AppLovin Q3'25 vs Q4'25 — Delta analysis

Purpose: identify what weakened, went silent, or decelerated between Q3 2025 (released 2025.11.05) and Q4 2025 (released 2026.02.11). AppLovin's Q4 looks extraordinary on every axis — Rule of 40 = 150 is a statistical rarity. This note separates operational acceleration from seasonal timing and tests for messaging omissions.

1. Headline comparison

Q4 seasonality context: AppLovin is a mobile advertising platform. Q4 is structurally their strongest quarter (holiday shopping + annual advertising budget sweeps), so QoQ acceleration is partly seasonal. The YoY comparison removes seasonality. YoY +66% revenue, +82% adjusted EBITDA are the real acceleration numbers.

Q4 came in above the midpoint of management's own guidance ($1,570-1,600M → actual $1,658M; Adj EBITDA $1,290-1,320M → actual $1,399M). This is a non-trivial beat even by AppLovin's recent track record of beating by single-digit percentages.

2. Apps business divestiture — the structural change

The single biggest change between Q3 and Q4 context is that Q4 2025 is the first full quarter as a pure ad-tech company (Apps/mobile gaming divestiture completed June 30, 2025).

Transaction structure (AppLovin press, 2025.07.01):
- Sold 10 game studios (Lion Studios, Machine Zone, PeopleFun, Zeroo Gravity, Belka Games, Athena, Clipwire, Leyi, Magic Tavern, Zenlife) to Tripledot Studios.
- Consideration: $400M cash + $400M equity (~20% of Tripledot fully-diluted).
- Closed 2025.06.30.
- Goodwill impairment $189M taken on Apps business during 2025 (Q3 YTD cash flow statement).
- $99M loss from discontinued operations through Q3 YTD.

Q4 2025 implication: 100% of reported Q4 revenue is advertising-platform revenue. No gaming/Apps revenue anywhere. The 84% adjusted EBITDA margin — up +700bp YoY — partly reflects mix shift from a lower-margin Apps segment to a higher-margin pure software platform.

3. Tax rate — a subtle QoQ shift worth flagging

Q3 2025 effective tax rate: 18.1% ($185M provision / $1,021M pre-tax). This matched or exceeded Q4 2024 guidance assumptions.

Full year 2025 ETR: approximately 13.1% ($520M / $3,953M pre-tax). Implied Q4 2025 ETR: approximately 9-10%.

This is a material QoQ shift. AppLovin didn't specifically break out the reason in the press release (the release is skeletal), but possible explanations:
- Full-year true-up of stock-based compensation tax deduction (common Q4 event)
- Discrete tax benefits related to Apps divestiture (favorable basis differences)
- R&D credit true-up

Without the shareholder letter or transcript detail specifying the source, this is a watch-item. At a normalized 18% ETR, Q4 net income would have been approximately $945M vs reported $1,102M — roughly $0.46 of the $3.24 EPS came from tax rate favorability (about 14% of EPS).

4. E-commerce self-service — what's known, what's opaque

The biggest forward story out of Q4 2025 is AppLovin's e-commerce platform transition. Management disclosed on the call (Motley Fool transcript, 2026.02.11):

• Self-service platform launched in Q4 on referral-only basis
• General Availability (GA) targeted for H1 2026
• 57% of qualified leads currently go live (i.e., 43% breakage) in referral pilot
• One example: an Israeli cookware company scaled from $4M to ~$80M annualized revenue through the platform
• Generative AI tools for ad creative are in development to reduce breakage before GA

FY 2026 revenue guidance: $8.0B-$10.3B — a $2.3B range, roughly 26% of the midpoint. This is extraordinarily wide guidance for a SaaS-comparable company, and the breadth directly reflects e-commerce rollout uncertainty. At the low end ($8.0B), revenue grows +46% YoY — still very strong. At the high end ($10.3B), +88% — requires e-commerce to compound dramatically.

Q1 2026 guide (more precise, shorter horizon): $1,745-1,775M = 5-7% sequential from Q4's $1,658M. This is narrower sequential growth than Q3→Q4 (+18%) even accounting for Q4→Q1 seasonality (Q1 is typically slightly weaker post-holiday). If e-commerce rollout was accelerating, we'd expect Q1 guidance to at least match or exceed seasonal baseline.

5. What's silent vs Q3

① Short-seller allegations not addressed: In February-March 2025, short-sellers Culper Research and Fuzzy Panda published reports alleging ad fraud / click manipulation / attribution issues at AppLovin. Stock dropped 15-20% at the time before recovering. Q4 2025 earnings call: no management commentary on the allegations a full year later. This is management's consistent posture (they declined to address on Q1/Q2/Q3 calls as well), but it's a notable silence in what is otherwise detailed business disclosure.

② Limited segment breakdown for e-commerce: AppLovin does not break out e-commerce vs gaming-ad revenue separately. The $2.4B single-quarter e-commerce number that VRT disclosed has no AppLovin equivalent. We know e-commerce is growing rapidly but the absolute contribution remains a management-framed narrative, not a disclosed line.

③ Apps 20% equity in Tripledot: AppLovin holds ~20% of Tripledot's fully-diluted equity as a result of the sale. Q4 release doesn't update the carrying value or fair value of this investment. Tripledot is private, so marking is subjective, but shareholders effectively hold a $400M+ stake in a gaming holdco that received AppLovin's best studios. This is not in accounts-receivable or prominent balance sheet disclosure.

6. Buyback pace — aggressive, but at elevated prices

• 2025 full year: repurchased + withheld 6.4M shares for $2.58B
• Average buyback price: ~$403/share (with withholding)
• Year-end shares outstanding: 338M (vs 340M at start of 2025)
• Remaining buyback authorization at end of 2025: ~$0.8B (started year at ~$3.3B after October $3.2B increase)

At $2.58B in 2025 buybacks against $3.95B FCF, AppLovin deployed ~65% of FCF to buybacks. This is aggressive capital return for a growth company. At Q4 2025 year-end, AppLovin stock was in the $400-450 range (post Apps divestiture + Q3 earnings strength); Q1 2026 opened in the $450+ range. A buyback at $403 avg price that now shows $450+ would have been value-accretive by ~$0.3B mark-to-market.

7. Rule of 40 = 150 is real

The "Rule of 40" in SaaS investor parlance = revenue growth % + EBITDA (or FCF) margin %. 40+ is good. 80+ is excellent. 150 is essentially unprecedented in public markets. Calculated as:

• Revenue growth (YoY Q4): 66%
• Adj EBITDA margin: 84%
• Rule of 40 score: 150

CFO Matt Stumpf called this "almost unheard of" on the call. A check across SaaS comparables in Q4 2025:
- Snowflake: ~90 (30% growth + 60% gross margin)
- Datadog: ~95
- CrowdStrike: ~100-110
- Palantir: ~95-100
- ServiceNow: ~75-80

AppLovin's 150 is in a different tier. The challenge this creates is that it is mathematically unsustainable — a mature company cannot simultaneously grow revenue 66% YoY and hold 84% EBITDA margins indefinitely. Either growth decelerates toward SaaS norms (15-30%) or margins compress as AppLovin invests in new verticals (e-commerce sales/marketing spend, international expansion). Both scenarios bring the Rule of 40 score back toward 80-100 over 2-3 years.

8. The two findings that most matter

1. Q1 2026 sequential guidance 5-7% vs Q3→Q4 sequential +18% — a meaningful deceleration in the sequential trajectory even accounting for Q1 seasonality. This should be seen in light of the wide $8-10.3B FY guidance range. AppLovin is signaling that Q1-H1 will be a pre-GA period for e-commerce where the self-service platform is still in referral-only mode.

2. Full year ETR 13.1% vs Q3 YTD 18.1% implies Q4 ETR ~9-10%, flattering Q4 net income by ~$150M (~$0.45/share). Adjusted EBITDA is unaffected by this, but GAAP EPS of $3.24 includes this tax benefit. At a normalized rate Q4 EPS would be ~$2.80.

9. What's genuinely strong (counter-critique)

• Rule of 40 = 150 is statistically extraordinary and operationally real (not an accounting artifact).
• 84% Adj EBITDA margin on $1.66B revenue = almost no other public company delivers this combination.
• $3.95B annual FCF on $5.48B revenue = 72% FCF margin. Very rare.
• Apps divestiture executed cleanly (on schedule, for stated price, with equity retention).
• $2.58B buyback at $403 avg while generating $3.95B FCF = capital return discipline.
• AXON AI improvements continue to ship and management describes "more improvements will come" as research advances.
• E-commerce platform launched (referral-only) with at least one documented case study of >20x scale (Israeli cookware $4M→$80M).
• Balance sheet cleanup: discontinued operations line fully zero by Q3-end 2025; Apps noise gone from GAAP starting Q4.