AppLovin's Adam Foroughi: Axon self-serve ad platform launches in two days — from 2,000 to millions of advertisers

numeral HQ to get started. Well, on that note, I believe we have our first inerson guest of the day. The day of the week from App Loving. Welcome to the show, Adam. Thanks so much for joining us. Thanks so much for taking the time to come on down to the TVP Ultradome. Welcome to the show. Here we are.

Much uh great to have you. Uh would love to have you kick us off with an introduction on yourself and the company. Just kind of set the table for us. Yeah, totally. We're probably the biggest company in the world that no one actually understands what we do.

That that's a good title thumbnail for some I guess it's bullish now if people don't understand what you do, right? Palunt. Yeah. Yeah. Same thing. So let's break it down. So we started the company in 2012 um advertising inside mobile games and really the goal back then hasn't changed to now.

We wanted to give advertisers the opportunity to market themselves but do it on a revenue based pricing model so that they can come in and say look I want to generate $1,000 of revenue by day 30 be able to get me $1,000 of media spend against that so I can break even make a profit. I'll scale unlimited if I can do that.

And I was I've been in ads for 20 years. I hated the construct of selling someone to convince them to run advertising. You don't know what works, what doesn't. You try to convince them an audience was there. It's all a bunch of makebelieve until you go to top offunnel revenuebased pricing.

And so we ended up working with a lot of game developers. These are companies all over the world, a lot of smaller indie businesses. And when you're working with a customer that where the CEO is the founder and they're on the front lines of marketing, it's a great dynamic because usually they're technical.

They want to actually understand the math around marketing. And so we built our business really being inside this niche, working with these casual mobile gaming developers all over the world. Well, business started growing a ton. We were serving a video advertisement inside their games for other games. Yeah.

And over the decade that we ran up up until going public, um, this business super lucrative, super successful for our customers. We helped these game developers really build their businesses, but no one had heard of us anywhere. We weren't VC, right? Well, not not by desire.

I'll say we're one of the biggest misses for Sand Hill cuz we we couldn't raise a million over four back in 2012. This was after we launched the product. Business was growing like a weed. We were we were doubling every single month for what was the what was the push back?

Do people just think like oh Google will take it all? I think I'm a horrible seller. I don't know. I I mean I've worked on the skills over the years, but it it the push back was Google will eliminate you. If it's not Google, Facebook will eliminate you. If it's not Facebook, here comes Amazon.

So it was why is this goofy named company going to be able to do well inside advertising? Really quickly, you said you've been selling ads for 20 years. 2005 is the start of this. What's the first ad you sold? We we were we were affiliate marketers back in social.

So a couple of us one of and that's how that's the most hardcore marketers started out affiliate because it's just you you eat what you kill. If you can make money in affiliate, you can make money doing anything. It is brutally tough. That's amazing.

Uh and and what was the mood in Silicon Valley or the the actual games that were uh the backbone of the launch? Is this the Zingga era? This is the tough part. It's a lot of casual games, right?

Like the the environment we have now is a billion plus daily active players playing solitire and puzzles and words match three. Yeah, totally puzzles. So if you walk down an aisle of a plane, you'll notice like puzzles on everyone's phone. Those are the types of people playing.

It turns out it's a lot of middle-aged women. It skews female. It's a really strong audience of people, but a lot of us don't relate to that. So you think mobile gamer, well shooter games and like traditional gamer.

And so that was always one of the things we had to to that was a challenge in the business is how can that audience be good for especially as we get into e-commerce and talk about our launch for broader brands because people they control so much purchasing. I was about to say it's all the heads of households.

We're on the platform. We only have adults. We we don't work on any child apps. So you've got adults head of household skew female people with money. These are people on a thousand phone playing for 45 minutes a day. It's perfect audience. People just never realized it until we became more mainstream. Yeah.

How do you think about there there's this critique going around with OpenAI, doing deals with Oracle, doing deals with Nvidia, and there's this like circularity to the economy, but then of course there's outside demand. People want to use the products.

How do you think about circularity within the mobile app ecosystem where a lot I'm I'm sure you have advertisers who are advertising mobile games and other mobile games and it feels like oh well like you know is this just all circular? when does the actual money come in?

H how how have you addressed that throughout the the the the history of the company and like where do we stand now? I mean again you got to find a purchaser at some point in that circularity, right? So if we were just advertising mobile games to mobile games that were ad supported be a house of cards.

But at the end of that you have two inside gaming two things you're trying to do. Take a user from a game where they're playing x time per day and take them to a game that they're going to be more engaged with and play 2x or 1. 5x. You can do that. You create more ad inventory.

So the real growth driver in the category is increased supply. It's not any different than social networking. You get more ads viewed, then you got to make more use of the advertising. So as the technologies get more powerful, you can match up the advertiser and the customer together for something that's transactional.

So you you get them going from a solitire to Candy Crush. They pay $2,000 a year. You've created transactional value in the middle there that there's a lot more value than where the user started. Yeah. H how did the company actually scale?

Like how how uh like meat and potatoes were the first deals and then I I imagine everything's like you know full tech platform now but were were the first uh advertisers who were onboarded on the platform kind of like a handshake and a contract or uh how do you actually inventory?

How do you solve video in your app at this moment and we'll give you you know Yeah. How much of like doing things that don't scale?

There was a ton of wheeling and dealing actually cool being in your office because we were a garage made into an office and we had a gong and like it was a bunch of one of the guys who runs my business team and now runs growth slept in the office for the first six months. So we were affiliates, right? Wheeler dealers.

So you'd go up to the early game developer and just give them a value prop of you can actually buy an install. Back then there was not even the chance of buying an install. Now I said sell them revenue. You're talking about 13 years ago. Being able to sell them an install on a pricing model was innovative.

And so when we went out to the market and pitched the Zingas of the world on instead of just buying ads and not really knowing what happens by installs, that in itself was, oh wow, that's great. And then you pair that with video. The power of the video clip on a mobile device is really understated.

35 seconds of average viewing time on our ads, you can't do anything else. So, when you get a brand opportunity to place an ad on a mobile device and the user's not distracted, non-s skippable.

Well, it's a mix of skippable and non-s skipable, but half the ads the user actually, it's a great point, opt into watching the ad for currency inside the game. That's up to 60 seconds of viewing time unskippable. You can't do other things. So, we live in add-rich environment.

This is the most nonad thing anywhere in the world. You've got to watch that advertisement. Yeah, that's fascinating. Um, how do you think AI is going to change mobile gaming? I I've seen there's these famous ads on Instagram where they show you this like character that's right. I'm sure you've seen these.

And then people are like, "But that's not the real game. It's a P three underneath. " And then I found someone actually built the real game because game development's getting cheaper.

It feels like we might be entering like a new there's a couple companies that have been pitching like vibe coding platforms for games that will uh you know like become their own app store. It feels like we could just be entering like a new primordial or Cambrian explosion of gaming.

How are you thinking about like vibecoded games or more custom software? Obviously they'll all need ads to power this but uh how are you thinking about I mean short answer to it for me right like like we we create discovery. So the more content that's out there the better.

So if AI creates an environment where anyone can build a game and we're all creative and most people humans now today have played games at some point in their life, everyone can write an idea down, create a game, they're going to need discovery on the game, they're going to need monetization on the game.

So so that's a great opportunity for us as an ad platform. More broadly, our system is predicated on our model actually working. So we we always talk about AI as LLMs, but if you just talk about AI as modern usage of neural nets, recommendation systems are one of the most economically viable use cases for AI today.

Yeah, they power Facebook, Instagram, Tik Tok, they power our system.

This is why we always felt like you know uh Zuck you know spending however many hundreds of billions like it was it was very easy to underwrite because you could just get there's a lot of ways to just get benefit from this infrastructure spend in the core advertising business right to totally you get I mean there's a simple function to all these models it's more data more complex model more compute better output and if your whole business model is built on an advertising business that this technology continue to evolve at the pace it is is fant Fantastic.

Yeah. How do how do you think about AI on the ad side?

Uh everyone's kind of predicting I mean there's already probably some tests out there that uh a lot of what Meta is doing will be uh just look at your product catalog, look at your inventory and then generate the video creative, generate the audio, generate the text. Uh how are you thinking about implementing that?

Are you already running tests? How how have the test gone? This this is one of the reasons investors are super excited about advertising platforms like ours. It's the the simplistic of you've got a shop, you've got a catalog, create an optimized appearance of that catalog through recommendation system.

We already do that. But more complex is today people upload a video ad into our platform. Not always adapted for the platform. We're we're new in the marketplace, especially in shopping. So, a lot of times they'll bring a 10-second clip from Facebook, upload it to us.

Well, if you could have gotten 45 seconds of viewer time, that sucks. As a brand, that's a total miss.

They don't have the resources at every level to go let us create a hundred ads a week for for this company now we'll rebrand as Axon and we can talk about that in a second but this Axon advertising platform they just don't have the resources to create that ad count at the scale of the time that you need into our platform and so when you can apply LLMs to go here's a couple cool ads out there let me create a 100 versions of this a week and put it into the system that'll be a huge uplift in user response to all the brands products that are being marketed on our platform.

Yeah.

When you when you think about uh we're very close with the the Ridge wallet team and so I got to early on in the Ridge days uh and I know that they they're an apploving customer, but early on I would just see the effort that Connor the CMO is just this constant treadmill of trying to make more and more creative, never being able to make enough, knowing that the more creative you make, the more money you'll make, but still just not even having the time in the day.

And so thinking about being able to multiply out creative across different user types, different games, being able to update it a lot more rapidly as games develop new features, bring back old users, etc. It's it's wild. Uh wanted to ask how you're thinking about uh M&A uh this year, next year, in the future.

you guys market cap is has um you know grown tremendously and you guys are now in a position where you could there's there's a variety of different platforms out there that that have a lot of attention that are maybe undermonetized or at least not monetizing as well.

I'm curious if that's at all something you guys think about or the opportunity in in um you know gaming is big enough that it makes sense. You're thinking specifically of like a Snapchat or a Pinterest or Yeah. or I mean Tik Tok was in the news last week.

you guys were uh you know and that felt like a transaction that was more of like a political transaction than a purely economic deal. But yeah, but I guess the broad question is like vertical integration.

Do you want to own games at some point or or where media is developed or social media platform or does that just make no sense for you? Well, I mean like we bid on TikTok, so the industrial logic of we can monetize inventory better with our ads models than anyone else out there is is sound at least to us.

But there's a couple things that constrain us. One, our culture is pretty unique. We've got very few people doing an exceptional large amount of output. Um, company is still under a thousand people. So, you just don't tend to see scale like ours at that little um headcount.

The and so putting companies together into that is really tough. The other piece is we're really excited about the organic growth opportunities in front of us. We're we're at a point now, I mean, I put out a couple months ago that early in the year we were 11 billion of ad spend on our platform.

Plus, now we've grown a couple quarters, so bigger number today. But and we only work with somewhere between a thousand and 2,000 advertisers on the platform. We've been completely managed. We haven't opened up the platform.

So, if you think about that, when Facebook opened up Ads Manager, they had five $5 billion a year revenue and we've got more than double that today on a very large platform. We haven't even opened it up. We're going to open up our platform, the Axon advertising platform in two days.

So once we take the average account from the very low thousands to the tens, hundreds and of thousands and then subsequently millions, we think the opportunity in front of us is many quarters to possibly years and that and so distracting you know it would have been it would have been cool to own Tik Tok.

It's a solid asset at least uh from uh from observing from the outside but the opportunity just immediately with the business is so massive that that it you know you may as well just focus on what you guys already do to totally and with a lean team we don't have the resources to distract ourselves.

So it's it's full speed ahead. We we want to give our solution to all the advertisers in the world. We know that we've made a lot of businesses inside gaming more successful. We want to do that across the board. We've seen it in a very small pilot in e-commerce over the last year.

The few hundred like Ridge Wallet and others that you you said um they're seeing a lot of spend on our platform comparable to what they're getting on Meta and it's on us to execute to get this in front of all the advertisers out there.

If we can do that and become another destination like Meta where there's a ton of scale and it's top of funnel to bottom funnel revenue priced all these businesses around us are going to grow. So we don't want to distract ourselves from that opportunity. How uh how are you rallying the team around Q4?

I'm assuming Axon is timed with the the the you know championship season for every you know retail brand but like what does it look is it going to be people sleeping in the office more I mean there's been a bunch of that it's around the clock right now because we've been closed and managed right so opening up an ad platform at our scale is a pretty complex undertaking you worry about like I mean fraud is an obvious case you don't want to see fraud you don't want to see scammy ads on your platform um that's the first thing that happens and like we're probably arguably the biggest ad platform that's ever opened up self-service in any sort of form um at this at this moment.

Uh and so when you've got that, you got to do a bunch of work building tools and controls to make sure the quality of advertiser and all the checks you have in place are as automated as they possibly can be and as good as it would have been in a human-managed environment.

So you've got that, you've got user experience, you need to make sure a customer can go live on their own all the way through to being able to pay you. And so we had to build a whole bunch of tools to get this right. We've been doing that for the last year. We've been tuning it with the customers on the platform.

Um so the team is no joke working around the clock right now to get this out the door in a couple days. Yeah. You can't you don't have the luxury of of missing Q4 if you want to open up an ad, you know, in the ad business. That's the most critical time. Uh how are you thinking about compute infrastructure?

I feel like we're seeing Meta build these massive clusters and it feels like maybe that's because they're going to be generating so many ads or doing so many so much Gen AI that they'll need a ton of compute. They still need a ton of core AI just to do ad matching.

Um, have you had to do uh on premise deployments, cloud stuff? Like what how do you think about uh building out your infrastructure for the long term and potentially like a very different just economic dynamic around how compute it like falls on the income statement. Yeah.

I mean, we work with Google Cloud, so it runs through our income statement. So, it's easy for investors to understand because we're not capitalizing the cost. Yep. um because we're so lean and we because we were bootstrapped maybe going back to the beginning.

We think about every dollar we spend anywhere across the organization as everyone's a founder, you're spending money out of your own pocket. And this holds true with compute as well.

If I I went into my team and said, "Look, we're going to go spend $5 billion, buy a bunch of compute, and in our scale, we can't do hundred billion dollars of investment. So, let's say five billion. " Um we we'd be wasting that money, and it would make them uncomfortable.

We try to constrain it to to trail or maybe be on par with revenue growth. So it becomes a cost of your future revenue growth. But what does that mean? If our engineers create a more complex model, they could consume more GPUs. They if we had unlimited capacity, they just deploy the model.

Well, in today's world, if they see more GPU consumption and they go, "Holy crap, this is fantastic in result, but is going to consume too much GPUs. We just don't have it. " They've got to go optimize their code. And you may get like you get it.

Okay, maybe we lose one point on accuracy, but the cost all of a sudden goes down 90%. So across the org, we're we're wired to think about things that way. Optimize and automate and cut costs as much as possible so we have the best business so we can fund our mission long term. Makes a ton of sense.

Partnering with Google makes so much sense given where they are on Nano Banana and where they are on Gemini and like just the cost per performance. Even if you go crazy into generating a bunch of ads, it feels like a great partner there. Uh how how did the Google will kill you like narrative?

It clearly didn't happen, but uh walk me through the various It seems like you were getting that criticism from venture capitalists early days. Uh how did that actually play out? Like how do you sit together in the market of ad buying because you do compete even though you work together on the cloud side.

Uh how has that relationship evolved? like why did the market play out the way it did and uh and was anything were there any decisions where you think like we made a really key decision?

If you if you talk to advertisers let's use the Ridge guys as an example again you want performance and you want to constantly be experimenting and you want you don't want platform dependency.

don't want in a perfect world you would be spending like 10% of your budget here 10% of your budget here 10% so that one that as prices fluctuate based because they're all markets right you can your business isn't you know you know and we saw the businesses that were you know in the DDC era that were dependent purely on you know a meta and and how bad things could get companies going bankrupt based on you know small changes in in price but anyways yeah I I mean Google is a is sort of a bottomfunnel channel for most dollars spent, right?

If you're selling a wallet, you go to Google, you search for a wallet, Ridge wallet appears there. Well, the consumer already knows they want to find a wallet. So, we were always going to be additive to the Google advertising story.

So, they did a really good job finding us early and betting on us across multiple business lines. And it became a really, really good partnership. Um, if you think about the business model we deploy, it's top of funnel to bottom of funnel priced on revenue. Facebook does that.

And if Facebook caught everyone's attention for every single ad um for every single product in the world where where there would be no other market, Facebook would capture all the topfunnel attention and Google would be bottom funnel. Turns out not everyone in the world notices an ad on Facebook.

So along we come with these really immersive ads, capture users attention and we were able to go partner with a wallet company and sell more wallets. So in a way when you do it this way they're they're getting an arbitrage. They're getting a customer that pays them immediately.

They ship the product after the fact and they make a spread on that sale. If they can do that, there's no budgetary constraint. So, we don't have a sales force at all really at the company because we're not going and asking for for budget.

We're going and saying, "Look, if we you can sell more wallets and make money on us, you'll scale to infinite as many wallets as you can go get in inventory. " And so, with that business model, we became additive to these other platforms.

I think really that's the only way a small brand or niche company like us was able to have the large scale success.

If you're if you're advising a I don't know public company CEO $100 million in market cap but they have a platform and they're not monetizing with advertising and they're considering rolling their own versus partnering with you or partnering with someone else. Like what are the trade-offs?

We've seen Uber's added a bunch of ads, Netflix added ads, there's uh GP is going to have ads soon. Um what are the different uh levers that a CEO should think about pulling? Have you done partnerships? How do you think about uh partnering with like those large scale new platforms? Yeah, we haven't yet.

I think at some point in our our narrative, we're going to get to a point where we're going to want to take our demand out, but we don't have excess demand with so few advertisers today. If we got 100,000 advertisers, we'd probably be able to take demand out.

But I think a lot of times people forget one, the technology is really complex. You've only had in terms of a a model like ours is scalable, us and then obviously Meta, very large scale that's been able to execute on this this discovery type of product price on revenue. Yeah.

Um, and so if there's only a couple cases, it's not trivial and people overtrivialize it a lot. The other thing is the ad format has to be similar to the ad format that we're really good at. Otherwise, it takes a lot of work. And because we've got such an immersive ad format, you don't tend to have that in other places.

Now, in television, you do. And that's an area that's that's been interesting to us. Um, you know, specifically like streaming or anywhere where you can see 15 to 30 second clip. Yeah. So streaming is a great great uh example.

Um pre-roll video is a good example, but a native ad that someone's scrolling by and probably misses or a little banner ad, those aren't going to create a lot of intent. So if the ad format isn't going to create intent, that's not a great matchup for us. But it's something we can execute on over time.

And I think that'll benefit the publishers because really these third party publishers, they shouldn't build their own ad tech for the most part. And they there's not another solution out there that's really that focused on them with Google maybe depp prioritizing their thirdparty ad stack for them. Yeah. Yeah.

What about Apple? I feel like there was a Rumble years ago about a you know a developer kit and SDK for adding ads into mobile games and uh it certainly hasn't stopped your growth. So what yeah how did that play out? Yeah, I mean they they have ads on search. They can monetize their own properties presumably.

I always say, look, if there's more ad inventory and more discovery, the P&L of the businesses in the market swells. So, let's say Apple rolls out a great ads product and you go to Apple Music, you discover a game. Well, you weren't going to discover that game before.

So, now that developer who will almost always be an arbitrage marketer, got a better arbitrage, their business, their P&L grows. What happens when that happens? Well, they have more money to reinvest in us. They have more money to reinvest in meta. And so if performance is there, that's fantastic for the ecosystem.

If it's non-performant, then that the customers are getting ripped off. And these customers in these categories, they don't have the business model or the brand to get ripped off.

I had seen a few pitches uh over the last couple years for businesses that were trying to effectively create apploven specifically for LLM type products. Why or why not is that an interesting opportunity? Because I would assume the app loving for an LLM is apploving.

You you probably never thought you'd be saying that statement. I think the day people started saying I want to be the app loving of this was when we knew we made it. I guess um like LMS they don't have a lot of space. So like you don't want to interrupt the user experience. You can't alter the output.

The referral program you guys were talking about a few minutes ago. Yeah. One one thing though is like a deep like if somebody put in like a deep research query somewhere waiting 10 minutes waiting 20 minutes accelerated. You really sit in front of everyone's bailing. Come on. Yeah, everyone's failing. True.

Yeah, that's a non viewed ad that's totally worthless. Yeah. So, yeah, I think it's going to be a lot more referral and embedded into the output and automated transactions, which again for us is good. If you think about dollars spent, a lot of dollars today go to search when the user already knows what they want.

If they go to the LLM and they they get an output that's something that they want and they can click purchase, all that's going to happen is those dollars are going to shift around from search to these LLM outputs. And that's good because that's already bottom funnel. The user knows that they want it.

And in theory, it's good because it's just going to be cheaper to the advertiser. So again, if it's effectively cheaper, that wallet company now has more money to go spend on a platform like ours where they're truly getting discovery from the customer. Yeah.

Uh talk to me about the company culture, the type of person that succeeds. Do you have economists on staff that are understanding the the nature of the market? Do you have a lot of data scientists? Is do you do you expect everyone to be able to think in statistical ways? Yeah. To so it's engineering first for sure.

Um research science runs the models. Those are those are typically like very intelligent math math people alongside they're also engineers. Um so they can implement their own research.

And then on the business team, most of the business people are math, econ, um they know how to do Python, SQL queries, like so we're we're mostly more technical across the board. Now core advertising business, our business and engineering teams are around 300 to 350.

So if you just slimmed it down to what drives our the 98% of the value of the the company, it's a very very small team. it. I always liked when I I listened to Steve Jobs talking about the original Mac team.

It's A players like to work with A players and then they they despise that like everything dilutes when you go to B and C's. Yeah. And we've got a really cutthroat ecosystem where you're either an exceptionally strong IC where you can output a ton or you're not going to make it.

And so we we end up churning through a lot of people who can't cut it, but the core team that remains loves working with each other because everyone's an A player. Yeah. Makes sense. How how do you think about, you know, what's uh what's your headcount plan look like over the next few years?

Is is it I mean I'm assuming you want to multiply revenue by a lot and without adding heads. I mean look, I'd love a world where we could run with 20 people, but we're probably not getting back to that world. But like I I remember Yeah.

I mean up up until 100 I knew the story and the name and every every everyone that was at the business and so we've surpassed that. But being run this lean allows us to maintain a really high quality.

So even our go to market team on e-commerce which every investor looks at as a huge opportunity it's 30 people today it's really hard to add a players if you maintain a high quality bar and you don't need to go into the state of look we've just got hiring quotas second you do that I think you blow up your culture in your organization so we're slow and steady now the the company's built on automate first hire second and so LLM let us do a ton most of the coding at the company is done by LLMs now a lot of the business functions are done by LLM.

So, we now are in a world where that culture is perfectly paired with this technology that's evolving so quickly. So, we should be able to do more with less. And if we can do that, we're going to be able to scale the revenue without having to add heads.

And then it's nice that the product's also so good, it sells itself. Well, thank you so much for coming by the TVPN Ultra. This is fun. Nobody Nobody loves ads more than maybe than than us than you maybe. So, yeah, the the the chat says because you're a gong respector, a gong ally, we have to hit the gong for you.

There we go. There we go. [Music] We'll we'll talk about Axon later this week. Excited. And in uh before we bring in our next guest, let me tell you about Finn.

AI, the number one AI agent for customer service, number one in performance benchmarks, number one in competitive bakeoffs, number one ranking on G2, and let me also tell you about Adio, customer relationship magic. Adio is the AI native of CRM that builds, scales,