Runs an AI engine trained on both sides of mobile ad auctions to make bidding predictions no rival can match.
Returns appear driven by leverage
AppLovin runs a mobile advertising business built around a machine-learning model called AXON 2.0, which sits between app developers buying users through its AppDiscovery platform and publishers selling ad space through its MAX platform. Because AXON 2.0 is trained on both sides of each auction at the same moment — seeing what an advertiser is willing to pay and what a publisher will accept — it can predict how valuable a user will be at the exact instant it places a bid, a calculation that a platform operating only on one side of the auction can never make. That accuracy is what keeps advertisers and publishers inside the system: advertisers see better returns on their spending, publishers get higher prices for their inventory, and each new auction adds more training data that sharpens the predictions further. The whole arrangement depends on volume staying high on both sides simultaneously — if either advertisers or publishers start leaving, the model loses calibration, performance drops, and the remaining participants have less reason to stay, which is a spiral AXON 2.0 itself cannot reverse once it begins.
How does this company make money?
The company takes a commission on every programmatic advertising transaction that passes through AppDiscovery or MAX, so revenue rises and falls with the total volume of ad spending on the platforms. It also charges subscription fees to app developers and marketers who use Adjust for attribution tracking and analytics.
What makes this company hard to replace?
Advertisers using AppDiscovery cannot move quickly because AXON 2.0's bidding algorithms need weeks to retrain on a new platform before they perform at a comparable level. Publishers using MAX have their revenue optimization wired into their existing ad stack configurations, and unwinding that requires technical integration work across their systems. Developers using Adjust face the most concrete barrier: the Adjust tracking code is embedded directly inside their mobile apps, and removing it requires writing new code, resubmitting the app to the app store, and waiting for approval.
What limits this company?
AXON 2.0 needs a steady, dense flow of auction data from both AppDiscovery and MAX to stay accurate. If volume on either side thins out, the model's predictions get worse. Adding more advertisers cannot fix a shortage of publisher data, and adding more publishers cannot fix a shortage of advertiser data — the model needs both signals together, from every auction, all the time.
What does this company depend on?
The company cannot operate without the iOS and Android app store ecosystems, which are the channels through which user acquisition campaigns are delivered. It relies on IDFA and Android Advertising ID for tracking and attribution through its Adjust platform. It needs access to programmatic ad exchanges and supply-side platforms for publisher inventory. AXON 2.0's computing runs on Amazon Web Services. And its connected TV ad insertion depends on Wurl's streaming technology stack.
Who depends on this company?
Mobile game developers using AppDiscovery would lose algorithmically-optimized user acquisition campaigns and would be forced back to manual bidding, which converts users at lower rates. Publishers using MAX would see revenue fall because yield optimization would become less precise and advertiser demand would shrink. App marketers using Adjust would lose the fraud detection and attribution accuracy they rely on to measure whether their campaigns are working.
How does this company scale?
As more auctions flow through AppDiscovery and MAX, AXON 2.0's predictions improve, which attracts more advertisers and publishers, which generates even more auction data — a cycle that compounds at very low added cost per new participant. What does not scale automatically is the engineering talent required to keep developing and improving the proprietary machine learning algorithms that power AXON 2.0; that expertise cannot be automated or easily hired in bulk.
What external forces can significantly affect this company?
Apple's iOS privacy changes have reduced the availability of IDFA, which directly weakens the attribution accuracy that Adjust provides to advertisers. GDPR and similar privacy regulations in other markets limit how much data can be collected for training AXON 2.0. When the Federal Reserve raises interest rates, venture capital funding tightens, and mobile app startups — who are a large source of advertising demand — have less money to spend on user acquisition.
Where is this company structurally vulnerable?
If a regulator or platform rule ever barred a single company from operating both a demand-side and a supply-side platform in the same auction — the way financial exchanges are sometimes barred from running both sides of a trade — AppDiscovery and MAX would have to be separated. The moment they split, each loses sight of the other side's price signal, AXON 2.0 falls back to one-sided training data, and the accuracy advantage that keeps advertisers and publishers on the platform disappears.
Structural observations derived from financial data, industry benchmarks, and supply chain position.
Companies that share the same coordination system — how they create, deliver, or capture value.
Companies that share active interpretations — structural patterns currently present in both stocks.
CompanyGraph