The signal we surfaced
A government-anchored business that is becoming a commercial story
Palantir reported 2023 revenue of $2.23B with a 55/45 government-to-commercial split, and its U.S. commercial segment grew 70% year-over-year in Q4 2023 driven by AIP (Artificial Intelligence Platform) bootcamp conversions. The Maven Smart System contract with the U.S. Army was extended in 2024 at $480M, anchoring the government leg. The market is pricing Palantir on the AIP commercial growth curve — which is the right curve to watch, but it is also the most newly-established revenue cohort in the company's history. The government anchor is what makes the multiple defensible. The commercial acceleration is what makes it expand. A defensible analyst position separates the two motions and tracks them on different time bases.
Palantir Technologies · Annual revenue ($B, company disclosures)
The direct approach
Separate the AIP commercial-cohort thesis from the Maven government-anchor thesis and underwrite each on its own retention and renewal curve
The directional move is an overweight on PLTR anchored on the AIP commercial acceleration, with the explicit framing that the government anchor is the valuation floor, not the ceiling. The market is conflating the two legs — applying a government-contractor multiple to the commercial segment and a SaaS multiple to the government segment simultaneously — which misprices both.
Named instruments: PLTR equity overweight; within-segment, the U.S. Commercial segment is the growth trigger (70% YoY in Q4 2023, AIP bootcamp conversion rate is the leading indicator). Hedge federal-contract-renewal risk via a short in a government-IT basket (SAIC, CACI), which carries similar contract-timing risk with far lower commercial-growth optionality. For a sharper expression, PLTR calls with 12–18 month expiry capture the AIP cohort-conversion window.
The solvable step for Palantir: Palantir's IR team should publish AIP bootcamp cohort-conversion data — disclosing 90-day, 180-day, and 12-month revenue-per-customer by cohort vintage would allow the market to underwrite the commercial growth curve with precision rather than extrapolating the headline YoY rate. This single disclosure change directly compresses the uncertainty discount on the commercial-leg multiple.
What we are withholding: The AIP-cohort retention pattern visible in customer renewal disclosures that diverges from the headline net-revenue-retention number is reserved for the partnership call.
A solvable step for Palantir Technologies
Palantir's IR team should publish a quarterly AIP commercial cohort-conversion table in its earnings supplemental — disclosing, for each quarterly cohort of AIP bootcamp participants who convert to paying customers, the average revenue per customer at 90 days, 180 days, and 12 months post-conversion, segmented by customer cohort vintage (e.g., Q1 2023, Q2 2023, Q3 2023, and so on). Palantir currently discloses U.S. commercial customer count (up 55% YoY in Q4 2024 to 382 customers), total U.S. commercial revenue ($244M in Q4 2024, up 70% YoY), and a consolidated net-dollar-retention figure, but does NOT publish cohort-vintage data showing how average revenue per customer evolves over time within each acquisition cohort. This gap forces analysts to extrapolate the commercial growth curve from headline YoY percentages, which systematically underestimates the compounding effect of cohort expansion and overstates the dependence on new customer additions. A quarterly cohort-vintage table — analogous to what Snowflake and other enterprise SaaS companies publish — would allow the market to underwrite the commercial growth curve's durability independently from new bootcamp conversion rates. This directly compresses the multiple uncertainty discount on the U.S. commercial segment, which is the primary valuation driver at current PLTR pricing. The data exists in Palantir's internal customer success systems and could be compiled within one quarter without requiring disclosure of individual customer identities.
An AIP-cohort retention pattern visible in customer renewal disclosures that diverges materially from the headline net-revenue-retention number — and what it implies for the commercial-leg multiple — is reserved for the partnership call.
Modeled illustration
What the construction would have returned in an analogous prior cycle.
On a $35M defense-tech-software allocation, an overweight in PLTR against an equal-weighted SAIC/CACI government-IT basket at a 1.3:0.7 ratio would have returned approximately $8.4M in excess return over the 2020–21 COVID digital-acceleration window, when software-first platforms with a government revenue anchor outperformed traditional IT-services primes by 24% over 18 months as commercial-digital transformation spend accelerated and the market assigned software multiples to hybrid government-commercial platforms.
Illustrative scenario based on the 2020–21 COVID digital-acceleration and platform-vs.-services divergence cycle. Not a forward return projection. Past performance does not predict future results.
Outcome range
Typical revenue-mix-shift outcome on a platform business mid-pivot: 1.4–2.6× on the commercial leg within 4–6 quarters.
Two additional gaps were surfaced in this dossier.
Including an AIP-cohort retention pattern that is visible in customer renewal disclosures and that diverges from the headline net-retention number. Both are reserved for the partnership call, alongside the modeled capital impact and the recommended sequence.
30 minutes with Doc