Wall Street loves AppLovin (APP 2.53%) right now. The company last month reported 70% revenue growth in 2025, to $5.48 billion, and generated $3.95 billion in free cash flow. Its Q4 revenue hit $1.66 billion with a pretty remarkable 84% adjusted EBITDA margin, and guidance calls for another revenue step-up in early 2026. Following the report, it seems like every analyst has rushed in with upgrades and higher price targets.
The numbers are spectacular, but that's exactly why I think they're dangerous.
On paper, AppLovin looks diversified: It owns an AI self-serve ad engine (Axon), a mobile ad exchange (Max), and a growing e-commerce ad platform. Management talks about "omnichannel performance marketing" and "multi-vertical expansion."
Image source: Getty Images.
Strip away the labels, and AppLovin is really one thing to me: a performance advertising engine built around mobile attribution and user tracking. In 2025, its Software Platform -- basically ads -- grew 88% to $4.81 billion and drove nearly all the profit, while the Apps side, which looks more like gaming or content, actually shrank as a share of the business.
The catch? AppLovin doesn't control the rules. Apple and Alphabet's Google do. The company itself warns that changes to mobile OSs or privacy frameworks, such as Apple's App Tracking Transparency, can affect its ability to measure, target, and optimize ads. Workarounds exist, but they're just that: workarounds.
The bull case assumes the current setup continues to work. The bear case is that AppLovin's advantages sit on rented land. A new privacy rule, attribution change, or app store policy could shift the math overnight.
NASDAQ: APP
Key Data Points
AI moat or model commodity?
Management leans hard on the AI angle. AppLovin's Axon is billed as a "machine-learning, real-time auction engine" that picks creatives, bids, and placements across millions of auctions per second -- and it works: 70% revenue growth and 84% margins speak for themselves.
My big question isn't whether AppLovin's models work today; it's whether that advantage can last. The underlying technology isn't proprietary, meaning competitors could build similar models using similar data. In digital marketing and tech, I'm already hearing about new AI tools that do what Axon does.
My view is that soon, rivals may deliver comparable results at lower cost.
What truly matters is exclusive signal and inventory, and AppLovin doesn't have either. Like everyone else, it's essentially buying access.
Growth still leans heavily on its mobile gaming ad business, while its new e-commerce platform and AI ad tools remain unproven. Rising competition from Meta Platforms is a concern, and it's unclear if the Axon 2 model can sustain an edge in a crowded gaming ad market.
At recent prices, AppLovin stock trades at high multiples for a business so dependent on external rules -- around 50 times trailing earnings and a double-digit multiple of free cash flow, even after a big 2025 run. In other words, Wall Street is paying a "platform" multiple for a business whose economics can be rewritten by a few lines of policy text in an iOS or Android update.
I'm going to stay away.
