Date: 2026-02-22 | Quarter: Q4 FY25 (ended Dec 31, 2025) | Reported: Feb 2, 2026
long PLTR 0% (not held — watching)
| | Q1 FY24 | Q2 FY24 | Q3 FY24 | Q4 FY24 | Q1 FY25 | Q2 FY25 | Q3 FY25 | Q4 FY25 | | | Mar-24 | Jun-24 | Sep-24 | Dec-24 | Mar-25 | Jun-25 | Sep-25 | Dec-25 | |---|---|---|---|---|---|---|---|---| | Revenue ($M) | 634 | 678 | 726 | 828 | 884 | 1,076 | 1,266 | 1,407 | | YoY % | +21% | +27% | +30% | +36% | +39% | +59% | +74% | +70% | | QoQ % | +2% | +7% | +7% | +14% | +7% | +22% | +18% | +11% |
YoY trend: +21% -> +27% -> +30% -> +36% -> +39% -> +59% -> +74% -> +70%
That's sustained acceleration from +21% to +70% over 8 quarters ^^. The slight pullback from +74% to +70% is noise — the underlying trend is intact and the comps get harder from here. More importantly: the COMPOSITION of this growth is what matters.
Here's the thing. PLTR isn't growing at +70% because of some contract win or one-time catalyst. It's because AIP (Palantir's AI Platform) created a fundamentally new GTM motion — the bootcamp — that is converting enterprise prospects to paying customers at a rate nobody predicted.
Segment breakdown Q4 FY25:
| Segment | Revenue ($M) | YoY % | QoQ % |
|---|---|---|---|
| U.S. Commercial | ~507M | +137% !! | ~+25% |
| U.S. Government | ~577M | +66% !! | ~+7% |
| International Commercial | ~208M | +13% | ~+4% |
| International Government | ~115M | +12% | ~+3% |
U.S. Commercial at +137% YoY is the headline. This is the AIP bootcamp thesis proving out in real time. Companies come in for a 5-day bootcamp, see AIP solve their actual operational problems, and convert to paying customers. The bootcamp model is essentially a frictionless self-serve GTM for enterprise — no 18-month SI engagement, no RFP theater. You show up, solve a real problem, sign a check.
NUANCE: International is the weak spot (+12-13% YoY). This is meaningful — Europe in particular is growing slowly, which limits TAM realization. But U.S. is compensating overwhelmingly right now.
Architecture: Cloud-native !!. PLTR's Ontology is the key architectural moat — a semantic data layer that creates a persistent, versioned model of a customer's operational reality. This is not a chatbot wrapper on existing data. This is infrastructure that becomes load-bearing over time. The more it knows about your operations, the more irreplaceable it becomes.
GTM: Self-serve enterprise !! via bootcamp model. This is the GTM shift that changed everything. Prior to AIP, PLTR was SI-dependent and relationship-driven — slow, lumpy, opaque. Now: bootcamp → pilot → expansion. Conversion rates are apparently very high and deal cycles are collapsing.
Building block vs process tool: Building block !!. AIP sits underneath customer workflows. Once you've mapped your operations through the Ontology, you can't rip it out without rebuilding your operational data model from scratch. This is exactly the kind of embedding I look for in platform companies.
Competitive position: Unique. Nobody else has the Ontology + multi-decade government classification history + enterprise AI deployment experience at scale. Snowflake is a data warehouse, not an AI OS. ServiceNow does workflow automation, not operational intelligence. Databricks does data engineering, not decision intelligence. PLTR is playing a different game — or trying to define a new category entirely.
Karp's "n of 1" comment is not hubris. It's an architectural claim. The Ontology as an abstraction layer for enterprise AI is genuinely differentiated.
| Metric | Q4 FY25 | Note |
|---|---|---|
| Revenue | $1,407M | +70% YoY ^^ |
| Adj Gross Margin | ~82% | [Non-GAAP] !! |
| Adj Op Margin | 57% | [Non-GAAP] !!! |
| GAAP Op Margin | ~16% | [GAAP] — SBC is real |
| FCF | $791M | 56% margin !! |
| Rule of 40 | 127% | !!! |
| SBC | ~$400M+ | Heavy — ~29% of revenue [GAAP] |
Rule of 40 at 127% is extraordinary. I've tracked a lot of software companies and this sits near the top of what I've ever seen sustained at scale. FCF conversion at 56% is real cash. These are not accounting artifacts.
SBC is the GAAP/Non-GAAP gap. PLTR's GAAP margins look very different from non-GAAP. SBC running at ~29% of revenue is heavy — this is dilution that matters to long-term shareholders. Management knows this and has been reducing it, but it's a legitimate concern. The GAAP P/E of ~200x reflects this.
| Indicator | Q4 FY25 | YoY % | Signal |
|---|---|---|---|
| TCV | $4.26B | +138% !! | Bullish divergence vs 70% revenue |
| U.S. Commercial Customer Count | High | +68% YoY | Healthy top-of-funnel |
| U.S. Bootcamp Deal Volume | High | Accelerating | GTM model working |
| Remaining Deal Value (RDV) | Growing | N/A | Future revenue lock-in |
TCV at +138% vs revenue at +70% = meaningful bullish divergence ^^. This is the most important number in the quarter for thesis purposes. It means the bookings pipeline is growing roughly 2x faster than revenue recognition. That gap has to close — either revenue accelerates to catch up, or TCV growth slows. The FY26 guide of +61% YoY suggests management believes revenue will accelerate to catch up.
| Metric | Guidance | YoY % | vs Consensus |
|---|---|---|---|
| Total Revenue | $7.18-7.20B | +61% | Beat ~$290M |
| U.S. Commercial | >$3.14B | +115% !! | Beat significantly |
| Adj Op Income | $4.13-4.14B | ~57% margin | — |
| Adj FCF | $3.93-4.13B | — | — |
Guide at +61% with U.S. Commercial at +115% is NOT a typical "sandbagged" guide. This is an aggressive number. Management is telling the market: the AIP flywheel has durability. I take this at face value given they've been beating and raising consistently, but I want to see Q1 FY26 actually deliver before adding conviction.
Prior (no prior muji coverage — first analysis): No prior thesis on file. PLTR has historically been a company I watched but didn't own — the government revenue dependency and murky commercialization made it hard to model, and Karp's compensation was egregious.
Updated: The AIP bootcamp model changed the thesis fundamentally. U.S. Commercial at +137% YoY is not luck — it's a GTM innovation that's working. The Ontology architecture creates genuine switching costs. The Rule of 40 at 127% is exceptional. The TCV/revenue divergence is a bullish leading indicator.
What changed my view: The bootcamp model is self-serve enterprise GTM done right. This is exactly the "convenience over price wins enterprise" dynamic I've written about. No SI dependency. Fast conversion. Expanding deal sizes. This is how platform companies should grow.
What still concerns me:
NUANCE 1: The valuation debate is legitimate but incomplete. Atlas correctly notes 70x TTM revenue. Here's what analysts miss: at 127% Rule of 40 with 57% FCF margins, PLTR is compounding FCF at a rate that makes the entry price more defensible at a 3-5 year horizon than it looks on a snapshot basis. If FY26 delivers $4B+ FCF and the business continues at this pace, the multiple compresses rapidly. The question is whether growth sustains.
NUANCE 2: The Ontology is underappreciated outside technical circles. Every analyst focuses on AIP as "AI platform" = chatbots. That's wrong. The Ontology is a persistent operational knowledge graph of a company's business. Once it's trained on your supply chain, logistics, maintenance schedules, and personnel data — it becomes the operational brain. This is not a tool. It's infrastructure. Analysts who compare PLTR to other AI companies are comparing apples to jet engines.
NUANCE 3: Government is a floor, not a ceiling. The U.S. Government segment at +66% YoY is actually impressive — especially given DOGE-related budget uncertainty. The concern about government dependency is valid structurally, but in practice, PLTR's government contracts (classified, operational) are likely more durable than discretionary IT spend. Government is a moat, not a millstone.
NUANCE 4: International is the sleeping bear. If Europe ever gets serious about AI sovereignty and industrial AI, PLTR is the most credible Western alternative. The current +12% is a policy-constrained market. This could inflect without warning. Not in the base case, but worth monitoring.
Tier 1 business. Revenue at +70% YoY ^^, Rule of 40 at 127% !!!, FCF at 56% margin !!. The AIP bootcamp GTM is genuinely innovative. The Ontology architecture creates compounding switching costs. TCV divergence signals continued acceleration. FY26 guide at +61% with U.S. Commercial >+115% is an aggressive, credible commitment.
But not at any price. At ~70x TTM revenue and ~141x EV/FCF, PLTR is the most expensive large-cap software stock on earth. The business quality justifies a premium. Whether it justifies THIS premium is the only question.
My stance: Not a current position. If I owned it, I'd hold — the platform thesis is real and accelerating. To initiate, I'd want either (a) a material pullback (30-40%) to a more defensible entry, or (b) Q1 FY26 to confirm the FY26 guide is tracking (U.S. Commercial >$780M would be the key data point). The business is extraordinary. The price requires flawless execution indefinitely. That's a risk profile I want to size carefully.
DING DING DING on the operating performance. NUANCE on the entry point.
-muji