AppLovin Corporation (APP) Earnings Call Transcript & Summary

Okay. All right. We're going to get started. Very happy to have back at the conference, first time in person, Adam Foroughi, Chairperson, CEO and Co-Founder of AppLovin. Adam, thanks for being here.

Thanks for having me.

Cool. I want to start on the app side of your business. Q1 earnings, you announced a strategic change in that AppLovin will run its content studios as a stand-alone business rather than as an integrated asset with software and that the portfolio would undergo strategic review. You had messaged this for a couple of orders, but it is a notable shift when we think about your business going back a year ago. So can you discuss a little bit the backdrop to the decision?

Yes. So to understand it better, you got to go back to why we got into apps in the first place. And we've been in Ad Tech business building software to help mobile app developers market themselves for a decade. 4 years ago, we wanted to upgrade our systems to become more personalized with machine learning. In order to drive success in advertising models, we needed data. And in particular, we needed third-party advertisers to share data around engagement and transactions with our platform to be able to execute on what we wanted to do. But since we didn't have the systems, they weren't willing to share the data. 2018, we decided we're just going to take the problem into our own hands. We got into game, started accumulating assets and then marketing them aggressively to grow them running the business as a cost center, really with the goal of generating enough revenue to cover the R&D plus the cost of user acquisition. We did that for 3 years. We then launched an upgrade to our ad platform in Q4 2020. So it's a month -- a quarter before we ended up going public. And at the time, the business was majority gaming revenue. The software business was 14% of our revenue. The gaming business was 86% of our revenue was growing quite nicely because we were buying so many users to fuel the software side with data. Then as the market started understanding how strong our new software platform was performing, advertisers were coming to the platform and wanting to buy aggressively now because they understood they needed to share data to get that personalization and the performance out of the system that they wanted to go achieve. We ended up expanding the software business quite dramatically. We also ended up getting a lot more data into our system from third parties. That led us to, in a year since we went IPO to now, software business is quadrupled. We've gotten reach across the mobile app ecosystem, across many big developers sharing data with their platform and gotten to the point where we're no longer need that data feed only from ourselves, ourselves are a small part of the whole of the business. And so the change that, that leads to is we can run the mobile games much more cost efficiently. And that typically means that in the market, you're trying to make somewhere around 20% flow-through margin. In the games business, we were running at a breakeven, maybe even worse when we first started and ramped it up. And so that's a shift for us. And over time, we'll be able to take a $2 billion revenue stream and hopefully bank a couple of hundred million to $400 million of EBITDA from it, whereas traditionally we didn't, which creates a world where we were able to uniquely lower revenue guidance and increase EBITDA guidance in the same quarter.

And how do you look at the value of that 1P data now? Is there still kind of an important signal that you get from it? Or is it just on a relative basis based on all the signal you get from the third party, it just becomes less?

Yes. The 1P data to us is the same as any advertisers' data. We don't take the personally identifiable data from our apps. So if someone logs into one of our apps with an e-mail, that data never enters our Ad Tech systems. What we're built on is taking an app, allowing it to market itself for a return on ad spend. So an advertiser can plug into our system that they want to spend $1,000 and generate $100 in the first 24 hours. It's a compelling arbitrage model for the advertiser. Everything is programmatic in our system. To fuel it, you need to understand just engagement data and transactional data from the app at the device level when the user decides you can have my device ID. So when that happens, a third-party app or an owned app is sharing the exact same data to our system. And the disconnect there back in the day was most of our data was our own. So it was just about quantity, not about quality or differentiation of data. And now our own data is just such a small sliver of the overall. It changed the outlook.

Got it. And can you expand on what changes we can expect that the apps portfolio as a result of this? You mentioned the profitability before, what you'd be looking for in growth. And assuming a scenario where the portfolio is largely the same, what would that ideally look like 3 years from now?

I mean, look, ideally, if you're running your business, you want to make money and grow it. So that's the best-case scenario. Another interesting scenario for us is if we can convert into cash because it's not strategic for us or parts of it into cash, that would be interesting as well. It will give us more to invest into, whether it's a buyback like we've announced of our own stock or acquisitions in the software-related spaces, those things could be interesting for us in terms of expanding the things that we're really focused on as a business strategically. But today, as we look at how we're operating it, we're just focused on the take it and restructure it so that we can cut out the things that we're ineffectively running there with the goal of just reach and make it much more of a cash flow business.

And then considering each of those 2 scenarios, right, one where you keep the portfolio on where you don't, where you do keep it, how do you think about kind of the levers to drive margin higher? Is that all just kind of on the UA side and reducing that? Or are there other things like in development?

There's 2 costs, and we always over indexed on user acquisition. Most gaming companies, their #1 cost is how much they're spending on user acquisition. The mobile gaming business is entirely driven by user acquisition. So revenue is directly correlated to how much you're willing to spend on marketing. The second line item is how much you're willing to spend on R&D or pipeline of games. And we talked about how we over index on both. We have 3,000 developers building games for us. We always overinvested there because of a goal of just put as many titles into market to get reach. And so now we can optimize the cost down both on head count globally as well as on user acquisition investment. And the toggle on user acquisition is just we'll be willing to take more of a margin on the buy. So if we increase the target ROI that we want to achieve from the user acquisition dollars spent, we'll spend less. We'll make more margin dollars than we would have traditionally gotten when we justified the investment because of the data play.

Got it. And then on the other side of it regarding your strategic review, how do you plan to kind of evaluate the portfolio in terms of assets that you would retain versus sell versus spin, right? You mentioned kind of interest in all the different structures. And then you've been one of the more active buyers of studios over the past few years. How does that kind of inform your view of the potential market?

Yes. A couple of quarters ago, we signaled we weren't going to buy any more and that the data was a critical mass. And at that point, it hadn't reached the tipping point where the vast majority of the data in our system was third party. And so now we're signaling that we're certainly not acquisitive. This isn't obviously a seller's market. And so if we get wave of magic wand and find a great multiple for that business to convert it to cash, we would certainly take a look at that. Our goal with it is to maximize shareholder value. So we believe we can do that by maximizing the amount of money that these assets generate and then long term, if we find homes for them with companies that are explicitly and strategically focused on running games, that's great. It will expand the studios, and these studios will be software clients. So the other interesting financial leverage we get on the software side is, today, we have intercompany elimination on the spend that these studios have in our software business. And if we separated them and sold them off to a third party, we could actually invest in them and grow them. all those dollars of investment would be recognized as revenue and then presumably be 100% flow through to the bottom line on the software side as well.

But even absent that, right, that pullback in UA spend, the revenue for the software segment was already kind of moving towards that TSTV, right? I would imagine that would accelerate now.

Yes. The 2 numbers all compress to a steady state, but there's going to be a differential, probably going to be 9 figures, quite substantial in the scale of the software business that could eventually report as software revenue if we didn't own the games.

Got it. I want to shift to software. This past quarter, excluding the contra revenue impact related to the MoPub transaction, run rate revenue for the segment, I think, was around $1.3 billion. At a high level, can you kind of bridge that number to the $2 billion that you're guiding for, for 2023?

Yes. When we look at the 2, we look at the different components of revenue that we now have, which are a little bit expanded off of where we were, even if you look at Q4. In Q4, we exited the year at $1 billion run rate on software. Now we're at $1.3 billion. But just using the Q4 comparison, we had AppDiscovery, which we think of as our ad network, our DSP on behalf of advertisers, all net revenue reported, so it's our margin. The gross spend through the platform is materially larger than it. But of the $1 billion, $900 million of the contribution was from Discovery and $1 billion run rate. And then $100 million was from Adjust or roughly $100 million from Adjust. We then get to Q1, and then going forward, as we end up with 2 incremental software revenue streams, we bought a company called MoPub. We turned the service off in the marketplace 331, transitioned all the customers over. MoPub was a couple of hundred million dollar a year business owned by Twitter, and the revenue generated there came from demand-side platforms that we're buying through the MoPub marketplace to reach audience. That business model is one where you take a company like a Trade Desk that we announced a partnership with or a Twitter Audience Platform that we've announced a partnership with or even a company like LinkedIn, and they want to access audience. On our MAX marketplace, we have 700 million daily active devices that we're servicing. We are the trading floor effectively for mobile in-app advertising and we are the majority of the market. And so these companies are now coming into the auction. And when they spend $1, we're able to take a 20% transparent cut from that dollar. That's an incremental revenue stream that we didn't have before. So we talked about MoPub contributing over $200 million of run rate revenue by the end of the year. So you can tack that on to the software business. And then the other piece that we will have that's materially expanding is MAX's fee for any company that's an ad network to buy on that exchange in a real-time basis. As the auction, we mediate both ad networks. That includes Google, Facebook, AppLovin and Unity and others. We mediate these DSPs, Trade Desk, Twitter, LinkedIn, et cetera. And so you end up with these 2 constituents. The ad networks have both side relationships, supply and demand, so they only pay us 5%. But we end up with 5% incremental revenue there. That also will be a few hundred million dollars a year of business. The entirety of the auction we've talked about will be $10 billion a year run rate by the end of the year. So if we get there, it's $10 billion next year. Some portion of it will take 5%. Some portion of it will take 20%. Some portion of it, we take more than 20% on our AppDiscovery product, and the sum becomes quite compelling against the 2 based on where we are today.

And regarding the 5% opportunity, can you maybe say what percentage of publishers are out there kind of adopting the [indiscernible] app solution now so we can just kind of level set towards the goal?

Yes. So the market itself is probably around 50% to 60% in [ app ] bidding. Publisher adoption is forced by us. So it's more platform adoption, and that's forced by us desiring to get the market to header bidding. We obviously want to charge for the services we provide. We want to bring efficiency to the market as well. It also moves forward by ad networks, inclusive of Google and others, to adopt header bidding. And most companies in the market are now in a trial phase of moving toward header bidding. We do think, in the next 12 months, the entirety of the market will get there. So it gives us confidence that next year, for at least most of the year, the components of that $10 billion auction that we don't touch with DSPs or ourselves will be all taxed at 5%. That ends up as a substantial business. Now the other interesting thing that I think investors didn't understand is all the software revenue other than a just and now world, which we can talk about later, is all net revenue. This is all margin effectively. So we size that, and I think it was just a misunderstood number on the flow-through of the business. 70% EBITDA margins roughly on the software business. The only costs we have against these dollars are servers and limited team.

Got it. And just to go back to what you said before on the DSPs, right, the 20% they pay, does that number include the 5% on the [ in-app ] or is that?

No, no. So the 5% to a Facebook or Google, a 20% to Trade Desk. So 2 separate fees.

Got it. Okay. And there were a lot of questions on it on the earnings call. So just -- with regards to the contra revenue, can you just cover this for those that are kind of less familiar with switching mediation providers and kind of the impact that has on serving ads to certain types of devices?

Yes. So I'm not going to get an accounting nuance.

Not the accounting, more -- yes...

Want but more than -- we ended up confusing. But in the quarter, we -- by shutting down MoPub, publishers had to migrate to our platform. The service had been existed for a decade. A lot of devices over time and downloaded apps with MoPub on them, servicing publishers, there was around 1,500, 2,000 developers with 50,000 apps that had to move over in 90 days. The issue there is the developers can do the work. But if a consumer doesn't actually update the app, they're still going to remain on MoPub. And all of us probably have new devices, new operating system. App update is automatic. Old devices were not automatic. User had to manually go in and update. So say 3% of users never update the apps. Well, 3% of users today are no longer monetizable. MoPub was a substantial business. So if you have an app business and you're generating $100,000 a day of ad revenue that's very sustainable, and you lose $3,000 a day, there's terminal value to that loss. We didn't want to start the relationship with all these publishers saying, "Hey, you're on the hook for this loss, come over to our platform." So we just quantified it for all the big publishers, paid it out in a bonus. That bonus onetime access contra revenue, works against the quarter's revenue number but is not a go-forward number and won't be broken out nor is there any expectation that there'll be material bonuses in the future. And we've never had those in the past as well.

Got it. Ones that kind of stood out to me from your earnings was that 90% plus of MoPub publishers migrated over, and that was despite what we observed to be some concerted marketing efforts from your competitors during that process. Were you surprised by that conversion? And then for those that have migrated over, what are they generally seeing in terms of relative yields?

Yes. The interesting thing with MAX was we came to market, I think it was around 3 years ago. MAX had already grown to become the largest in the sector in terms of publisher adoption. Publishers had to come from somewhere. And the only other 3 solutions to service publishers at scale were MoPub, ironSource and Google. By bringing in MoPub is then number two, we put the majority of the inventory on one platform. So we already had the market's leading solution. We enhanced it by bringing some MoPub development functions into the new platform, the unified platform. We also added material amounts of demand, things like Twitter Audience Platform, Trade Desk, LinkedIn and others. And so when you expand the demand, you introduce increased liquidity, it ends up an even better solution that was already the best one in the market. So we ended up converting a lot over. Efficiencies go up in the short term and long term. As users adapt to -- the publishers adopt to using the new technology, there should be expansion of business for them and then subsequently for us in the coming months and years.

And for the publishers that see the higher yields, like how much of that then goes back into user acquisition? And then what kind of opportunity does that then create for AppDiscovery?

Yes. So that's really the hook into our business. What's interesting about our advertising business is the vast majority of our advertising dollars on the AppDiscovery side come from our MAX publishers. So these publishers tend to be casual games. Puzzle-based games is a big category. Think a game like a word game. Word game goes on to MAX. Say they generate 10% more lifetime value because the system is more efficient than what they were generating before. That 10% all gets reinvested back in the user acquisition on top of whatever base they had. And it's really easy for these developers to then plug into our AppDiscovery product. We end up their largest partner, both on the monetization side and the growth side. And then now that we have 2/3 roughly of the in-app ad ecosystem on our platform, it creates this strong closed loop, where our most important customers are app developers. They're performance buyers. They're buying on AppDiscovery. We've sized AppDiscovery on a net revenue basis, but on a gross revenue basis, billions of dollars a year are being spent through the platform. And most of these businesses are simple puzzle apps. They're lifestyle businesses, they're developers that are running profitable businesses. We're powering the biggest chunk of that business for them on both sides. And so that creates a very sticky formula that coincidentally, and relevant to today, fairly shielded from the macro economy is some of the trends that we're seeing in advertising because our customers might end up a 5-person development shop out of Turkey that spends a ton of money on our platform on a performance and arbitrage basis. These types of companies will spend $1 to make 2 all day long in any economy. And the consumer of these games isn't paying much to play these games. These are free-to-play accessible global forms of content that are just relaxing.

Got it. And for the publishers that migrate, you kind of talked about the relative technology before. But as you noted, I think in the past, MoPub wasn't a kind of a core product for its prior owner, right? What are they finding in terms of features and level of service impact?

I mean, we better do it better because it is our core product. So that was one of our thesis around it. That's also why MoPub -- and we're close to some of the executives over there for a while, brought it to us. The goal was as they were looking to exit the business, put it in the hands of someone who could do more good for the ecosystem. It's important for their business, too, as it is for ours, that these publishers do well. Many of them are advertisers that Twitter's own audience products as well. So I think we'll do a good job with that hopefully.

And can you say since March 31, when MoPub shut down, how much of that demand that was previously there has been kind of cut over with the supply that migrated?

A lot of it is ramped and a lot of it will take time to ramp. So some are just automatic. They plug in from 1 exchange to another. We've got twice plus the amount of scale if we were bigger than MoPub, think of it as more than double to access. So some will just ramp up. Some will take time, and that's because -- so for example, a Trade Desk is hooking up plumbing, but they need advertisers on the other side to be sold through to go access this audience. So we have a sales force. We go sell to the agencies. Agencies say go spend some dollars to AppLovin's marketplace. NIO comes through. We take your tax, they get their cut the agency spends the dollar. So that takes time to ramp up. And that's a good growth factor for us. Brand dollars are something we've never touched before, and omnichannel nongaming dollars are something we've never touched before. So to us, this is a new TAM, new opportunity. but we sit on top of 700 million daily devices that are playing games being monetized by the MAX platform. So it's a substantial audience and it gives us a good opportunity for growth in the future.

I'm sure you received questions about this today about possible ecosystem changes. Obviously, it's topical at the moment, just kind of given Apple's Worldwide Developer Conference in a few weeks. I guess much of the investor focus we get is on potential changes that would impair the ability to either attribute installs or actions based on an IP address. I think that's based on solutions Apple already has around what they give to iCloud subscribers for Safari. Can you provide your view of how AppLovin's potentially positioned for this change? And is this something you've contemplated in your software outlook?

Yes. So I mean, starting with that, yes, we don't think it's very impactful to our business, and I'll unpack why, but we've factored it in. And by not thinking it's very impactful, we're confident in our numbers. The reality is we have structural advantages in the ecosystem on the advertising side. We have the marketplace. We have multiple ways to monetize the marketplace. You end up with a market-leading product, 2/3 of an ecosystem, and you can monetize it anywhere between 5%, 20% and 30%. You end up with a substantial amount of dollars you're taking from a substantial market. Players are going to play games. That's just a fact. The other fact that it exists in our world are people who discover content through ads, and this is a lot of people, we've driven over 10 billion installs through our platform through ads over the life of the company. The people who discover content through ads are 50% likely to agree to share data for personalized advertising. So that's our audience. Our audience's consent rates are 50%. So that means 1 out of 2 users have an IDFA present personalized advertising, 100% deterministic attribution, no decay. Then you have a 50% of the audience that wants to go contextual. They don't want to share data to see ads. So IP address over there exists and matters for attribution so that an advertiser understands at a cohort level 75% accuracy, how did half the dollars that they spent perform. If Apple obfuscates IP address, we don't know what they're going to do. But if they roll out private relay the way they have done on the web, 20% of the audience will have an obfuscated IP address. So you end up with 20% of the 50%, 10% of the total that no longer will be able to run granular attribution. And so to us, that feels like a rounding error. Advertisers don't need precision. They went from 100% to half of it, went to 75% to now that 75% will go to 60% but you can extrapolate value. And so we don't look at it like anything material is changing in the ecosystem.

I mean -- and stepping back, whether it's platform changes or potential legislation like what we saw introduced last week, there's obviously a lot of flux within digital advertising. I guess for investors that might be focused on that dynamic, what's important for them to know about AppLovin, its culture, the way you position the company to react quickly to change?

Look, assuming we're all going to use our phones, people are going to play games, there's -- consumption is there, anytime there's a change in an ecosystem, someone wins, someone loses. There has to be how it shakes out. If the baseline is that the consumption is going to be there, someone's going to win, someone's going to lose. Usually, the companies that end up the winners are the fastest moving. We've always prided ourselves on having a lean team, very nimble technology and being able to move quickly. Also having the control of the ecosystem across our technology stack is a material advantage. Because we've got the marketplace and 2/3 of the ad ecosystem flowing through our technologies, we've got a lot of visibility to understand exactly what's coming. And at least today, we also have games. So we have all parts of the advertising stack covered in the in-app ecosystem. That gives us not only visibility into what's going on. It gives us relationships with the app stores on the other side. So we end up in an advantaged seat to be able to move fast, not to mention our technology team and our core technologies allow us to adapt as quickly as anyone else in the market, and we've consistently done that for years.

I want to touch on the macro environment. At earnings, one of your peers in the software space flagged some caution related to headwinds at mobile publishers, the economy. And then yesterday, we had one of the large digital platforms lower their rev guidance partly citing the economy. How would you frame the health of the mobile games or apps economy right now? And then given the industry never really existed during a traditional recession started in 2010, 2011, how do you assess that risk?

Yes. Mobile games has 2 parts. One is the in-app purchasing part that's very well measured and understood. Public data suggests $100 billion category. There's not much growth in that category. The reason is, is that once IDFA went away for half the audience, these publishers of content needed to go target purchasers on the other side of their marketing to grow those businesses. And now you're a year in, and it's much harder to do that. And so when it's harder to do that, you can't market it as well, you can't grow. That's the largest part of it, call it 80%. 20% of the market is the ad-supported casual part. That's the part that really is our market. And that's the area where you've got a casual game. They don't need our target purchasers. These games are free to play. I'd urge you to try to pay more than $10 in Wordscapes. It's really hard to do. And so these games are the most affordable, most accessible forms of relaxing entertainment in the world today, and they monetize with all parts of our stack. And so that part of the market is actually growing quite a bit more than the other side. Don't need to target purchasers. The other side becomes less effective at marketing. So the core puzzle casual games can market more efficiently. The other thing that ends up happening is the IP gaming companies need more revenue. They got to grow somehow if they haven't used ads traditionally, they might. And if you look at King from Activision, the rosiest part of their earnings reports for a couple of years were the ad-supported side. Zynga reports slowdown in growth on the IP games but the ad-supported part of the business is doing really well. And so that's what we're excited about because we built tools and technologies to service a very core part of the market that has now structural advantages that are going to continue to increase on both platforms. So we think we're in a very good place there.

And so did you just see kind of a gradual shift across the mobile games ecosystem or from that targeted hardcore game to casual to the hybrid model?

We're going to see it over time. Yes. A lot of these studios, too, it's like you're in development cycles that are 1, 2 years, and this IDFA has only been a year. So to change the development road map, it takes a while. And you're going to see over the coming years, if games are not as successful because they can't target payers, well, all this talent is going to shift to a much more casual mainstream type of game that fits right into our strengths.

Got it. We have less than 10 minutes. So if anyone in the room has questions, just raise your hand, and we'll get a mic to you. All right. You acquired Wurl about 2 months ago, which puts AppLovin into the connected TV space. I'm curious, how would you compare the connected TV ecosystem now to mobile kind of 10 years ago? What are the similarities? What are the differences in terms of both the hardware and the software side?

Yes. So upper level, I'm going to start with just talking about how we think about new initiatives and dive into what's exciting to us about Connected TV. We look at our business and we always traditionally are ready to generate cash flow. It's difficult to raise money private. It was easier to raise money a year ago in the public market. It's pretty difficult in the public markets today. But if you look at our current EBITDA multiple, at least after today, we're trading low teens on today's EBITDA. Our ability to generate cash flow is only going to compound from here. We don't have substantial CapEx. We don't have -- we have normal rate of taxes but the flow-through of a software $1 to cash is roughly 50%. So if we do the 2 next year, it's $1 billion of cash flow generation in the core business. So we think, foundationally, our core business, we made a lot of decisions over the last couple of years to structurally strengthen it, to have a market-leading position and be able to grow for years to come and generate a lot of cash doing so. So we have a baseline that's really strong. Then we go in new initiatives. Okay, this is a multibillion-dollar business. what we do over here has to be big. It's not going to be quick. Any new initiative, I'm trying to build businesses, you think about them an MVP state to gradually growing, these are 3- to 5-year bets. And the goal with these new initiatives are multibillion-dollar potential bets. They can pay off in 3 to 5 years and complement the growth that we'll see from the core business and the cash flow generation from the core. So when we looked at CTV, we see a big screen. We see video advertising, and we don't see performance ads. We see a lot of brand advertising and a lot of unfilled inventory. That's exactly how mobile in-app was when I started the business a decade ago. And so the opportunity feels a lot similar. Wurl as a potential acquisition and partner for us was interesting because they're sort of the CDN for these brands that have content, so take a Bloomberg or an AMC. They need to schedule their shows, put it in a channel directory and stream it across a whole bunch of devices on the other end. Wurl makes that happen. So they have supply, 300 million devices, an immense amount of hours played watched on a daily basis. That supply gives us access, and we've got the demand side platform, and the difficulty going to Connected TV for performance has been there's no attribution. Well, we also have an attribution company [ one of the ] leading and Adjust. So we're trying to lay the pieces to go execute on this idea. Now no one's done it. So therefore, it's not trivial to do. If we can do it, it will unlock billions. If we can't, we've got a very strong cash flow generation off the base of the business. So all these new initiatives that we talked about are option value for a huge return, but the baseline should be a substantial return in itself.

And can you talk a little bit about the adjust role on that. I mean we saw that they -- I think it was a few weeks ago, rolled out a new product, CTV AdVision, right? Is that sort of just the first step to developing that measurement process?

Yes. Like we look at attribution like why was it just strategically important to us. They're at the front door to any marketing software business. You have attribution that an advertiser is to hook up and then you can go market yourself. Just being able to run attribution on Connected TV starts the funnel. And so now that we've got that, we go connect our Ad Tech into Wurl. We've got the supply. We've got the attribution. Can we then make the advertising work attribute back properly and use similar models to what we've learned over a decade in mobile.

And what kind of is the basis of the model, right? Like initially, the network that you had was built off context. Someone downloads one game, that -- then you build the app graph, right? Within television, kind of what does the model look like as...

There's a couple of models. There's one is the exact same. You have 2 screens that are connected screens, and you can understand if the audience is similar on the 2, you can get a similar recommendation on the 2. The other is context on the television device. Users the watches Show A, -- I don't want to much TV. So I'll just use Show A as an example. User that watches Show A is interested in certain kinds of apps. And once you get the data on one side, you can combine it with the other data and the machine learning can run personalization through that. And so you can understand, okay, someone's watching a news channel, they also like these apps. And once the data sets merge, our model should be able to output recommendations. At least that's the theory. And again, we're not competing with anyone here. So if we can make it happen, the only competition is brand advertising and brand advertising cannot fulfill every impression that's available on television, so there is no performance fallback. And by not having one, the market is clear for a company to step right into.

Another area for investment you previously noted is in the blockchain gaming space. There's a lot of private capital flow into this area to produce games, which we would describe as varying in quality and not always the most player-friendly. I'm kind of curious what is it about this area that interests you? And what kind of broader adoption do you see for play-to-earn model over time?

Yes. We look at it the NFT side and the blockchain side is interesting. We want to be able to build software to enable developers to give their content ownership to the consumer for the purposes of trade, repurchase, whatever it may be. If you think about $100 billion in-app economy, today, the consumer is playing games, immense amounts of spend. They might buy a gun, they might buy a car. Whatever they buy in the game has no value to them in the real world. Why isn't that car getting a depreciated value and resold to someone else who could accelerate their progression into the game? The reason it's not is because there is no digital asset identity today, in mobile gaming at least, and there's no marketplace to allow those kinds of transactions. So as a software platform, what we'll bring to market is the ability for these developers that we work with to unlock that in their games. And fortunately, in our business, we've got thousands of developer relationships. We're already in a deeply embedded technology layer into the ecosystem. So it gives us advantages to bring something like this to the market that a lot of these well-funded start-ups around us don't have.

And you've also talked about getting into the OEM solution space. Maybe just talk about what opportunity you see there based on the current set of assets you have.

Yes, same thing like OEMs, carriers, they want to make more money off their devices. Today, there's a couple of players that are selling boot up screen, install flow, give them, I think, $1 per device or whatever it may be. We've got strong marketing systems that understand what a user should go download because they're predicting engagement and transactional value, so relevant advertising. If we can plug that into the life cycle of the device and give these carriers a much more native marketing experience, the expansion of that market could be substantial and our ability to power software for the carriers to generate, again, a multibillion-dollar revenue opportunity for us could be there as well. And so all 3 of these new initiatives, the same thing. If we execute, unlocks multibillions over the coming years. If we don't, we fall back on a very strong business that's generating a ton of cash flow as it is.

Got it. I think we've got time for one more. Maybe I'll just kind of follow up on the capital allocation part. When you look at your software portfolio today, are you kind of satisfied with it? Or do you see opportunities in the market just given a lot of the volatility we're seeing? And then how do you kind of weigh that against your share repurchase authorization?

I think we did authorize $750 million. We think the stock is well worth buying at this point, so we are buying as well. And we also look at it on the software side, if we're going to be able to generate $1 billion of cash for plus for the coming years, it will give us a strong asset to be able to expand that software business with. But when we talk about the new initiatives, we've been acquisitive on CTV. Core business, we've been acquisitive, but we have all the pieces we need. And then blockchain is so new. No one's executed on us, so we're going to build it. And carrier marketing solutions doesn't have incumbents that we feel like we can't build to surpass. And so the core stack is pretty hardened. So it will be hard for us to find opportunistically something that strengthens it. That said, we're always going to look to expand if we can into businesses that have the same profile of their current business. What's important to us is, if we're going to build something, it's got to be able to generate a lot of cash. We will always operate under the core competency that built this company, which is we build businesses and we cash flow ourselves. So regardless of market dynamics, we're able to exist and strengthen our business going forward.

Great. All right. With that, we're out of time. Adam, thanks so much for being here.

Thanks for having me.