Date: 2026-02-23 Quarter: Q4 FY25 (Dec 2025) Market cap: ~$82B | EV/TTM Rev: ~15x (post-41% YTD selloff) | Revenue growth: +66% YoY
AppLovin delivered the strongest quarter in its history — $1.66B revenue, $1.40B EBITDA, $1.31B FCF, Rule of 40 score of 150 — while the stock sits 41% below its year-start price on narrative fears about AI competition and a live SEC investigation. The operational reality and the stock price are the most disconnected I've seen for a company of this quality. The thesis is intact and strengthening on every financial metric. The non-financial risks (SEC probe, short-seller allegations attacking AXON's data collection mechanism) are real and unresolved — they prevent a Conviction 5. I hold Conviction 4/5.
Conviction: 4/5
| Criterion | Threshold | Actual | Pass? |
|---|---|---|---|
| Revenue YoY growth | >30% | +66% | ✅ Pass |
| Gross margin | >60% | 88.9% GAAP | ✅ Pass |
| Revenue per quarter | >$50M | $1,658M | ✅ Pass |
| Data availability | 4+ quarters | 12 quarters | ✅ Pass |
| Share dilution | <10% annual | ~-1.8% (net buyback) | ✅ Pass |
| GAAP profitability trajectory | Improving | Net margin 67% and rising | ✅ Pass |
All gates pass. Net share count is declining — buybacks exceeded all equity dilution in FY25.
| Factor | Rating | Detail |
|---|---|---|
| Growth | Strong | +66% YoY Q4 FY25; FY25 full-year +70% YoY; Q1 FY26 guide implies ~52% YoY — still far above threshold |
| Trajectory | Mixed | YoY decelerating from peak 77% (Q2 FY25) → 68% → 66% → ~52% guided. BUT QoQ re-accelerated Q4 (+18.0% vs +11.6% prior Q). E-commerce must compensate to inflect YoY. |
| Margins | High | Gross 88.9%, EBITDA 84.4%, FCF 79.0% — all-time highs, all expanding. 95% incremental EBITDA flow-through in Q4. Structurally fortress-like. |
| Dominance | Dominant | 85% iOS ad SDK reach (Pixalate Q3 2025). 80%+ mobile mediation share via MAX. AXON 2.0 moat from 10+ years of ML training data. Unity in retreat (rev -10% 2024). AXON e-commerce competitive from day one ("at least 5 of 10 advertisers say performance is really good"). |
| Valuation | Cheap | EV/TTM Rev ~15x at 66% growth. EV/Forward Rev ~12x at 52% growth. PEG well below 0.3x. Down from ~25x at peak. Cheaper growth-adjusted than Meta (~12x at 15-20% growth). |
| Special | Present | 41% stock dislocation from all-time-high fundamentals. E-commerce self-service GA (H1 2026) = meaningful TAM unlocking catalyst. SEC resolution = binary signal. OpenAI ChatGPT ad partnership (unconfirmed Feb 21 report) = potential incremental TAM. |
Data comparability note: Q1–Q3 FY24 and Q1 FY25 include the Apps gaming studio segment (divested June 30, 2025). Q4 FY24 onward is pure advertising platform. YoY comparisons across the divestiture boundary are mixed-composition. FY25 full-year advertising-only revenue = $5,481M per press release (using $3,224M FY24 continuing ops as comparator = +70% YoY). Q4_FY23 gross margin/profit anomalous due to impairment charges — exclude from trend analysis.
| | Q1_FY23 | Q2_FY23 | Q3_FY23 | Q4_FY23 | Q1_FY24 | Q2_FY24 | Q3_FY24 | Q4_FY24 | Q1_FY25 | Q2_FY25 | Q3_FY25 | Q4_FY25 | | | Mar-23 | Jun-23 | Sep-23 | Dec-23 | Mar-24 | Jun-24 | Sep-24 | Dec-24 | Mar-25 | Jun-25 | Sep-25 | Dec-25 | |---|---|---|---|---|---|---|---|---|---|---|---|---| | Revenue ($M) | 715 | 750 | 864 | 577¹ | 1,058² | 711² | 835² | 1,000 | 1,484³ | 1,259 | 1,405 | 1,658 | | YoY % | 14% | -3% | 21% | —¹ | —² | —² | —² | 73% | —³ | 77% | 68% | 66% | | QoQ % | 0% | 5% | 15% | —¹ | — | -33%² | 17% | 20% | 77%³ | -15%³ | 12% | **18%** | | Gross Margin [GAAP] | 63% | 66% | 69% | —¹ | 72% | 83% | 86% | 85% | 82% | 88% | 88% | **89%** | | Op Margin [GAAP] | 9% | 18% | 22% | —¹ | 32% | 54% | 64% | 63% | 45% | 76% | 77% | **77%** | | EBITDA Margin | — | — | 49% | 71% | — | — | — | 77% | 68% | 81% | 82% | **84%** | | Net Margin [GAAP] | -1% | 11% | 13% | 33% | 22% | 44% | 52% | 60% | 39% | 65% | 60% | **67%** | | FCF Margin | 40% | — | — | 59% | 37% | 63% | 65% | 70% | 56% | 61% | 75% | **79%** | | EPS GAAP (dil.) | -0.01 | 0.22 | 0.30 | 0.47 | 0.67 | 0.89 | 1.25 | 1.72 | 1.67 | 2.39 | 2.45 | **3.24** | | FCF ($M) | 289 | — | — | 340 | 388 | 446 | 545 | 695 | 826 | 768 | 1,049 | 1,309 | | EBITDA ($M) | — | — | 419 | 411 | — | — | — | 770 | 1,005 | 1,018 | 1,158 | 1,399 |
¹ Q4_FY23 distorted by impairment charges. ² Mixed-composition (includes Apps). ³ Q1_FY25 includes Apps through June 2025.
Advertising-only trend (clean comparison):
| Q1_FY24 | Q2_FY24 | Q3_FY24 | Q4_FY24 | Q1_FY25 | Q2_FY25 | Q3_FY25 | Q4_FY25 | |
|---|---|---|---|---|---|---|---|---|
| Ad Rev ($M) | 678 | 711 | 835 | 1,000 | 1,159 | 1,259 | 1,405 | 1,658 |
| YoY % | 91% | 75% | 66% | 73% | 71% | 77% | 68% | 66% |
| EBITDA Margin | 71% | 72% | 78% | 77% | 81% | 81% | 82% | 84% |
Source: Company notes (ad-only) and Q4_FY25 press release.
Prior beliefs (entering Q4 FY25 earnings):
| Metric | Expected | Basis |
|---|---|---|
| Revenue | 1, 648–1,680M | Guide midpoint $1,585M + 4–6% systematic beat |
| EBITDA margin | ~82–83% | In-line with Q3 FY25 guidance range |
| FCF margin | ~74–76% | Extension of improving trend |
| Management tone | Defensive on competition | Stock -41% YTD, SEC active, short-sellers unresolved |
| E-commerce update | Self-service progress + client count | GA timeline reconfirm expected |
| Beat characterization | Modest to in-line | Beat streak narrowing toward guidance range |
Updated beliefs (actuals vs. expected):
| Metric | Expected | Actual | Verdict |
|---|---|---|---|
| Revenue | ~$1,660M | $1,657.9M | ✅ In line (4.6% beat on guide) |
| EBITDA margin | 82–83% | 84.4% | ✅ Beat — +1.4pp above top of guidance range |
| FCF margin | 74–76% | 79.0% | ✅ Positive surprise |
| Incremental EBITDA flow | ~90% | ~95% | ✅ Exceptional execution |
| QoQ sequential growth | 10–13% (seasonal decel) | 18.0% | ✅ Large positive surprise — re-acceleration |
| Rule of 40 | ~145–148 | 150 | ✅ At theoretical ceiling |
| Management tone | Defensive | Confrontational/proactive — opened prepared remarks on competition | ✅ Matched expectation; scale exceeded |
| E-commerce breakout | No breakout | Not provided — "unified platform" rationale | ✅ Expected; maintained |
| SBC as % revenue | 2.5–3% (continuing Q3 trend) | 4.9% (Q4 annual grant cycle) | ⚠️ Spike, but annual grant timing not deterioration — FY25 SBC down 44% YoY |
Delta assessment — what surprised me:
Sequential re-acceleration (11.6% → 18.0% QoQ): Not expected given Q4 seasonal strength already baked in. The "material model unlock" Foroughi described — rolled out weeks before the earnings call — is embedded in both Q4 actuals and Q1 FY26 guidance. This is the most bullish signal in the report.
EBITDA margin exceeded guide by 140bps despite marketing spend initiation. If performance marketing scales (they're testing now with LTV:CAC breakeven at day 30), margins could compress modestly — but that would indicate explosive e-commerce growth, which is unambiguously positive. The margin "floor" appears to be ~82–83%.
CTV commitment silently dropped: Q4 FY24 call included an explicit CTV performance advertising initiative for 2025. Zero mention in Q3 or Q4 FY25. Management quietly deprioritized this. Not a thesis threat given gaming + e-commerce dominance, but worth noting that commitments can disappear without acknowledgment.
57% qualified lead conversion (43% breakage): More friction than expected. The binding constraint is that e-commerce advertisers don't have video ads in the format AXON needs. Gen AI video model ("shortly") is the direct fix. GA timing (H1 2026) appropriately held until conversion improves.
APP is a performance advertising platform — no traditional SaaS leading indicators (RPO, billings, ARR). Relevant forward signals:
| Indicator | Current | Trend | Signal |
|---|---|---|---|
| E-commerce pixel count | 5,306 clients | Rapid growth from zero (Oct 2024) | Bullish — expanding supply-side data; each new pixel improves model |
| Qualified lead → live conversion | 57% (43% breakage) | Unknown prior trend; gap vs 100% target | Neutral — friction is real but addressable via gen AI video tools |
| Existing e-commerce cohort spend | "Material YoY increase" + "big uplift weeks ago" | Accelerating | Bullish — retention + expansion from maturing cohort |
| Revenue QoQ growth | 18.0% Q4 vs 11.6% Q3 | Re-accelerating | Bullish — largest sequential dollar add in company history (+$253M) |
| Revenue beat consistency | 4Q avg +5.75% above midpoint | Stable | Bullish — systematic under-guidance implies embedded conservatism in Q1 FY26 guide |
| Gen AI video creative pilot | 100+ customers on interactive page; video "shortly" | Imminent | Potentially bullish — directly addresses 43% breakage rate |
| MAX marketplace growth | "Really quickly"; multiple competitors growing simultaneously | Positive ecosystem | Bullish — rising tide in MAX expands APP's take rate income |
| Model unlock frequency | ~quarterly material improvements | Ongoing | Bullish — model continues improving independent of advertiser count |
| Conversion rate potential | ~1% current vs 5% achievable | Long runway | Bullish — 5x embedded monetization upside without new advertisers |
Divergence check: Revenue (66% YoY) is growing broadly in line with what pixel count and cohort spending trends would predict. No bearish divergence visible. The sequential re-acceleration argues for leading indicators (model improvement, existing cohort spend) moving ahead of reported revenue — a mild bullish divergence.
Sources: stages/scuttlebutt/APP/2026-02-23.md
Customer sentiment — gaming (core):
Customer sentiment — e-commerce (nascent, mixed):
Employee sentiment:
Competitive landscape:
SEC investigation:
Management:
Note on stock price: APP ended 2025 at ~400/share.YTD2026declineof41236. After OpenAI report rally (~+14% on Feb 21) = ~$269 at analysis date. All figures approximate — use directionally.
| Metric | Current (post-selloff) | At Peak (~end 2025) | Peer: Meta | Peer: Trade Desk | Assessment |
|---|---|---|---|---|---|
| EV/TTM Revenue | ~15x | ~25x | ~12x | ~15–20x | Cheap vs growth |
| EV/TTM EBITDA | ~18x | ~30x | ~15x | ~35x | Fair to cheap |
| EV/TTM FCF | ~21x | ~35x | ~20x | ~40x | Fair |
| P/E GAAP (TTM) | ~24x | ~40x | ~26x | N/M | Fair |
| Revenue growth | 66% YoY | 66% YoY | ~15% | ~25% | APP's growth far higher |
| EBITDA margin | 84% | 84% | ~40% | ~18% | APP's margin best-in-class |
FY25 full year: Revenue $5,481M, EBITDA $4,512M (82% margin), FCF $3,952M (72% margin), Net income $3,334M.
Q1 FY26 annualized run rate: $1,760M × 4 = $7,040M. EV/Forward Rev at $82B EV ≈ 11.6x. At 52% growth, this is extraordinarily cheap.
Key valuation insight: The market is pricing the narrative (AI disruption of AXON, SEC risk, competition) not the numbers. Every operational metric hit all-time highs in Q4 FY25. At 15x EV/TTM Revenue with 66% growth, 84% EBITDA margin, and 79% FCF margin, APP is cheaper on growth-adjusted basis than Meta. The PEG ratio (EV/Rev ÷ growth %) is ~0.23x — one of the lowest I've seen for a hypergrowth software platform.
The bear case on valuation: If growth decelerates to 30–35% by FY26 without e-commerce compensating, and if the SEC probe impairs AXON's data collection, the multiple could compress to 8–10x forward revenue. At $7B forward revenue and 10x, market cap = $70B — roughly flat from here. Not catastrophic, but not asymmetric.
Secular trend: Performance advertising in a mobile-first, attention-scarce world. The shift from managed service (human optimization) to AI-native (model optimization) is structural and irreversible. AXON 2.0 is the purest AI-native performance advertising model in mobile today.
Platform or point solution? Platform. MAX mediation is the exchange layer; AXON is the demand-side intelligence; the creative tools (gen AI video, interactive pages) are becoming the supply-side layer. The three elements compound: more publisher inventory → better model training → better advertiser returns → more advertiser spend → more publisher revenue → more publishers. Classic two-sided marketplace flywheel.
TAM penetration:
AI-native moat: AXON's competitive advantage is not just the ML architecture — it's 10+ years of training data from the world's largest mobile gaming mediation network. A new entrant with equivalent architecture but no training data would take years to match AXON's predictive accuracy. CloudX has OpenAI's models but zero mobile gaming training data.
SEC investigation (binary): If the probe finds AXON's training data was collected via fingerprinting in violation of Apple/Google ToS, APP could be forced to retrain models from scratch with compliant data — degrading AXON's performance significantly. No timeline for resolution. This is the existential tail risk.
YoY growth deceleration below 40%: Q4 FY25 at 66% → Q1 FY26 guide ~52%. If e-commerce doesn't scale fast enough (GA H1 2026, conversion rate from 57% to >90%), YoY comps could compress to 30–35% range by H2 FY26. That would break the growth qualification gate.
E-commerce ROI skepticism from advertisers: Ridge and Jones Road pullbacks demonstrate that the "arbitrage window" closes as the model becomes known. AXON must prove durable incrementality, not just early-mover ROAS. Measurement transparency gap vs Meta is real and could slow enterprise adoption.
Platform risk (Apple/Google): APP's business depends on iOS/Android ad ecosystems. IDFA deprecation was absorbed; a more aggressive privacy policy change (e.g., mandated ATT opt-in for all ad delivery) could reduce AXON's signal. Low probability but high impact.
CTV promise silently dropped: Q4 FY24 included an explicit CTV initiative for 2025. Zero mention in any subsequent quarter. Shows management will quietly deprioritize commitments without acknowledgment. Worth monitoring whether e-commerce GA timeline receives the same treatment.
E-commerce self-service GA (H1 2026): Opening from referral-only to GA removes the 43% breakage friction, enables SMB at scale, and could compress the pixel count gap with Meta significantly. First indicator of true e-commerce scale.
Gen AI video creative model launch ("shortly"): The binding constraint for the 43% lead breakage is e-commerce advertisers lacking video in AXON's format. When the gen AI video model goes live, this constraint dissolves and conversion rates should inflect upward.
**Q1 FY26 beat on 1.76Bguide : * * 4 − quarterbeataverageof5.751.86B. If the model unlock from Q4 FY25 is sustained, the beat could exceed the historical average.
SEC investigation resolution: A clean bill of health from the SEC — or even a consent decree that doesn't impair AXON's data practices — would remove the single largest overhang on the stock. Binary upside catalyst.
First e-commerce revenue quantification: Management has refused to break out e-commerce revenue ("unified platform"). As the business matures, investor pressure to quantify will grow. First disclosure — even an approximate range — would reset the e-commerce valuation narrative and could be a significant re-rating catalyst.
Long APP: 6.8% of portfolio (common shares) + 6.4% LEAPS. Total exposure ~13.2%. Established prior to this analysis.