All-In Sep 20, 2026 24m 10m saved
With Adam Foroughi, co-founder and CEO of AppLovin
In 2022 AppLovin's market value fell from about $40 billion to about $3.8 billion. In the same year the company earned a billion dollars in EBITDA.
A chief executive watching that usually goes on the road to defend the story. Adam Foroughi did the opposite: he told his team he would stop meeting investors altogether, and spent the cash the business was throwing off buying the stock those investors were selling.
"But guess what? We generated a ton of cash. Let's start buying our own stock. Let's become our best investor."
Adam Foroughi, co-founder and CEO of AppLovin, on All-In, built the company with no venture funding in its early years, took it public in April 2021 at about $28 billion, watched it fall 92%, and bought back roughly a fifth to a quarter of the shares on the way down.
The full interview is covered here so you can skip it. 24 minutes of audio, 14 minutes of reading.
Here are the 11 lessons that matter.
Key Takeaways
Over a billion people a day play mobile casual games, and they are adults running households rather than teenagers
AppLovin's own platform carried $11B of ad spend when it last disclosed the figure, nearly two years ago
The whole mobile game advertising market is roughly $50B a year, which is about where social advertising was not long ago
Advertising was machine learning 1.0, and the research now moves in both directions between ad ranking and large language models
Chatbot advertising competes with Google's search business, not with discovery — the transaction was going to happen anyway
The stock fell 92% while the business was earning a billion dollars, which he blames on who owned the shares at the IPO rather than on the business
He stopped meeting investors, bought back roughly $6B of stock and retired 20–25% of the shares
One week of investor meetings in September 2023 took the stock from $80 to $150
He does not believe agents will take over shopping for most people, because the transaction itself is the reward
Margins of 84% have not been competed away, which he puts down to model complexity and proprietary data
The company bought game studios only to get training data, then sold them all once third parties would share
1. A $50B Ad Market
Asked to frame the business for people who have never heard of it, Foroughi said the quiet build was a consequence of raising almost no venture money early, plus a name nobody took seriously. What the company does is sell advertising inside mobile games.
The audience is a billion adults a day, not a niche of teenagers
You've got over a billion people a day playing mobile casual games. These are all adults, heads of households, and the scale of the opportunity is just humongous.
Adam Foroughi
He built the market size up from the last number the company disclosed.
The platform carried $11B of ad spend at the last disclosure
We disclosed last January, so nearly two years ago, that on our own platform, there was 11 billion a year of ad spend.
Adam Foroughi
Growth of roughly 60% a year since then takes that to about $20 billion, he said, and AppLovin is not the only company selling in the market.
The whole mobile gaming ad market is about $50B a year
So then you'd probably more than double that again and round it off and say there's probably about $50 billion of advertising being spent every single year in this mobile gaming ecosystem.
Adam Foroughi
The comparison he wants is social advertising a few years ago
It was not very long ago that social was a $50 billion opportunity.
Adam Foroughi
The change he says has investors interested is that the same ad slot, which used to send a player from one game to the next, can now be used to sell that player a physical product.
2. Ads As ML 1.0
One of the hosts asked whether this generation of internet advertising is producing foundational technology the way Google's ad systems did. Foroughi said the value has moved to language models, but that the plumbing came out of advertising.
Advertising is where these methods were first made to pay
But advertising is a very profitable implementation of a deep learning model.
Adam Foroughi
Recommendation systems and language models are built differently, he said, but the work transfers.
The research crosses over in both directions
So a lot of the research that's being done in the space in the large language model space can port to recommendation systems and vice versa.
Adam Foroughi
The structural advantage of an advertising model, on his account, is the feedback loop: the model predicts a future outcome and finds out immediately whether it was right, which is not true of most machine learning problems.
3. When Ads Became Content
Asked whether people themselves have changed in how they respond to advertising, Foroughi compared the start of his career with now.
The ads of 2005 were spam, because the technology could not do better
I started my career in 2005, so I saw the ads back then, complete garbage.
Adam Foroughi
Now the recommendation is where the shopping happens
And if you talk to most people who shop today, most of their shopping recommendations are coming from Instagram. The ads have become very much like content.
Adam Foroughi
He said the same holds inside games, where playable previews of other games get real engagement.
People play the ads
And in our domain as well, people love the ads that we show.
Adam Foroughi
4. Discovery, Not Search
Asked what advertising looks like when people run five or six queries through a chatbot instead of a search engine, Foroughi split advertising in two. There is the bottom of the funnel, where the buyer already knows what they want and is completing the transaction, and there is discovery.
A chatbot ad business takes share from search, not from discovery
So that ads model is almost going to exclusively compete with the Google search business.
Adam Foroughi
His point is that closing a transaction the buyer had already decided on adds little to the economy: search would have completed it anyway. The other kind creates demand that did not exist.
Discovery is where the economic expansion is
But when you show a consumer an ad for something that they had no idea existed, they didn't know they needed to buy.
Adam Foroughi
That is the business he is measuring himself against
And this is what makes Meta so amazing in their ad business and what we aspire to do.
Adam Foroughi
5. The Mic Is Not On
A host put the common suspicion to him: you mention something at lunch and the ad appears that afternoon. Another host pushed further, asking whether shared location data lets an advertiser treat four people at the same table as a group and show them all the same watch.
Foroughi said neither mechanism is what is happening, and that his own company does not collect location at all.
Location tracking is not how the ad got there
That's what I'm told. I don't think advertising companies can track location. So, we don't track location at all.
Adam Foroughi
The explanation he gave is more ordinary: a search, a browse or a product lookup you have forgotten about, followed by a conversation on the same subject. Parsing microphone audio at that scale, he said, is not realistic. He allowed that a social network does have the relationship data, so a friend's search can shape what you see.
His closing argument on the creepiness question was economic.
Relevant advertising is now a measurable part of GDP
And there's a big part of GDP that's now coming from this digital ad economy. The better these technologies get, faster GDP growth.
Adam Foroughi
6. From $28B To $3.8B
The hosts walked him through the market-cap history. AppLovin listed in April 2021 during the run of pandemic-era IPOs.
It went public at about $28B and peaked near $40B
We went out in April 2021. It was about $28 billion.
Adam Foroughi
Then the share register worked against it. Private-equity holders, ex-founders and team members were all sellers at the listing, and with so many companies going public at once, he said, the large institutions never did the work on a company with a goofy name. No demand met a lot of supply.
The stock fell every day in 2022, while earnings went up
We got as high as 40 billion. And then in 22, the stock went down literally every day. We got to about a $3.8 billion market cap. In that year, we did a billion dollars in EBITDA.
Adam Foroughi
Asked how a chief executive holds a company together at that point, he described what the drawdown does to people outside the building.
The calls he was getting were about his state of mind
I would get phone calls from family members, friends, are you suicidal?
Adam Foroughi
And everyone on the team was getting the same calls, with less equity and less standing
But it's very tough then because you realize as a CEO, your team is getting those same phone calls from their family members.
Adam Foroughi
7. Its Own Best Investor
Foroughi's read, from what he described as a finance background, was that the price reflected the quality of the shareholder base rather than the business, and that a beating on that scale creates an opportunity on the other side.
He stopped selling the story and started buying the stock
But guess what? We generated a ton of cash. Let's start buying our own stock. Let's become our best investor.
Adam Foroughi
The buyback was deliberately aggressive
So we kicked off a super aggressive buyback program.
Adam Foroughi
Roughly $6B of stock, and a fifth to a quarter of the shares retired
Retired 20 to 25% of the shares outstanding. At peak that six billion was worth over 50 billion.
Adam Foroughi
Internally, he said, the message was an us-against-the-world framing, backed by a compensation change: a performance stock plan of the kind normally reserved for the chief executive, extended to key people across the company.
The plan went wider than the usual recipients
we implemented a performance stock plan which typically goes to CEOs but we did it across key people in the company
Adam Foroughi
8. ML 2.0 And A Week In NY
The recovery, on his account, was a technical change rather than a communications one.
The model moved from regression to deep learning
We went from a regression model to a deep learning model and the outcome was we're driven by our advertising algorithm.
Adam Foroughi
Because advertisers pay on performance, a better algorithm produces better returns for them, which means they spend more, which shows up directly as revenue. The model launched in April 2023 and, because he still was not meeting investors, nobody outside knew.
The new model shipped in April 2023 and the market had no idea
Now we turned into 23 and launched that model in April. We still weren't talking to investors. So people hadn't found out.
Adam Foroughi
By September 2023 the stock was around $80, the buyback was becoming less practical at the higher price, and he went to New York.
One week of meetings nearly doubled it
And in that week, the stock went from 80 to 150.
Adam Foroughi
He described sitting in meetings and being able to read the room as people called their desks. The round trip, on his numbers, ran from a $3.8 billion market value to $250 billion.
The share price moved from $9 to $750 in two and a half years
We ended up going from $9 to $750 a share in a matter of 2 and 1/2 years.
Adam Foroughi
His conclusion is that public investors are not a different species from private ones: most follow trends late, and the good ones are early. Damned if you do, damned if you don't, is how he put it.
9. Privacy Rules And Apple
Asked how he manages a business that Apple and European regulators keep tightening the rules around, Foroughi said clarity matters more to him than looseness.
Clear rules are what technology can build against
Look, in any of these spaces you want the regulations to be clear.
Adam Foroughi
When precise targeting is taken away, he said, the user gets grouped with others and served a worse ad. His evidence that users notice is the complaints the company received after Apple's change: people asking for more relevant advertising, not less.
Privacy rules and useful advertising are not opposites in his telling
So, there is this notion that you need privacy regulation so that technology companies can do exactly what's expected of them.
Adam Foroughi
Discovery is the part users actually want
On the other side, consumers do want relevant ads. It helps them discover products.
Adam Foroughi
The specific trade he described is a player watching 30 seconds of advertising to earn an extra life, which has real monetary value to them.
The 30 seconds gets spent either way
Now, if you're doing that, do you want to sit and watch garbage for 30 seconds or do you want to watch something relevant?
Adam Foroughi
10. Agents vs Window Shopping
Asked what advertising becomes when software agents do the buying and the screen may not even be a phone, Foroughi drew a line between restocking and shopping.
Agents will handle the repeat purchase
For instance, I might put my supplement subscription into an agent and have it optimized every single month and deliver on time.
Adam Foroughi
Discovery, he said, is a different behavior, and the people who assume agents will absorb it are describing themselves rather than the market.
The typical shopper wants the shopping
The typical shopper wants to find a product and wants to actually go through that shopper behavior. They want to window shop.
Adam Foroughi
Saving 20% does not beat the experience on a small basket
I don't think that matters on a $50 transaction because the dopamine hit from going through it is what they enjoy.
Adam Foroughi
His customer is not the early adopter on X
I sort of say like our audience is the New York Times audience. There's still a ton of people using Yahoo properties every single day.
Adam Foroughi
11. Lean Beats Big
A host asked the question the whole segment had been circling: how a small company beat Meta and Alphabet at advertising, when both have spent two decades and their best engineers on it.
The operating answer is refusing to believe they have won
We never think we won. We think every day we wake up and we're probably going to get screwed right now and we better work hard.
Adam Foroughi
Focus and headcount are the advantage
And so I think there's this ability to take on giants. If you're very focused, you remain lean and you can just move faster than them.
Adam Foroughi
Pressed on where the business leaks money, he said his margins are the best in the market at 84%, which leaves little to recover. The model he described is an arbitrage the advertiser runs: a seller of lipstick buys a customer from AppLovin for less than the item's price minus the cost of goods, the sale covers the cost immediately, and the advertiser scales up. AppLovin is not the merchant in that chain, and he is not trying to be.
Asked why nobody has competed those margins away by offering the same service at 60% or 50%, he gave a two-part answer about complexity and data.
Complexity plus proprietary data is what holds the margin
Because these technologies are really complex and if you can innovate and you have differentiated data, you can build an advantage.
Adam Foroughi
Scale of adoption is the moat, which is why he says Anthropic has not run away with language models
the power of a model that then reaches a point of scale and gets adopted by a large scale community becomes something that is a moat that is hard for other people to overcome
Adam Foroughi
Bonus Insights
The game studios were bought for training data and then sold
A host asked whether AppLovin intends to become a game publisher, and whether owning studios puts it in conflict with the customers it sells advertising to. Foroughi said the studios were never the point.
Yeah, we sold all those games. We bought them originally as a data play.
Adam Foroughi
So we bought our own studios, we seated the training data in our first model, we built a model that was really successful in market, we started growing really quickly.
Adam Foroughi
Once outside developers were willing to share data with the model, he said, the studios were divested.
Engineering runs out of Palo Alto, Beijing and Singapore
Foroughi is based in Los Angeles and the company is headquartered in Palo Alto. Asked about the China team, he answered with an assessment of the people rather than a cost argument.
Chinese people are very humble. They're very, very hardworking. They're very sharp.
Adam Foroughi
when I sit in a room with some of the people on my team, I know I'm probably the dumbest person in that room
Adam Foroughi
The same episode had Bending Spoons booked
One of the hosts framed a question about buying games by referring to Bending Spoons, appearing later the same day, and its strategy of acquiring slower-growth businesses that venture investors have lost interest in.
Foroughi's bottom line is that AppLovin's advantage is a deep learning model with data nobody else has, pointed at an audience of a billion adults a day that the rest of advertising treats as a games business — and that the moment the market priced the company at $3.8 billion while it was earning a billion dollars was the moment to buy the stock rather than to explain it.
Products, Companies & Tools Mentioned
AppLovin (His company. Sells advertising inside mobile games, ran a roughly $6B buyback through the 2022 drawdown, and now wants to sell physical products to the same audience)
Meta, Facebook and Instagram (His benchmark for discovery advertising, and where he says most people's shopping recommendations now come from)
Google and Alphabet (The search advertising business he says chatbot ads will compete with directly, rather than expanding the market)
OpenAI and ChatGPT (Raised by a host as the new ad product to learn from, on the premise that most users will never pay $20 a month)
Apple (Tightened targeting on iOS; he says the complaints afterwards were from users asking for more relevant ads)
Anthropic (His example of why a technically strong model does not automatically win: scale of adoption is the moat)
Amazon (Named by a host as the case study in a company absorbing the places its business leaks)
Bending Spoons (Booked on the same episode; the host used its acquisition strategy to ask whether AppLovin should own game studios)
Yahoo and The New York Times (His shorthand for the mainstream shopper his ads reach, against the early-adopter audience on X)
Listen to the full episode
🔴 YouTube | 🔗 Episode page
Watch the full episode:
More Adam Foroughi interviews and podcast appearances
If this was worth your time, send it to someone who follows the name.
Get the latest market chatter as it happens: