AppLovin's Q2: A Timing Blip That's Already Behind It

Model uplift landed just after quarter-end; consumer vertical hits records and Q3 guidance points to reacceleration.

APP · Earnings Call · 2026-08-05

The Quarter That Wasn't to Their Standard

AppLovin's second quarter 2026 was a rare miss against the company's own high bar. Revenue of $1.92 billion and adjusted EBITDA of $1.61 billion both came in just below guidance. The explanation was not demand weakness but a timing gap in model improvements. As CEO Adam Foroughi put it, “In Q2, we didn't have the same amount of uplift that we normally have in any other prior quarter. That came right after the quarter.” — Adam Foroughi · 2026-08-05 This is a familiar narrative — the company has consistently ridden model-driven step-ups, as seen in the model uplift keyword that has dominated its transcript history. What's new here is the explicit admission that the cadence can dip, even if the long-term trajectory holds.

In this case, in Q2, we didn't have the same amount of uplift that we normally have in any other prior quarter. That came right after the quarter.
Adam Foroughi · 2026-08-05

The market's reaction, though, has been severe: the stock has fallen about 22% in the last 90 days, with a drawdown of over 50% from its December peak. That suggests investors may be reading more into this miss than a mere timing issue — perhaps questioning whether the model performance improvements can keep up the pace. Yet management's confidence is anchored in what happened after the quarter closed.

Consumer Vertical: The Accelerator

The standout performer was the consumer vertical, which management described as having an "outstanding quarter." In prepared remarks, Foroughi noted that advertiser spend set a record, “finishing 28% above Q4 2025 levels.” — Adam Foroughi · 2026-08-05 This is particularly notable because Q4 is the seasonal peak for e-commerce advertisers, and to exceed it in a seasonally slow Q2 signals that the model is finding traction with a broader base of merchants — and that the consumer vertical is no longer a sideshow but a genuine growth engine. The strategy is deliberately focused on mid market advertisers, where the platform performs best today. Foroughi discussed partnerships, citing Triple Whale as an example, to bring in "very targeted customers" rather than opening the floodgates to long-tail SMBs who may not have the creative assets or the patience for a learning curve. This is a departure from the earlier “let everyone in” posture, and it signals a more disciplined approach to scaling demand as the model matures.

EBITDA Margin: The Compute Trade-off

The margin story this quarter was about deliberate investment. Adjusted EBITDA margin still expanded ~300bp year-over-year to roughly 83%, but the sequential increase in costs was driven by higher compute for training and new model development. CFO Matt Stumpf noted, “The primary driver of the sequential increase in costs was higher compute associated with training our existing models and with new model development.” — Matt Stumpf · 2026-08-05 Management has consistently framed these investments as yield-generating: every dollar spent on compute is tied to a return expectation. In the Q&A, Stumpf reiterated that the incremental spend is expected to remain around $0.10 per incremental dollar of revenue, and the long-term margin profile should hold in the low 80s. This is consistent with the company's historical pattern of investing aggressively when the math works. As a reminder of how far margins have come, the operating margin expanded by over 20 percentage points year-over-year to 78% in the most recent reported quarter, though that data lags the current period.

The Path Forward

The Q3 revenue guide of $2.055–$2.085 billion implies 7–8% sequential growth, above seasonal norms, and management is explicit that this “reflects the model improvements that are already live and performing.” The narrative is that the Q2 miss was a one-off technical timing issue, not a structural slowdown. Long-term, Foroughi reiterated the target of “compounding at roughly 30% annually,” driven by ongoing improvements in gaming and the expansion of consumer. Prior quarters have been filled with talk of breakthrough model jumps. In the May 2026 call, for example, Foroughi said, “The one we just had a couple of weeks ago was quite substantial.” — Adam Foroughi, Co-Founder and CEO · 2026-05-06 This quarter's more measured language — acknowledging that “we may not always have lifts in every single period” — may be a more realistic, if less exciting, view. The company is also now free of the SEC inquiry, which it notes concluded with no action, removing a potential overhang. The key question for investors is whether the market will pay for the same multiple when growth is acknowledged to be non-linear. The recent drawdown suggests some skepticism. But with the consumer vertical growing 28% above its peak and the model uplift already live in Q3, the company's own evidence points to reacceleration. The tape will be the ultimate judge.