PLTR vs APP: Why 6x SBC Spread Reveals AI Scaling Economics
Palantir (14.6% SBC) and AppLovin (2.4% SBC) both grew revenue ~65% in Q3 2025. So why does one require 6x more equity compensation? Our SEC filing analysis reveals diverging trajectories that matter more than static spreads: APP's SBC is declining 38% YoY while PLTR's is accelerating 42% YoY. The difference: human-intensive vs AI-engine scaling economics.
Palantir vs AppLovin: Why 6x SBC Spread Reveals AI Scaling Economics
Last Updated: January 10, 2026 Data Currency: Q3 2025 10-Q filings. PLTR | APP
TL;DR: Two AI stocks with ~65% revenue growth have 6x different equity compensation profiles. The static spread (PLTR 14.6% vs APP 2.4%) tells part of the story. The diverging trajectories tell the rest: AppLovin's SBC is declining 38% YoY while Palantir's is accelerating 42% YoY. This isn't random—it reflects fundamentally different scaling economics. Human-intensive (PLTR) vs software-engine (APP) business models create different paths for shareholder value creation.
SBC and Capital Efficiency Comparison (Q3 2025)
| Metric | PLTR | APP | Spread |
|---|---|---|---|
| SBC/Revenue | 14.6% | 2.4% | 6.1x |
| SBC YoY Trend | +42% | -38% | Diverging |
| ROIC (Q) | 24.5% | 82.9% | 3.4x |
| Revenue Growth YoY | 63% | 68% | Similar |
| Gross Margin | 82% | 80% | Similar |
| Operating Margin | 33% | 50% | 1.5x |
| Unvested SBC Liability | $947M | Minimal | — |
| Earnings Quality Score | 6/10 | 7/10 | APP |
Source: MetricDuck 5-pass filing intelligence extraction
Analyze these companies: MetricDuck extracts SBC analysis, hidden liabilities, and risk factors from SEC filings. View PLTR Filing Intelligence | View APP Filing Intelligence
The Question That Matters
Palantir and AppLovin both grew revenue approximately 65% year-over-year in Q3 2025. Both trade at premium valuations. Both are AI-narrative stocks.
So why does one company require 6x more equity compensation to achieve similar growth?
This isn't a rhetorical question. The answer reveals which shareholders will capture more of future growth—and which will see it diluted away.
The Answer: Two Different Scaling Models
The 6x SBC spread isn't random variation. It reflects fundamentally different business architectures:
Business Model Comparison
| Factor | Palantir | AppLovin |
|---|---|---|
| Growth Driver | Headcount + Custom Deployments | AXON AI Engine |
| Customer Model | 911 customers, top 20 avg $83M | Pure advertising platform |
| Scaling Mechanism | Add engineers per customer | AI improves without headcount |
| SBC Trajectory | +42% YoY (accelerating) | -38% YoY (declining) |
| Revenue/Employee | Lower (human-intensive) | Higher (software-intensive) |
Palantir: Human-Capital Intensive
Palantir's business requires deployment. Each customer needs engineers to implement, customize, and maintain their Foundry or AIP installation. The company has grown from 629 to 911 customers, with top 20 customers averaging $83 million in revenue each.
This creates a direct relationship between customer growth and headcount growth. More customers = more engineers = more stock compensation.
SEC Filing Evidence: "Our average revenue for the top twenty customers during the trailing twelve months ended September 30, 2025 was $83.0 million, which grew 38% from an average of $60.1 million in revenue from the top twenty customers during the trailing twelve months ended September 30, 2024."
AppLovin: Software-Engine Intensive
AppLovin's AXON AI recommendation engine scales differently. The software improves itself—net revenue per installation increased 75% while installation volume stayed essentially flat.
This creates operating leverage. Revenue grows faster than costs because the AI engine handles the work humans would otherwise do.
SEC Filing Evidence: "For the three months ended September 30, 2025, our revenue increased by $569.9 million, or 68%, compared to the same period in the prior year due primarily to improved Axon Advertising performance, where net revenue per installation increased 75%, partially offset by a decrease in the volume of installations of 1%."
Why Trajectories Matter More Than Static Spreads
The 6x SBC spread is significant. But the diverging trajectories are more important for forward-looking analysis.
SBC Trajectory Analysis
| Company | SBC/Revenue | YoY Change | Interpretation |
|---|---|---|---|
| PLTR | 14.6% | +42% growth | Dilution accelerating |
| APP | 2.4% | -38% decline | Past inflection point |
AppLovin has entered a new phase: "There were no material equity award issuances during the nine months ended September 30, 2025."
Palantir is still investing: "$947.1 million... which the Company expects to recognize over a weighted-average service period of three years."
What This Means for Shareholders
AppLovin shareholders benefit from a business that no longer needs significant equity dilution to grow. The company is buying back shares, not issuing them. Future revenue growth flows more directly to existing shareholders.
Palantir shareholders face continued dilution. The $947M unvested RSU liability will become stock over the next three years. If SBC continues growing 42% YoY while revenue grows 63%, the dilution rate may stabilize—but it's not improving.
Deep Dive: Palantir (6/10 Earnings Quality)
The Bull Case
Palantir bulls make legitimate points:
- Sticky government contracts — Mission-critical systems have high switching costs
- Customer diversification — Top 3 customers = 17% of revenue (down from 18%)
- Geopolitical tailwinds — Defense spending likely remains elevated
- Commercial acceleration — 73% YoY growth in commercial segment
The Concerning Metrics
| Metric | Value | Assessment |
|---|---|---|
| SBC/Revenue | 14.6% | Elevated |
| SBC Growth YoY | +42% | Accelerating faster than efficiency gains |
| Unvested RSU Liability | $947M | 3-year dilution pipeline |
| SBC Sustainability | "Elevated" | Per SEC filing analysis |
| Government Concentration | 54% | Still majority government |
Hidden Liabilities
| Liability Type | Amount | Duration | Risk |
|---|---|---|---|
| Cloud Commitments | $1.95B | 10 years | Moderate |
| Class Action Litigation | Ongoing appeal | — | Low-Medium |
| Indemnification Obligations | Unquantified | Ongoing | Low |
SEC Filing Quote: "Significant cloud hosting commitment ($1.95 billion through 2033) represents a substantial future operational expense."
Key Insight
Palantir's SBC isn't a quality problem—it's a business model characteristic. Human-intensive enterprises require talent, and talent in AI/ML commands equity. The question is whether the platform becomes sticky enough to eventually reduce per-customer deployment costs.
Deep Dive: AppLovin (7/10 Earnings Quality)
The Bull Case
AppLovin bulls point to operational efficiency:
- Operating leverage — 50% operating margin vs PLTR's 33%
- SBC declining — 38% YoY reduction signals efficiency
- ROIC excellence — 82.9% return on invested capital
- Focus — Divested Apps business, now pure-play advertising
The Concerning Metrics
| Metric | Value | Assessment |
|---|---|---|
| Platform Risk | #2 Ranked Risk | Apple/Google dependency |
| Debt/Equity | 2.38x | Leveraged capital structure |
| Single Segment | 100% advertising | Concentration after divestiture |
| Cybersecurity Risk | #1 Ranked Risk | Ad ecosystem vulnerability |
Hidden Liabilities
| Liability Type | Amount | Duration | Risk |
|---|---|---|---|
| Cloud Commitments | $1.3B | 3 years | Moderate |
| Related Party Impairment | $50M (Humans, Inc.) | One-time | Low |
| Litigation | Ongoing | — | Medium |
SEC Filing Quote: "Changes to the policies or practices of third-party platforms, such as the Apple App Store and the Google Play Store, including with respect to Apple's Identifier for Advertisers ('IDFA')."
Key Insight
AppLovin's low SBC is sustainable because their AI engine scales without proportional human capital. But this efficiency comes with concentration risk—Apple or Google policy changes could materially impact the business overnight.
The Honest Framework: When Is High SBC Acceptable?
Not all high SBC is bad. Here's a framework for evaluation:
SBC Acceptability Framework
| Condition | Acceptable | Concerning |
|---|---|---|
| Company Stage | Early-stage, building platform | Mature, established product |
| SBC Trajectory | Declining as % of revenue | Growing faster than revenue |
| Growth vs. SBC | Revenue growth >> SBC growth | SBC growth ≈ revenue growth |
| Cash Generation | Strong FCF despite SBC | SBC masks cash burn |
| Competitive Moat | SBC builds defensible position | SBC just retains commodity talent |
Palantir Assessment: Mixed. Early-stage platform argument has merit, but SBC growing 42% vs 63% revenue growth isn't significantly declining. Needs to show inflection.
AppLovin Assessment: Past acceptable phase. SBC declining 38% with 68% revenue growth is the ideal scenario. Shareholders capture growth.
Risk-Adjusted Perspective
Neither company is "better" in absolute terms. They have different risk profiles:
Palantir Risks
- SBC acceleration — Dilution faster than efficiency gains
- Government budget uncertainty — 54% revenue exposure
- Human capital intensity — Operating leverage limited
AppLovin Risks
- Platform dependency — Apple/Google policy changes existential
- Advertising cyclicality — More volatile than government contracts
- Single segment — All eggs in advertising basket
The Tradeoff
Palantir offers: Stickier revenue, diversified customers, lower platform risk At the cost of: Higher dilution, lower ROIC, human capital constraints
AppLovin offers: Higher ROIC, declining dilution, operating leverage At the cost of: Platform dependency, advertising cyclicality, concentration
What the Data Actually Shows
Cutting through the narratives, here's what SEC filings reveal:
| Claim | Evidence |
|---|---|
| "Human-capital intensive" | PLTR: 911 customers, top 20 avg $83M each requiring deployment |
| "Software-engine scales" | APP: Revenue/installation +75%, volume flat |
| "SBC diverging" | PLTR: +42% YoY; APP: -38% YoY |
| "APP past inflection" | "No material equity award issuances" in 9 months |
| "PLTR still investing" | "$947.1M unvested... over three years" |
| "Platform risk real" | APP's #2 ranked risk factor |
| "Customer diversifying" | PLTR: Top 3 = 17% (down from 18%) |
Bottom Line
The 6x SBC spread between Palantir and AppLovin isn't a bug—it's a feature of fundamentally different business models.
Palantir is building a human-intensive enterprise platform. High SBC is the cost of acquiring and retaining the talent needed to deploy at each customer. The bull case depends on eventually achieving platform leverage that reduces per-customer costs.
AppLovin has already achieved software-engine leverage. The AI improves itself, revenue scales without proportional headcount, and SBC is declining. The bull case is simpler math: shareholders keep more of the growth.
The diverging trajectories tell the story better than static spreads. One company's dilution is accelerating. The other's is declining. Over five years, this compounds into meaningfully different shareholder outcomes.
Choose based on which risk profile matches your thesis—not based on which SBC number looks "better" in isolation.
Explore More
This analysis is part of our Earnings Quality Analysis framework.
Related Analyses:
- Stock-Based Compensation: How Tech Dilutes Shareholders — Comprehensive SBC methodology across sectors
- AppLovin's 75% ROIC: AI Adtech Analysis — Deeper dive into APP's capital efficiency
- Enterprise AI Earnings Quality: Palantir, Snowflake, Datadog — Broader AI software comparison
Methodology
This analysis uses MetricDuck's 5-pass SEC filing intelligence extraction:
- Pass 1: Narrative intelligence from MD&A
- Pass 2: Accounting quality and SBC analysis
- Pass 3: Hidden liabilities identification
- Pass 4: Risk landscape assessment
- Pass 5: Segment performance analysis
All quantitative metrics derived from XBRL-tagged financial data. Qualitative assessments grounded in specific SEC filing quotes.
Disclaimer: This analysis is for informational purposes only and does not constitute investment advice. Stock-based compensation analysis reflects point-in-time SEC filings and may change with future disclosures.
MetricDuck Research
SEC filing analysis and XBRL data extraction