All-In PodcastAdam Foroughi, Applovin CEO: Surviving a 92% Drawdown, Ads as ML 1.0 & the $50B Game Ad Market
All-In Podcast
What Applovin Actually Does 0:30
Adam Foroughi explains that Applovin is an advertising company that helps mobile game developers make money from their games. He points out that more than a billion people play mobile casual games every day, and that these are ordinary adults, not a niche audience. Almost two years ago the company disclosed that eleven billion dollars a year of ad spend moved through its own platform, and that figure has grown roughly sixty percent year over year since then, putting Applovin's own volume near twenty billion dollars. Counting the rest of the market, he estimates the whole mobile gaming advertising ecosystem at around fifty billion dollars a year, a size he compares to where social media advertising once stood.
Advertising As Early Machine Learning 3:31
Foroughi describes advertising as ML 1.0, meaning it was the first real proving ground for the deep learning technology that now powers artificial intelligence more broadly. He notes that recommendation systems and large language models follow similar research paths, and that researchers often move between the two fields. The advantage of advertising, he says, is that when a model predicts an outcome, that prediction turns into measurable economic value almost immediately.
Discovery Ads Versus Search Ads 7:01
He draws a line between two kinds of advertising. One type, like Google search ads, catches people who already know what they want and are just finishing the transaction, so it does not create new economic activity that would not have happened anyway. The other type, the kind Applovin and Meta specialize in, shows people something they did not know they wanted, which he calls pure discovery. On tracking fears, he insists Applovin does not track location or listen through microphones, and that the sense of being followed by ads comes from ordinary trackable actions like searches and site visits, not surveillance.
The Stock Crash And Comeback 10:30
Applovin went public in April 2021 at about 28 billion dollars, later hitting 40 billion, before collapsing to roughly 3.8 billion in 2022 even as the company posted a billion dollars in EBITDA that year. Foroughi attributes the crash to a mismatch between the investor base at IPO and demand once trading began. His response was to stop courting investors and instead buy back the company's own stock, eventually repurchasing about six billion dollars worth of shares, retiring twenty to twenty five percent of shares outstanding. He also describes the personal toll of that period, including family members asking if he was suicidal, and how he held the team together with a shared performance stock plan and an us against the world mentality. After the company shifted to a deep learning based advertising model in 2023, the stock rose from around nine dollars to seven hundred fifty dollars over two and a half years, taking the market cap as high as 250 billion dollars.
Privacy Rules And Staying Lean 16:00
On privacy regulation from Apple and the EU, Foroughi argues that clear rules let technology adapt, and that when Apple restricted precise targeting, users actually complained about receiving worse, less relevant ads. He explains that Applovin once bought game studios purely to gather training data for its first deep learning model, then sold them off once the model succeeded and outside developers began sharing data willingly. Looking ahead, he doubts that AI agents will replace typical shoppers, since most people still enjoy the process of browsing and comparing products rather than delegating it. He credits Applovin's ability to compete with giants like Meta and Google to staying lean, remaining deeply focused on one problem, and never assuming the company has already won.
AI-generated summary. It can be wrong or incomplete - check anything that matters against the original.
