Gotham, Foundry, Apollo and AIP — software for government and commercial decisioning. ~4,429 employees · commercial is re-rating the mix.
Commercial + AIP is re-rating the mix while government still funds half the base. Trailing P/E sits at the 4th percentile of Palantir's profitable history — earnings compounded under a −5.6% share price. Still: 63.9× sales is mid-pack in its own EV/S history and 3× the software peer median, so the multiple still needs the growth to print. Bullish on commercial crossing through; base below Street median because SBC (13.6% of sales) keeps the trust discount.
Palantir sells software platforms — Gotham for government, Foundry for commercial, Apollo for continuous deployment, and AIP as the LLM operations layer on top. Capex runs at ~0.7% of revenue: this is software, not fabs. On $6.16B of TTM sales the company converts $3.36B of free cash.
The growth story is commercial + AIP catching the government base. In FY25, Government was $2.40B and Commercial $2.07B — almost even. The Q2 FY26 run-rate implies commercial may lead soon. Everything on the next six pages is a judgment about how long that crossing compounds before the multiple has to compress.
Government still funds half the dollars, but Commercial is the re-rating engine. FY25 put Gov $2.40B against Comm $2.07B; AIP bootcamps and US commercial growth are what pull the mix through 50%. The multiple requires that crossing to keep printing.
United States grew at a ~42% three-year CAGR from FY22 to FY25; UK ~25%, Rest of World ~12%. The commercial acceleration is a US story first.
Four-year revenue CAGR is 30.5%, with FY25 alone at +56.2%. Operating margin swung from −26.7% to 31.6% in four years; TTM is already 42.8%. Free cash now exceeds net income.
| FY21 | FY22 | FY23 | FY24 | FY25 | What it says | |
|---|---|---|---|---|---|---|
| Revenue | 1.54 | 1.91 | 2.23 | 2.87 | 4.48 | Acceleration into FY25, not a one-quarter spike |
| Revenue growth | +41.1% | +23.6% | +16.7% | +28.8% | +56.2% | Commercial + AIP bent the curve |
| Gross margin | 78.0% | 78.6% | 80.6% | 80.2% | 82.4% | Software GM; TTM already 84.8% |
| Operating margin | −26.7% | −8.5% | 5.4% | 10.8% | 31.6% | Leverage arrived in FY23 and kept compounding |
| Diluted EPS | −$0.27 | −$0.18 | $0.09 | $0.19 | $0.63 | TTM EPS already $1.18 on the Q2 print |
| Free cash flow | 0.32 | 0.18 | 0.70 | 1.14 | 2.10 | 47% FCF margin in FY25 — cash follows earnings |
Eight beats from eight, wide band — +1.1% to +27.3%. The Street keeps underestimating operating leverage; the printed surprise is large and the commercial mix still matters more.
Against ten software peers, Palantir prints best growth and best operating margin in the set — and the richest EV/Sales by far (~64× vs ~21× peer median). The operating story is exceptional; the sales multiple still requires it to keep printing.
| Company | Market cap | Rev growth | Gross margin | Op margin | ROIC | EV / Sales | P / E |
|---|---|---|---|---|---|---|---|
| Oracle ORCL | $423B | +17.4% | 65.8% | 30.8% | 8.0% | 8.1× | 24.7× |
| Palantir PLTR | $395B | +56.2% | 84.8% | 42.8% | 25.6% | 63.9× | 145.8× |
| Palo Alto PANW | $297B | +14.9% | 71.9% | 9.6% | 1.7% | 27.9× | 305.8× |
| CrowdStrike CRWD | $218B | +21.7% | 75.0% | -3.9% | 6.0% | 42.1× | n/m |
| Salesforce CRM | $158B | +9.6% | 77.6% | 21.9% | 9.2% | 4.5× | 22.2× |
| ServiceNow NOW | $129B | +20.9% | 74.8% | 11.4% | 5.5% | 9.2× | 77.6× |
| Snowflake SNOW | $115B | +29.2% | 67.1% | -26.1% | -27.2% | 22.9× | n/m |
| Cloudflare NET | $107B | +29.8% | 72.6% | -14.1% | -6.8% | 43.2× | n/m |
| Datadog DDOG | $83B | +27.7% | 79.5% | 0.4% | 0.2% | 21.2× | 470.5× |
| MongoDB MDB | $32B | +22.8% | 72.0% | -4.2% | -3.5% | 11.9× | n/m |
| Zscaler ZS | $27B | +23.3% | 76.7% | -4.7% | -3.2% | 8.9× | n/m |
| Peer median excl. PLTR | $122B | +22.3% | 73.7% | -1.8% | 0.9% | 16.6× | 77.6× |
Best growth and best OM in the set — richest sales multiple by far. Peers cluster near ~21× sales; Palantir sits near 64×. That gap prices commercial durability and Rule of 40 — which is why our base sits below the Street median even with a bullish rating.
Cheap against its own P/E history, expensive against peers on sales. Revisions have drifted down as the share price digested 2025. What is priced in: FY27 ~$12B at ~76× EPS.
Cheap on earnings history, expensive on sales — against peers. P/E sits at the 4th percentile of Palantir's profitable history (earnings catch-up under a flat share price). EV/Sales is the 52nd percentile of its own range and still ~3× the ~21× software peer median. The multiple still requires the growth to keep printing.
Accepting ~$12B in FY27 at ~76× EPS prices continued commercial torque. Our $185 is 5.1% below the Street median — SBC and programme concentration keep the discount of trust high even with a bullish rating.
The mean of $172.17 is dragged by lowballs toward $80; the median at $195 is the cleaner read. Our $185 sits between — above the mean, below the median.
Every assumption behind the three targets on page one is on this page. Valued primarily on FY27–28E EPS × exit multiple. Change one input and the number moves.
| Driver | Bear | Base | Bull | Why it moves |
|---|---|---|---|---|
| FY27E revenue | $11.2B | $12.1B | $13.1B | Street lo / mean / hi for FY27 |
| FY27E operating margin | 28% | 45% | 50% | Commercial mix vs SBC / growth fade |
| FY27–28E EPS | ~$1.7 | ~$2.3–3.4 | ~$4.1 | Aligned to Street EPS path |
| Diluted shares, B | 2.30 | 2.30 | 2.28 | ~2.30B shares · float ~95.5% |
| Exit P / E | 55× | ~80× | 55–70× | Bull is earnings, not multiple expansion |
| 12-month target | $95 | $185 | $230 | −44.8% / +7.6% / +33.7% |
| FY27–28E EPS | 40× | 55× | 70× | 85× | 100× |
|---|---|---|---|---|---|
| $1.7 | $68 | $94 | $119 | $145 | $170 |
| $2.3 | $92 | $126 | $161 | $195 | $230 |
| $2.8 | $112 | $154 | $196 | $238 | $280 |
| $3.4 | $136 | $187 | $238 | $289 | $340 |
| $4.1 | $164 | $225 | $287 | $348 | $410 |
The cell at ~$2.3 × 80× ≈ $184 sits on the $185 base. Bull is an earnings path (~$4.1 EPS), not multiple expansion — exit multiples stay in the mid-50s to 70s on the high EPS case.
Seven events over the next eighteen months. The two heaviest are a November print against ~$2.16B and an early-February guide against the Street's $12.1B FY27.
Commercial growth and Rule of 40 framing decide the multiple. Consensus wants ~$2.16B and $0.41 — but mix is the scorecard, not the top line alone.
| Commercial share / growth | Reads as | Target |
|---|---|---|
| Comm >50%, Ro40 held | Crossing confirmed; bull path opens | $230 |
| Mix rising, growth intact | Base intact; trust discount stays | $185 |
| US comm. <50% YoY | Deceleration; bear becomes the base | $95 |
Six risks ranked by what they would cost rather than how likely they are, each with the observable that tells you it is happening.
| Insider | Role | Filings | Value |
|---|---|---|---|
| Peter Thiel | Director | 7 | $290M |
| Alexander C. Karp | CEO | 85 | $249M |
| Stephen A. Cohen | President | 70 | $198M |
| Shyam Sankar | CTO | 102 | $193M |
| All insiders | 11 people | 530 | ~$1.03B |
Read this as texture, not signal. Karp and early holders sell persistently; ~$1B is still only ~26 basis points of a $395B company. The strip shows selling into strength — consistent with liquidity after a long run, not proof the thesis is wrong.
Bullish, but trust-discounted. A base below the Street median is the point — not a soft rating.
Each case sets revenue, margin and multiple independently, then values primarily on FY27–28E EPS at a twelve-month horizon. The headline is the base case.
Prices, consensus, targets and Form 4s from Financial Modeling Prep; fundamentals from Palantir SEC filings. Peer medians use the eleven-name set on page 4. Data as of 7 August 2026.
Generated from structured data — it cannot interview management or price an untold event. Institutional ownership was unavailable and is omitted rather than estimated.
Every figure above comes from Palantir Technologies Inc.’s own filings and live market data, assembled and written up automatically. Ticker Alpha is building this for any listed company — so the research you need on a Tuesday afternoon takes a minute, not a week.