was still 10 people. We pride ourselves on being able to do this now at very large scale.
Great. So you've been asked a lot of times about the path forward for the core gaming ads business. And you've heard to kind of a 20%-30% growth target over a multi-year period going forward. So that's a fair bit faster than the mobile games market itself is growing from a consumer spend perspective. So take us behind the scenes for a few minutes and talk us through the moving pieces to get to that figure. So how do you build to 20%-30% growth?
Yeah, sure. So before we dive into the 20%-30%, I think it's really important to address this point that we've heard from some investors out there. I think most of the investor community has kind of worked past this. But we still hear it occasionally, this misconception about how we can grow faster than the TAM. And one important thing that you need to understand just about the ecosystem where we live is that our customers are arbitrage marketers. So they're setting return goals. And we're finding users that are hitting those return goals. So the more users that we can find for these advertising customers that are driving the returns for them, the more scale of spend we're going to have. And so that then is driving TAM growth over time.
As we're finding more users, as our technology improves, then the TAM is actually growing as a result. And the 20%-30% long-term goal of growth, how we think about that on the existing business, the 20%-30%, is really driven by two components, the first being reinforcement learning of our technology. So these AI systems are always learning. So as we show transactions, the system is learning from what happens within that transaction, what the user behavior is, and whether or not it's driving a result at the end of the day. And all that information is fed back into the system. And it's learning. And that's determining the future behavior of what the system is doing. So that reinforcement learning is driving, on average, about 5% per quarter. And then the second component are directed model enhancements.
So our team's constantly looking through new research that's being posted out there, investigating different techniques to potentially improve our algorithm. And those direct model enhancements, on average, we think could drive at least 10% growth on an annual basis. And over the last six quarters, we've had about four quarters of those types of improvements in the system. So we're pretty confident that we can continue to drive at least one of those types of improvements on an annual basis going forward.
Got it. OK, let's shift to a topic that I think a lot of people in the room are probably very excited about, which is e-commerce advertising. So let's just start with, what are you doing in e-commerce? How are the early tests going? And why do you think AppLovin will win in this category?
Yeah, so obviously, there's a lot of excitement around this. But let me start with why we ended up talking about expanding outside of gaming to begin with. When we launched AXON 2, we saw the power of this technology. And this is what I call personalization technology. It's a recommendation engine that's really powerful. Now, if you have a recommendation technology and you give it just one category of content, it can't really do what it's designed to do. An analogy would be it's easy to forget that TikTok once was Musical.ly. When they bought that company for under $1 billion, it was niche. They only showed music videos. Then they opened up to all types of content and let the technology do the rest to scale the product. Well, we've got a similar technology. It's very powerful.
And we let it only show gaming advertisements to the consumer. So once we saw the power of this technology, we said, look, this thing's going to be able to do anything. So let's get it to be open to everything. And so we started building the tools and technologies to enable any website with a business model to be able to advertise on our platform and reach this really large audience of 1.4 billion daily actives through us and do it on a performance basis. And so we had a lot of confidence that we were going to be able to do this. We did that work. We rolled it out. And now there's been a lot of noise on Twitter from customers that are early in a pilot that we're running that are seeing phenomenal results.
And so I'd urge investors to go to Twitter and look up AppLovin and just see the commentary to understand really how performant this product is for those advertisers. We haven't talked much about the details outside of that. But I did say on our earnings call last quarter, this is the best and fastest growing product we've ever released. That includes the algorithm itself. So we're really excited about what the potential is here. Now, I do want to talk for a moment about what winning means to us. Because as someone who does like to win, and I feel like as an entrepreneur, you have to aspire to win in some way. You have to be motivated to win and understand what that means to you.
We built our company to try to win by creating and expanding economies, not in the traditional sense of what people think about winning means in advertising. Let me give you an example of what that is. Typically, people in advertising believed in order to win, you had to have a product where you can take dollars away from another player in an ecosystem. We believe to win, we can create an economy. Now, imagine a world where there was one mattress company that existed. This was the whole world. There was a company like Meta out there that's exceptional at advertising, massive audience. They've got a super set of our audience, serve a lot of ads. They built a huge economy in the world. Let's say this mattress company was able to run ads through Meta and deliver 1,000 purchases of mattresses a day.
Now, if Meta was perfect at advertising, they showed every single user in the world an advertisement for that mattress, the user saw an advertisement for that mattress, that you couldn't create more purchasing power as a secondary platform, but it's just not the case that the audiences are perfectly overlapped, and every single user can notice an advertisement on the social network, so we come along. We write this technology, and we go, "Hey, mattress company, run your ads on our platform and reach this audience that's doing something different. At this moment in time, this audience is very, very large scale, is in a different mindset, and let us see what we can do for you," and they come onto our platform and they launch.
And let's say they're buying on a performance basis, aiming to get the same results that they get on the social network. And really, these advertisers are arbitrage marketers. So the return on ad spend that they target on a Meta has to be matched by us. And they launch a campaign. And they deliver 1,000 transactions per day. Well, we don't want the 1,000 to be pulled away. We want the 1,000 to be additive. And so if we can then measure that the 1,000 is completely incremental to their business, and now there are 2,000 or more mattresses being sold per day because the incremental advertising channel will have expanded the economy. And that's important. When we talk about winning, that's what we build products to do.
Great. I guess one thing that investors are curious about is the investment behind the e-commerce business. And it seems like, based on your comments just now and things you've said in earnings calls, like the technology is there. But what about the go-to-market? Do you have the sales and support infrastructure in place to scale e-commerce? And are there investments that you need to make in order to grow there?
Yeah, I mean, a lot of times when we talk investment internally, we're talking about engineering. And we paid for our engineers already. So we don't talk about costs. And we understand that investors really want to focus on what's going to be the incremental cost investment to scale things. We're an engineering-first company. And we believe that if we build a product that's so good, that's automated, and drives to measurable results on behalf of the advertisers, that product should sell itself. And so we believe we have that product. The magic was building the technology. Now where we are is we have to build a lot of automation and tools to allow advertisers to come onto our platform and be able to manage the campaigns themselves.
Once we have those tools, and it may be in the future, you have an advertiser who wants to talk to an account manager. We don't want a person in that seat. We want an AI bot account manager talking back to that advertiser and helping them work through their issue, and so that's the way we think about expanding, going into a new market, is not to go hire a bunch of account managers and then salespeople, but build a product that could sell itself so you don't need the sales layer, and then now in today's world, build an AI bot to become your account manager so that we can think about scaling this product to millions of advertisers without expanding our headcount, and in our business, it's not lost on us that the potential outcomes in the future are unpredictable. We may be unpredictably exceptionally good.
We may have unpredictably different outcomes. But the one thing that we can commit to investors about is that we don't have an interest in expanding our headcount materially to go after revenue. We think we can go after revenue and automate every step of the way. And if we can do that, we can build a business that has exceptionally high margins.
Great, so when we think about your machine learning engine, which is branded as AXON, so when we think about AXON and the ability to extend to e-commerce, does that imply that you could extend the business to other non-gaming ad verticals? And if so, how broad is that opportunity?
Yeah, we think it's really broad. We expect, now that we've seen the data that we've seen, that any company that is online with a website that has a business model that drives to some KPI will be able to buy advertising through our platform. Now, that KPI may vary. We've chosen specifically to go after mid-market direct-to-consumer shops right now. Now, their KPI is revenue and profit. They want to advertise and get users and know that they're making an incremental profit on that advertising. To us, that's the hardest KPI because it's so far down funnel. Now, you can imagine a different type of advertiser who's sending emails to sell financial tips could come online, well, we can certainly sell emails, that's easier to go build a cost per lead model, and that exists in the system already.
Another advertiser, like a streamer, might want to sign up subscribers. Cost per registration is also supported in the system, so because we have a KPI-based advertising model, it's completely automated, and we've proven we can expand out to web and make it work across a wide variety of different types of business models at all sorts of different prices in terms of cost per goods sold and revenue per goods sold. We know now that we can go out and expand this product to millions and millions of customers in the future.
I want to touch on connected TV for a minute, and I'm curious about a couple of things on this topic. Firstly, why is connected TV important to AppLovin? What does it allow you to do that you couldn't do before? And maybe most importantly, what capabilities can you bring to that ecosystem that you think advertisers are really looking for?
Again, we're focused entirely on performance. And what bugs me is that when we look at the connected TV device, this is an IP-enabled device. And the advertising is sold exactly the same way as linear TV. You had brand campaigns on linear TV that cut over to connected TV, very large space. But there haven't yet been economies that were created or expanded through this new way to access the consumer. And what I mean by that is if you had a linear TV campaign for Coca-Cola and it was running at some amount of dollars per day that were substantial, and some of the traffic shifted to connected TV device, those budgets shifted in the same amount of the audience shift. And nothing incremental happened. The market didn't expand.
Now, let's say through our product, we can serve an ad for an up-and-coming soda company like Olipop, as an example, and we can serve that ad and make a completely closed funnel attribution and show the user an ad, track to transactional value for the soda company that's up-and-coming, and make that ad performance-based. What will end up happening is that in today's world, the consumer is not discovering anything new through connected TV. They're seeing that same Coca-Cola ad they would have seen on linear TV, but if now they start seeing ads for all sorts of different products that have no means to advertise on television today, economies will expand. That's what gets us excited. We want to look at big opportunities where we can expand economies.
If we can deliver technologies to do that, we believe we're going to become a lot bigger than we are today.
Great. So there was a long period where investors were pretty focused on competition in the ad network space. And I think we're now at a point where most people believe that AppLovin is the clear leader in your space. So I want to ask you, given the success that you've had, who do you view as your competition now and why?
This is an interesting question. Because when I started the business, we couldn't raise a seed round. And the pushback was competition should eliminate you. You really have no reason to build this company. Go focus on another space. And really, when people talk about competition back then, it was not dissimilar to today. It was Facebook and Google have every advantage in the advertising world. They should win every single space of advertising in the future. And for one, when we built the business, we knew if you can focus on something specific, you can build a really large business if that's something specific, is a big market, and compete with the biggest companies in the world. More importantly, when we built our technology and always thought about our products, again, I use this theme constantly, we thought about expanding economies.
We didn't think of Facebook and Google as competitors to these advertisers. We didn't want the advertisers to go, we're spending so much money on Facebook and Google, and we're going to carve off some of that money and bring it to your platform. What we wanted to do was build a technology that was so good that we could be additive to those two companies. We looked at them for inspiration. They built exceptionally compelling products. We still do look at other companies around us today for inspiration. But we don't think of these companies as competitors. In fact, we're partners with a lot of the companies in the space. We want to expand the spaces and the economies that we operate in.
Got it. So you've been pretty active in repurchasing stock historically. You continue, I think, to have a large authorization there. You've also done some M&A in the past, including MAX and MoPub and Adjust. But looking forward, what are your priorities? Does the company need any capabilities that it can't build organically? And how do you evaluate the attractiveness of continued stock buybacks?
Yeah, that's a great question. So our focus today is not, at least in the near term, on M&A, and it's more on these organic growth initiatives that we've been talking about, which are primarily the continued development of our technology for the existing mobile gaming business, that algorithm, and then the expansion into transactional web advertising, so primarily e-commerce, so the investment there, this goes back to the question we were talking about previously about investment within the teams to support those initiatives. I mean, we grow our teams and headcount in a very lean and efficient manner, as you can see from our margin profile.
And that's the plan going forward as well, is to continue to invest, but very modestly, so to add on headcount, only finding the best talent that's available, and then at the same time, continually reviewing the rest of the organization to find areas to trim at the same time so that we continue to retain that margin profile that we've got today. So the business is continuing to improve in terms of its margin profile and cash generation. So then we're left with, obviously, tremendous free cash flow generation, what to do with that. So what we believe is the best usage of that cash is to continue to repurchase shares. Because what we're focused on is driving long-term shareholder value.
We've seen that thus far over the last couple of years, investing over $3 billion in repurchases, we've been able to drive tremendous value for our shareholders by actually reducing the overall share count. The plan going forward is to continue to invest the vast majority of our free cash flow in repurchases. Given the growth opportunities that we think we've got in front of us, the 20%-30% that we talked about before, the expansion into transactional web, that we really shouldn't be price sensitive in terms of purchasing behavior. We're going to continue to repurchase our shares on an ongoing basis going forward. We just received an increase in our repurchase authorization from our board this last quarter, increasing that by $2 billion. We have a $2.3 billion authorization today.
We continue to plan to be in the market going forward.
Got it.
I do want to add a point on M&A too. We were acquisitive in the past. So people assumed that we're just an acquisitive company. And a lot of those acquisitions were meant to enable our core advertising business to grow and become what it is today. Now, we've got a lot of organic opportunities in front of us that we're excited about. But when we look in the rearview mirror at the M&A that we've done and assess it, we realize that M&A is exceptionally hard to execute on. I didn't realize how hard. And what we found for us is that our culture is different. And when you want to go out and buy and make an impact, you've got to buy something that's substantial to the relative scale your business operates at.
Now, with the scale that we operate at today, that would mean that if we were to buy something, it'd have to be a pretty substantial acquisition. Those companies don't tend to be set up where you can go in, buy them, and let go of most people and dump the technology and recreate it, and really, as we look in the rearview mirror at the M&A we've done, the two most successful we had were MoPub and MAX, where we didn't inherit culture. We didn't inherit people, nor did we inherit technology. The rest of them, we've had a huge headache trying to absorb and get the value out of the M&A that you expect when you do the deal. Now, we were able to extract a lot of value because the deals were structured to help our advertising business develop over time, well, it's developed. It's valuable.
And we think we can make it much more valuable organically going forward. So that makes us, as an acquirer of companies, a really tough place because we end up having this differentiated culture that can't merge with other cultures all that well. And so that's why we talk about M&A not being much of a priority. If we fundamentally believed it could accelerate our business and it was easy, then we wouldn't say that. But we think it's much harder than people realize. And it's exceptionally hard for a company that's structured like us.
I want to tie that back to profitability with Matt in a second. But just to stick with that point for one second, when you look out over the next five or 10 years, do you see a situation where you'd need to significantly extend the engineering teams or sales teams? On the product roadmap that you envision, is this something that you can just leverage a really lean team for indefinitely? Because to the comment you just made, it sounds like you have a really, really effective culture and a pretty small team that's lean and flexible. And is there sort of any reason that that would ever need to change?
Yeah. I mean, on the last earnings call, I gave a new metric that we care about, and I said EBITDA per employee, but really, we care about cash flow per employee, and we expect the numerator to go up over time and the denominator to go down, so importantly, we think our headcount over time, even as these opportunities in front of us become bigger economically, we expect our headcount to go down. I don't know of another company that is scaling a business as quickly as us that would say that point, but the reason I fundamentally want to accomplish that is because we're working in a world now where AI technologies are allowing human beings to be more productive, and if our human beings that are exceptionally sharp want to work with other A+ players, you can't hire that many A+ players.
There's just not that many A+ players sitting in the world to hire. And so if we continue to optimize to those players and give them the capability to utilize more AI technologies to improve their output, we believe we can grow the company, execute on all of our initiatives, and lower headcount over time. And that's something we're really focused on trying to prove out.
And from a go-to-market perspective, is it just that performance sells itself? And so that's it, people come to the product over time because they see it works?
It's two things. It's performance and patience. So the product itself is really good. I don't know of another ad tech company that's gone viral on social media as we have on Twitter today amongst the population of advertisers we've allowed on. But then you also have to have patience. If we wanted to go out and get the largest brands in the world to advertise on our platform tomorrow, we'd have to ramp up a sales force and really go sell them. Because while some brands move quickly, the largest ones move pretty slowly. And in order to go get them quickly, you need to sell into them.
Now, if we were to have the patience to say, look, they're going to come anyways, if this product is really, really good, that it can sell itself, well, maybe the largest brands in the world will get five years from now and not tomorrow. But then we don't end up with that headcount. And so fundamentally, we think we're going to be able to get every single advertiser. We're just going to have to have patience and understand that those advertisers will come over time. And this isn't dissimilar to how we were able to scale in gaming. We originally built our products in gaming and targeted advertisers to come on board that were the fastest moving. These are normally called indie studios.
They're smaller businesses where even the CEO who's developed the game and eight other people are running it, the CEO is still making the decisions around marketing, and these companies would come to us, and so five, six years ago, if you talked about who our advertiser base was in gaming, it was a bunch of no-name companies, and when we went IPO, people did industry checks, and we're like, huh, the largest gaming companies don't even work with this company, and what is going on? Why is that? Well, it's because the largest companies take time to go adopt new platforms. They need to be sold to. They want to feel like VIPs, and we build technologies that sell themselves, and we don't want to have the staff to make anyone feel like a VIP. We want them to get extreme benefit from our products.
And we want them to come because they hear so much noise around them that they end up not having a choice. And that takes patience to wait for. And we're going to continue to execute on that same model.
Great. I mean, there's a lot of really interesting qualitative background there. But let's tie it back to the quantitative side now. So I mean, understanding everything that Adam just said, do you see, Matt, a path for the ad business to grow at 80% + incremental margins going forward the way it has for context for several quarters, quite a few quarters now?
Yeah. I mean, given the ethos that we have at the company in terms of making sure that we're growing in a very lean and efficient manner, we're very confident in our ability to continue to drive growth at this profile that we've had over the last, I think it's four to six quarters, we've had on average about 90% flow through from revenue to EBITDA. And really, the only cost that we've driven up over that period of time are data center costs. So our infrastructure needs to grow. And that's primarily Google Cloud that we work with. So as our revenue increases, obviously, our model continues to get more complex. And we need to support the scale of that existing revenue. And what we've seen over that period of time is around 10% on average is kind of the incremental cost as we continue to grow revenue.
And so we actually expect that our margin profile should improve from where we're at today as we continue to drive that kind of 90% flow through from revenue to EBITDA. And we're hopeful that to Adam's point, as we continue to grow, that we can actually find ways to be even more efficient than that and continue to drive margins up.
Great. Maybe we'll close out 1,000-ft view here. The company's obviously achieved a lot of success. Investors are very excited about what's coming in 2025 and beyond. So from your perspective, what do you see as the most or what would you characterize as the most exciting opportunity ahead? And then flipping it around, what do you see as the biggest challenge and thing that you need to focus to execute through?
Yeah. I'd say this ties back a lot to my own role as well over time. And if you talk to us a couple of years ago, I was embedded into every aspect of the company. But most of my time was spent running product. That's what I did for most of the company's existence. And as our technology became what it is today, the mathematical complexity got to a level that I have no idea what's going on there. And unfortunately, that means I was replaced. The engineering team absorbed product and took that away from me. And a few months ago, I was trying to figure out what to do. I was like, my hours per day of work had gone down to such a low amount. I wasn't sure what to do with myself. And I took on a new role. I took on head of HR.
So not many CEOs of technology companies are heads of HR. But I did this because I feel like our biggest challenge is executing. And I feel like the biggest challenge to executing is a cultural disruption. I have these principles. And I want to ensure that they're maintained. And a lot of times, as companies grow, you end up building out these different organizations, especially HR as the culture org. And the founder loses sight of what's going on inside the organization because the company scales beyond a point of that founder having visibility. And that happened to even us. And so when I saw the opportunity to take back control of that and ensure that my biggest fear, our culture getting disrupted and then leading to churn of our best talent, doesn't ever come true, it was an opportunity that I wanted to go after.
So that's what I'm focused on as far as addressing the biggest challenges that I foresee could come up into the future. Now, in terms of the biggest opportunities, we've written, again, this algorithm that's really powerful. We have, I think, quarters and years of growth in front of us. And importantly, we want to do that without expanding costs. So therefore, we have a lot of tools that we've got to build to continue to automate this process on behalf of the advertisers. But the great thing about what we can become is that we can go from a company that has expanded a niche market, the gaming market, the dollars that are spent on our platform are directly contributing to the expansion of that economy. But that's not a big economy in the world.
We now have the possibility of going after the whole economy that's connected through digital and creating more discovery, and if we're able to execute on that, we think we can be a company that creates economies, and when you think about the biggest companies in the world, Facebook, Amazon, Google, they've created economies. The biggest companies either enable the creation or do it directly, and that's what we aspire to do.
All right. That's a great place to close it out. Adam, Matt, thank you so much for being here.
Thanks, Matt.
Thanks for the great questions.