AppLovin grew Q2 2026 revenue 52.8% to $1.92 billion entirely on pricing — revenue per app install rose 58% while install volume fell 2% — with Q3 guidance implying growth slows to roughly 47%.
Revenue up 53% on price, not volume
AppLovin reported second-quarter 2026 revenue of $1.92 billion, up 52.8% from $1.26 billion a year earlier, with operating income of $1.49 billion and diluted earnings per share of $3.76. It was the company's fourth full quarter as an advertising-only business, following the June 30, 2025 sale of its mobile-games ("Apps") division.
The striking detail is not the growth rate but its composition. Management attributes the $664.9 million revenue increase "primarily to improved AppLovin Ads performance, where net revenue per installation increased 58%, partially offset by a decrease in the volume of installations of 2%." In other words: AppLovin delivered slightly fewer app installs to its advertising clients than a year ago, and earned substantially more money on each one.
Operating margin is the share of revenue left after the costs of running the business, before interest and tax. Adjusted EBITDA is a company-defined profit measure that strips out interest, tax, depreciation and amortization, and — importantly here — stock-based compensation; it is not a GAAP figure and AppLovin's definition may differ from other companies'.
Price-led growth: durable engine or a ceiling approaching?
Charging more per unit is the higher-quality way to grow, because it costs almost nothing extra to deliver. That shows in the cost lines: cost of revenue rose 46% to $225.8 million — slower than the 53% revenue increase — driven by "an increase of $52.1 million in expenses associated with operating our network infrastructure." Gross margin (revenue minus direct delivery cost) edged up to 88.3% from 87.7%.
But the volume figure deserves attention. Installations fell 2% year over year, and for the first six months the picture is starker: revenue per installation rose 75% while install volume fell 10%. The second quarter's −2% is therefore a meaningful improvement in volume trend against a first quarter in which volume fell considerably harder — while the monetization gain simultaneously cooled from the first-half pace of +75% to +58%. Both lines are converging toward a more ordinary growth profile.
AppLovin attributes the per-install pricing power to Axon, its AI bidding and recommendation system, which decides which ad to show which user at what price. As long as advertisers hit their return-on-ad-spend targets, they keep bidding more. That mechanism is real and has worked for several years. The open question is that it is a multiplier on a declining base of installs, and a multiplier cannot compound indefinitely against shrinking volume.
Geographically the growth was broad: US revenue (by user location) rose 50.3% to $989.6 million, and the rest of the world rose 55.6% to $934.1 million. The business is now almost exactly half non-US.
The Apps divestiture: what it does and does not distort
This is the comparability point that matters most, and it cuts in two directions.
It does not distort the revenue comparison. Because the Apps division is presented as a discontinued operation, the prior-year quarter was restated to exclude it — both the $1.92 billion and the $1.26 billion are advertising-only. The 53% growth rate is a genuine like-for-like figure.
It does distort the bottom line. Q2 2025's reported net income of $819.5 million included $47.7 million from discontinued operations, which in turn contained the $106.2 million pre-tax gain on the divestiture (after $18.3 million of transaction costs) offset by a $125.6 million write-off of deferred tax assets triggered by the sale's treatment as an asset sale for tax purposes. Strip that out and continuing-operations net income was $771.9 million — so the true underlying earnings growth is +64.1%, not the +54.5% a naive comparison of headline net income produces. The half-year comparison is distorted the opposite way: H1 2025 carried a $99.4 million loss from discontinued operations, including a $188.9 million non-cash goodwill impairment on the Apps unit recorded in the first quarter, which flatters the first-half growth rate (+77% reported vs. +65% on continuing operations).
A third framing is worth having for anyone comparing against pre-2025 headlines: the old AppLovin, including Apps, reported $1.57 billion of total Q2 2025 revenue. Against that, this quarter's $1.92 billion is +22%, not +53%. Both numbers are honest; they answer different questions.
One residual tie remains. AppLovin took 596.9 million Tripledot shares — roughly 22% of that company — as $285.0 million of the $715.6 million sale consideration, and the divested business is now a customer: AppLovin recognized $18.0 million of revenue in the quarter from "Tripledot and its subsidiaries' use of AppLovin Ads," about 0.9% of total revenue. Small, but it is related-party revenue and not fully independent demand.
Where the spending went — and why GAAP and adjusted figures diverge
Research and development rose 127% to $99.9 million. Almost all of that increase — $54.8 million of $55.9 million — was "personnel-related expenses related to an increase in stock-based compensation-related payroll costs." Total stock-based compensation more than doubled, to $85.8 million from $34.6 million, or 4.5% of revenue versus 2.7%.
This is the single largest reason Adjusted EBITDA grew faster (+58.5%) than GAAP operating income (+56.0%): the adjusted figure adds stock compensation back. Stock compensation is a real cost — it dilutes existing shareholders — so the 3.0-point expansion in Adjusted EBITDA margin overstates the improvement a shareholder actually captured. The GAAP operating margin gain of 1.6 points is the more conservative read, and AppLovin offset the dilution anyway through buybacks: diluted share count fell 1.5% to 337.0 million.
Going the other way, general and administrative costs fell 27% to $40.3 million, on "a decrease of $9.5 million in bad debt expense and a decrease of $5.2 million in professional services costs primarily associated with transaction-related expenses" — the latter being deal fees from the divestiture that did not recur.
Two items below the operating line also flattered the quarter. Other income swung to a $62.4 million gain from a $22.3 million expense, helped by a $31.3 million unrealized fair-value gain on private equity fund holdings (a paper mark, not cash), $21.3 million more interest income on a larger cash balance, and $12.0 million of foreign-currency gains. Against that, the effective tax rate rose to 15.9% from 12.7%, as a "decrease in stock-based compensation benefits" reduced the tax deductions AppLovin had been harvesting when its share price was rising fastest. That tax-rate drift is worth tracking; it is a structural headwind to EPS growth if it continues.
Cash and capital returns
Free cash flow was $863.3 million for the quarter and $2.15 billion for the first half, up 34.9% year on year. Note that free cash flow grew far more slowly than the 65% rise in first-half continuing-operations net income, because a $504.8 million net decrease in operating assets and liabilities absorbed cash — receivables build as revenue scales, and the filing cites "increased publisher payments, operational spending, and cash paid for income taxes" as offsets. Half-year cash conversion (free cash flow as a share of net income) was about 87%; the second quarter alone converted around 68%. Healthy, but no longer better-than-earnings.
AppLovin repurchased and retired 3.3 million Class A shares for $1.5 billion in the first half — an average of roughly $455 per share — with $1.8 billion of authorization remaining, and finished the quarter with $3.1 billion of cash against an untouched $1.0 billion revolving credit facility. Interest expense was flat at $51.2 million, so the balance sheet is doing no work either way.
Takeaway: The 53% growth is real and like-for-like, but it is entirely a pricing achievement layered on a shrinking base — installs fell 2% while revenue per install rose 58%. Third-quarter guidance implies that engine is decelerating on schedule: about 47% growth, down from 59% in Q1 and 53% in Q2, with guided Adjusted EBITDA margin of 83% below the 83.9% just delivered. AppLovin is converting from a hypergrowth story into a very high-margin, moderately-growing one, and the third quarter is the first period where management's own numbers say so.
Guidance and outlook
For the third quarter of 2026, management guided to revenue of $2.055–2.085 billion and Adjusted EBITDA of $1.710–1.740 billion, implying an 83% Adjusted EBITDA margin. Against Q3 2025 continuing-operations revenue of $1.405 billion, the $2.07 billion midpoint is 47.3% growth.
That sets a clear sequence: +59.0% in Q1 2026, +52.8% in Q2, and roughly +47% guided for Q3. The deceleration is orderly — roughly six points of growth shed per quarter — rather than a break, and the absolute dollar step-up implied for Q3 ($146 million sequentially, versus $81 million from Q1 to Q2) is actually an acceleration in sequential terms. But the guided margin of 83%, a step down from the 83.9% delivered, suggests management expects the cost base, chiefly the AI and engineering investment showing up as stock compensation, to grow at least as fast as revenue from here.
Our read: the business is unusually profitable and the cash generation is genuine, but three things determine the next several quarters. First, whether install volume stabilizes or keeps eroding — pricing power applied to a shrinking base has a mathematical limit, and the volume line is the leading indicator. Second, whether the expansion beyond mobile gaming into web-based e-commerce and connected TV, which the filing describes as markets that "remain nascent and may not develop as we expect," actually contributes measurable revenue; nothing in this quarter's disclosure separates it out, so investors are taking it on trust. Third, the rising effective tax rate and the doubling of stock compensation both quietly transfer value away from per-share earnings even while margins look excellent. AppLovin also flags real concentration risk: its revenue depends heavily on mobile gaming and on distribution through a small number of platforms — the Apple App Store, Google Play, and Meta — where Meta and Google are simultaneously significant partners, so a policy change at any one of them lands directly on results.
Source: AppLovin Corporation Form 10-Q for the quarterly period ended June 30, 2026, filed August 5, 2026 (SEC accession no. 0001751008-26-000059), and the Form 8-K earnings release dated August 5, 2026, Exhibit 99.1 (accession no. 0001751008-26-000057). All figures are as reported; prior-year periods reflect the Apps business as a discontinued operation.