AppLovin Corporation APP

Key data

The business

AppLovin sells mobile advertising. Its AXON engine is a self-learning system that decides which ad to show which person inside apps, then charges advertisers for the installs and actions it drives. After selling its own games studios, the Apps business, across 2024 and 2025, the company is now pure advertising: one software platform, almost no physical assets, and cost of revenue of just 12% of sales. The customer is an app developer, historically a game studio, who hands over an ad budget and gets installs back while the engine keeps a cut. The moat is a data flywheel: more spend teaches AXON, better targeting wins more spend. The newest move points that same engine at e-commerce and web advertisers, a market many times larger than mobile games.

The numbers

Growth is decelerating off a very high base, and the engine, not the ad count, is doing the work.

Year-over-year revenue growth slipped from 59% in Q1 to 53% in Q2, and sequential growth halved, from 11% to 4.4%, in a single quarter. That is the inflection. The tension flagged a month ago, whether price per install could keep outrunning the falling install count, held this quarter: revenue per install up 75% against a 10% drop in installs. Figures from Q3 2025 forward are continuing operations, the advertising business alone.

On the continuing advertising basis, revenue nearly tripled from $1.84B in 2023 to $5.48B in 2025, and net income rose ninefold, on an operating margin near 78%. Capital spending is under $5M a year, so operating profit converts almost fully to cash: trailing free cash flow is about $4.5B on $6.8B of revenue. Stock compensation is only 4% of revenue, so the cash the owner keeps sits close to what is reported, rare for software.

The market, having cut the stock 57% from its high, prices growth to flatten: analyst EPS barely moves from $25.83 in 2028 to $25.78 in 2029. The question is whether the e-commerce push is a second leg those flat estimates ignore. Revenue per install and the non-gaming mix are the prints that settle it.

Management

Founder-CEO (name and role not verified this run) controls the company through super-voting Class B stock. Over the last year insiders sold $126M across 85 transactions and bought nothing, even as the shares fell 57%; the largest were about $25M by Eduardo Vivas (role not verified) in June 2026, plan status not disclosed. The company repurchased $2.2B of its own stock in 2025, much of it near the highs above $500, a richer price than today's. The pattern, heavy selling into weakness and buybacks into strength, is the opposite of what you want to see from people who know the business best.

How it fails or surprises you

Price per install stops outrunning volume. Q2's engine charged 75% more per install while installs fell 10%. That works only while AXON's targeting improves faster than the gaming ad market saturates. If revenue per install decelerates toward the volume decline, growth stalls fast. Watch that one line over the next two prints.

The deceleration the read explains least. Year-over-year growth slipped 59% to 53%, and sequential growth 11% to 4.4%, in one quarter. At 16 times sales that slope matters. If Q3 sequential growth prints under 3% with no new revenue line named, the second-leg thesis is wrong and the flat 2028 estimates are right.

E-commerce advertising lands (right tail). AppLovin is aiming AXON at online stores and web advertisers, many times the size of mobile gaming. Estimates barely move past 2028, so the market pays nothing for it at $320. If non-gaming revenue becomes a disclosed, growing line, the 12x forward multiple is far too low. First print: management breaking out that mix.

Closing thoughts

Two prints settle this in two or three quarters: revenue per install tells you whether the engine keeps improving faster than gaming saturates, and any disclosed non-gaming number tells you whether e-commerce landed. The stock has already dropped 57%, so much of the plateau fear is in the price. An ambiguous print, price per install still rising but installs still bleeding and no e-commerce figure, leaves you holding a 25x-earnings business growing in the low teens, fairly valued and no more. The right tail is fatter than the de-rate implies if e-commerce shows up; the left tail is gaming saturation with nothing behind it.

The bet is still that AppLovin's ad engine keeps getting smarter and keeps charging more to reach each phone, and now spreads beyond games. What breaks it is two numbers moving the wrong way together: revenue per install decelerating while install volume keeps falling. When the price per install stops beating the volume decline, the story is over, whatever the multiple says.