AppLovin reported Q3 2026 results on November 4, 2026 that stopped the debate about whether its growth was decelerating. Net income grew 72% year-over-year to $1.44 billion on record revenue of $2.18 billion, up 55% year-over-year, with a 66% net income margin and 84% adjusted EBITDA margin. Those margins belong to a software company running near-peak efficiency, not an ad-tech platform one model cycle away from its next architectural leap.
That is precisely what makes the Q3 result worth examining carefully. The company said it partially rolled out the next generation of its models during the quarter, and that performance was strong, with advertisers able to deploy more spend profitably at target return-on-ad-spend goals. It also said the gaming business grew nearly 50% year-over-year, reaccelerating alongside the rollout. Read that again: a partial rollout beat the high end of guidance by approximately $100 million. Management also said the full rollout of the new model architecture is still pending and it cannot commit to a specific timeline.
CEO Adam Foroughi compared the new architecture to AXON 2.0 in technical complexity, saying it gives the company “a much longer roadmap” for processing more data and training bigger models. That framing matters. AXON 2.0 helped scale the advertising business from the hundreds of millions in 2022 to roughly an $8 billion analyst-projected figure in 2026. If the successor architecture compounds that base at anything like the same rate, current consensus estimates are almost certainly wrong.
Scotiabank said the debate has moved from “does AXON work” to “how repeatable are the gains.” That is exactly the right question, and the answer is not yet settled. Q4 revenue guidance of 8% to 10% sequential growth implies a deceleration from Q3’s 14% sequential growth, which gives skeptics something to point at. Management’s consistent explanation: they guide what they can see, not what they expect the next model iteration to contribute.
The ecommerce vertical adds another dimension investors are still learning to price. AppLovin management has said the non-gaming advertising market is 5 to 10 times larger than gaming, and ecommerce is a primary expansion target. What remains unclear is how small ecommerce is today. The commonly-circulated “5% of Q1 revenue” figure attributed to BofA could not be verified in primary company materials, and AppLovin does not break out ecommerce as a standalone revenue line item in its reported financials. The directional point still holds: at a low single-digit share, a vertical management believes is orders of magnitude larger than its current base barely registers in the financial model. That is an asymmetric gap.
AppLovin generated $1.4 billion in free cash flow, repurchased and withheld 4.1 million shares for $1.4 billion, and increased its share repurchase authorization by $3.2 billion. A company allocating buyback capital at that pace while guiding Q4 revenue to $2.365 to $2.395 billion is signaling that management sees the stock as cheap relative to what a fully deployed next-generation model can produce.
The risk is timing. Foroughi noted that the full rollout timeline remains uncertain and that gains may not follow a linear quarterly cadence. Investors who bought the AXON 2.0 story early waited through quarters of uneven model improvements before the compounding became undeniable. The pattern may be repeating. The difference this time is the starting base is roughly $8 billion in annualized revenue, not hundreds of millions.
