APP Q3 2026 Earnings Analysis
AppLovin delivered record Q3 revenue of $2.18B (+55% YoY) driven by next-gen model rollout that accelerated gaming 50% YoY, expanded EBITDA margins to 84%, and generated $1.4B free cash flow.
Key Metrics
要点总结
- Model rollout drove gaming to ~50% YoY; consumer scaled triple-digit; record $2.18B revenue with 55% YoY growth.
- 84% EBITDA margin (+190bps YoY); $1.4B FCF; 87% QoQ flow shows strong operational leverage and cost discipline.
- Q3 share repurchases: 4.1M shares for $1.4B; increased buyback authorization by $3.2B to $3.6B total authorization.
收听平台
可用语言
Transcript
// Full episode scriptWelcome to Beta Finch, your AI-powered earnings breakdown! Today we're diving into AppLovin's third quarter 2026 results, and folks, this is a big one — record numbers, a reaccelerating business, and a massive buyback expansion. Before we get into it, let's handle the housekeeping. This podcast is AI-generated content for educational and entertainment purposes only. Nothing we discuss should be considered investment advice. Always do your own research and consult a qualified financial advisor before making any investment decisions.
Okay, Jordan, with that out of the way — AppLovin just put up a record quarter. Where do we even start?
Start with the top line, honestly, because it's wild. Revenue came in at $2.18 billion, up 55% year-over-year. And this wasn't just a beat — it was the largest sequential dollar step-up in company history, over $200 million more than Q2. Both revenue and EBITDA blew past the high end of guidance.
And remember, last quarter there was some hand-wringing about a "miss" and competitive pressure in the space. Tonight, CEO Adam Foroughi basically said, "Told you we'd fix it." They rolled out a next-gen model architecture — he compared the complexity to the Axon 2.0 transition a few years back — and gaming reaccelerated to nearly 50% year-over-year growth.
Which is the real headline here. This was a partial rollout. Adam was very clear: there's a lot more runway left to optimize, and the new architecture performs best exactly where data is thinnest — newer advertisers, smaller categories. That's basically a description of their consumer vertical, which is now growing at triple-digit rates year-over-year.
Let's talk margins, because usually when a company says "we're training bigger AI models," investors brace for ballooning costs.
Right, but that's the counterintuitive part. CFO Matt Stumpf said data center costs actually declined as a percentage of revenue, even as training compute investment went up. Their infrastructure team optimized how the models get served in production. So you got net income up 72% year-over-year to $1.44 billion — that's a 66% net margin — and adjusted EBITDA of $1.84 billion at an 84% margin.
Eighty-four percent. For context, management pointed out that the big digital ad platforms this earnings season grew somewhere between 10% and the mid-20%s year-over-year. AppLovin grew 55% — multiples faster — and did it at higher margins than basically anyone in the space.
And free cash flow was $1.4 billion, 77% of EBITDA. Matt noted Q2 conversion looked weak because of tax payment timing, so this is really the clean number. Full-year guide is still 70-75% conversion.
Now let's get to what I think is the most shareholder-friendly part of this call — the buyback.
Massive. They repurchased 4.1 million shares for $1.4 billion in the quarter — a big step-up from Q2 — and the board just tacked on an incremental $3.2 billion to the authorization, bringing the total to $3.6 billion. Since the program started, they've bought back 104 million shares for $9.4 billion at an average price of $90, returning 85% of free cash flow to shareholders along the way.
And diluted share count has now declined for eight consecutive quarters. So you've got 55% revenue growth with a shrinking share count — Matt flat-out said very few companies can pull that combination off.
One analyst, Clark Lampen from BTIG, actually pushed on this — noting the stock is still down about 40% year-to-date despite tonight's pop. Matt's answer was basically, "yes, we saw the dislocation, and that's exactly why we stepped up buying this quarter — and we can keep doing that opportunistically."
Let's hit a few Q&A nuggets, because there was a lot of good color. Adam reiterated the long-term growth algorithm — roughly 30% annually across the combined gaming and consumer business — but stressed it's one unified auction, so you shouldn't judge gaming and consumer in silos.
There was also a fun moment on competition. After last quarter's competitive scare, Adam basically owned it — said it was "our own foot fault," not the market getting more crowded, and that other ad-tech names reporting strong numbers isn't zero-sum for AppLovin.
And CTV got a mention for the first time in a while — still early, pilot campaigns only, not expected to move numbers anytime soon, but Adam said internally the team is "very excited" about the product roadmap there.
Don't sleep on the data commentary either. Adam talked about potentially pursuing data acquisition or tuck-in M&A — specifically name-checking their MoPub deal years ago as the template — buying customer and data relationships that could compress a decade of organic growth into a much shorter window.
So what's the Q4 setup look like?
Guidance is $2.365 to $2.395 billion in revenue, 43-44% year-over-year growth, and adjusted EBITDA of roughly $2 billion. At the midpoint, that implies full-year revenue and EBITDA growth north of 50%. Some analysts pushed back a bit, asking why the sequential growth guide — 8 to 10% — looks more conservative than Q3's 14%, especially heading into peak e-commerce season. Matt's answer: this is just their standard conservative guidance philosophy, baking in known model improvements and normal seasonality, nothing more.
So, bottom line for investors — a business that said "trust us, it's fixed" last quarter and then delivered receipts this quarter. Gaming reaccelerating, consumer scaling at triple-digit rates, margins expanding despite heavier AI investment, and a very aggressive capital return program behind it all.
Before we wrap — everything discussed is AI-generated analysis for educational purposes. Past performance doesn't guarantee future results. Please do your own due diligence.
The story going forward is really about execution on this model transition — Adam called it still a "partial rollout," so there's more to come — plus how far consumer penetration can run, and whether CTV becomes a real third leg of the stool down the line.
Lots to watch heading into Q4 and the holiday season. We'll be back to break down the next print.
Thanks for listening to Beta Finch — see you next time!