A case study · as of June 4, 2026

Palantir: the most expensive stock in software, and one of the most contested

An independent, fully-cited, deliberately neutral teardown of Palantir Technologies — what actually drives its sudden growth re-acceleration, how durable its moat is, why its valuation splits Wall Street, and the ethics and governance questions that follow it.

NYSE: PLTR77 sourcesNeutral · evidence on both sides

Palantir spent its first two decades as a polarizing government-software company that lost money. Then enterprise AI arrived. In 2024–2026 it became the best-performing stock in the S&P 500, turned solidly profitable, and re-accelerated to growth most software companies never see at its size — while trading at a multiple almost no company has ever sustained.

In FY2025 Palantir grew revenue 56% to $4.475B, posted $1.625B of GAAP net income (a 36% margin) and a 127% Q4 "Rule of 40" score[39][41][42]; in Q1 2026 growth jumped to 85% as U.S. revenue crossed 100% growth for the first time[48]. Yet at roughly a $341B market cap the stock traded near 70× sales and 150× earnings[73][72][50] — and its work spans battlefield AI, immigration enforcement and national-health data, drawing both record contracts and sustained protest. The debate is no longer whether Palantir is real; it is whether the growth, the moat and the price can all be true at once. This site lays out both cases and leaves the verdict to you.

$4.475B
FY2025 revenue (+56% YoY)
GAAP net income $1.625B [41]
+85%
Q1 2026 revenue growth
U.S. commercial +133% [48]
127%
Q4 2025 Rule of 40 score
growth + margin [42]
~$341B
market cap (Jun 2026)
≈70× sales [73] [72]

From slow burn to vertical: the revenue trajectory

For years Palantir grew in the teens-to-twenties percent and burned cash. GAAP profitability arrived in 2023[45]; then growth inflected — $2.87B (2024, +29%) to $4.475B (2025, +56%), with the company guiding FY2026 to roughly $7.65B (~71%)[47][39][71]. The shape of that curve is the heart of both the bull and bear cases.

Palantir annual revenue (US$B), FY2021–FY2026E
202120222023202420252026E

FY2021–2025 actuals[44]; FY2026 is the company's own raised guidance[71].

The balance of evidence, at a glance

Why the bull case holds

  • Rare combination at scale: 56% growth and 36% GAAP net margin, a 127% Q4 Rule of 40, and ~$2.1B free cash flow in FY2025[42][46].
  • A re-accelerating engine: Q1 2026 revenue +85%, U.S. commercial +133%, net dollar retention 150%[48][49].
  • Government entrenchment: a $10B Army enterprise agreement and a ~$1.3B Maven ceiling anchor durable, hard-to-displace demand[11][12].
  • A claimed structural moat — ontology, forward-deployed engineers and FedRAMP High/IL5–6 accreditation[24][26]; bulls value it at $230–255[74][75].

Why the bear case holds

  • Valuation: ~70× sales and ~150× earnings vs. the S&P 500's ~3.6× and ~31×, leaving "no margin for error"[50][72].
  • Moat skepticism: short-sellers call it "locked-in consulting wrapped in software," and a16z warns the FDE model can fail to scale[29][28].
  • Concentration & governance: top three customers were 16% of revenue, founders hold ~49.99% voting control, and insiders sell ~9× for every 1 they buy[62][64][65].
  • Reputational risk: ICE deportation tooling, NHS data access and battlefield AI draw protest, divestment and former-employee dissent[51][57][53].
⚖️
What reasonable people disagree about: whether 2025–26 growth is a structural enterprise-AI inflection or an AI-cycle surge against easy comparisons[48]; whether the ontology + forward-deployed-engineer model is a durable moat or unscalable services[28]; whether any fundamentals justify ~70× sales[50]; and whether deep defense and government entanglement is Palantir's greatest asset or its greatest liability[11][51]. Each is genuinely contested in the sources.
🧭
This is an independent research compilation, not affiliated with Palantir and not investment advice. Figures are point-in-time as of June 4, 2026. See Methodology & Limitations for what may be wrong and Sources for the full bibliography.
Company & Timeline

A 20-year overnight success

Palantir spent most of its life as a secretive, money-losing government-software company. The story of the last three years is how that base became the launch pad for an enterprise-AI business.

Founded 2003HQ Denver, CODirect listing Sept 2020

Palantir was built on a post-9/11 bet that messy government data could be unified and acted on. That thesis produced Gotham (defense/intelligence, 2008) and Foundry(commercial/civil, 2016), but profitability only arrived in 2023 — just as AIP turned the AI wave into a growth engine[4][45].

What Palantir does

Palantir builds software that integrates an organization's scattered data into a single operational model and lets analysts and operators act on it. It sells three core platforms: Gotham, used by militaries, intelligence agencies and counter-terrorism analysts; Foundry, used across commercial and civil-government sectors; and AIP, launched in April 2023, which integrates large language models into customers' private networks[4]. Onboarding increasingly runs through five-day "boot camps" that build a working use case on the customer's own data[15].

As of the FY2025 annual report, Palantir drew 54% of revenue from government and 46% from commercial customers, with 74% from the United States and 26% from abroad, across 954 customers[9]. The company is headquartered in Denver, Colorado[8].

Timeline: founding to inflection

2003
Peter Thiel incorporates Palantir; Alex Karp becomes CEO, with cofounders Nathan Gettings, Joe Lonsdale and Stephen Cohen[1].
2004
The CIA's venture arm In-Q-Tel provides early backing; the post-9/11 mission is counter-terrorism data analysis[2][3].
2008
Gotham ships — the defense and intelligence platform that defines Palantir's first decade[4].
2016
Foundry launches for commercial and civil-government data integration, broadening beyond intelligence[4].
Aug 2020
Headquarters move from Palo Alto to Denver, Colorado[8].
Sep 2020
Direct listing on the NYSE (ticker PLTR) at a $7.25 reference price, ~$16B fully diluted, raising no primary capital[5].
Apr 2023
AIP, the Artificial Intelligence Platform, launches — wiring large language models into private networks[4].
2023
First full year of GAAP profitability[45].
Sep 2024
Added to the S&P 500, replacing American Airlines[6]; added to the Nasdaq-100 in December[7].
2024
Best-performing stock in the S&P 500, up ~340%[68].
Jul 2025
U.S. Army awards a $10B enterprise agreement, consolidating 75 contracts[11].
2025–26
Revenue re-accelerates: FY2025 +56% to $4.475B; Q1 2026 +85%[39][48].
🏛️
Why the origins still matter:the CIA/In-Q-Tel beginnings and the intelligence-community roots are simultaneously Palantir's deepest moat (trust, clearances, entrenchment) and the source of much of the controversy that follows it — a tension this case study returns to in Strategy & Moats and Risks & Controversies.
Market & Industry Structure

Two markets, one platform

Palantir straddles a vast, crowded commercial data-and-AI market and a smaller but stickier defense-and-government software market. The two have very different economics — and the second is increasingly where the money and the moat are.

Big-data analytics + defense softwareTAM est. ~$395B (2025)

On paper Palantir competes in a ~$395B big-data-analytics market growing in the low teens[10]. In practice its defensible position sits in defense and government, where 2025 brought a $10B Army agreement and a Maven contract ceiling near $1.3B — scale that is hard for rivals to replicate[11][12].

The commercial market: large, fast-growing, and crowded

The global big-data-analytics market was estimated at roughly $394.7B in 2025 and projected to grow to about $1,176.6B by 2034 (~12.8% CAGR)[10]. This is the headline TAM bulls point to — and it is real, but it is also where Palantir faces the most competition: cloud hyperscalers, Snowflake, Databricks, Microsoft Fabric and a long tail of analytics tools all sell into it (see Competitive Landscape). Market-sizing figures like these are third-party estimates and should be read as directional, not precise.

The defense market: smaller, slower-moving, far stickier

Palantir's heritage market is government and defense software, and it is in the middle of a structural shift: software-first vendors are taking share from, and being bought by, the traditional primes. In July 2025 the U.S. Army awarded Palantir an enterprise agreement capped at $10 billion over up to ten years, consolidating 75 contracts into one[11]. In May 2025 the Pentagon raised the ceiling on Palantir's Maven Smart System by $795M to nearly $1.3B through 2029, with more than 20,000 active Maven users across 35+ tools[12]; the department is moving to make Maven a broader, department-wide capability[13].

The same defense-tech wave has produced fast-growing peers: Anduril, founded by a former Palantir engineering director, is valued at about $31B and projects 2026 revenue of $4.3B[14]. That validates the category — and signals that the government-AI market is becoming contested too.

Where the money is

Why the market backdrop favors Palantir

  • A large, growing commercial TAM (~$395B, low-teens growth) gives the U.S. commercial business a long runway[10].
  • Defense budgets are flowing to software: a $10B Army agreement and an expanding Maven program anchor multi-year demand[11][13].
  • Procurement consolidation (75 contracts → 1) favors an incumbent platform with clearances already in place[11].

Why the market backdrop is a risk

  • The commercial TAM is the most competitive arena, with hyperscalers and well-funded data platforms attacking it[22][23].
  • Government demand is policy- and budget-dependent and can shift with administrations and appropriations[12].
  • The defense-tech wave is also producing direct rivals like Anduril, so "category creator" need not mean "category owner"[14].
🧭
The synthesis:Palantir's reported TAM is dominated by the commercial market, but its most defensible economics are in government. The investment question is whether it can win enough of the crowded commercial market to grow into its valuation without giving up the government stickiness that makes it special — explored in Strategy & Moats.
Business Model & Economics

Software with a services edge

Palantir sells long-term platform subscriptions, but it gets there in a way most software companies don't: by embedding its own engineers, running boot camps, and landing-then-expanding. That motion is its biggest strength — and the heart of the 'is it really software?' debate.

Platform subscriptions + usage82% gross marginNDR 150% (Q1 2026)

The model is a high-margin platform business — ~82% gross margin, 150% net dollar retention[46][49] — sold via forward-deployed engineers and boot camps that land a use case fast and expand it. Bulls call that a durable flywheel; critics call it services wearing a software multiple.

How Palantir makes money

Revenue comes from subscriptions to its three platforms — Gotham, Foundry and AIP — typically priced through custom, negotiated contracts rather than a public price list; expansion is driven by additional use cases, users and data volume[18]. The defining go-to-market mechanic is the forward-deployed engineer (FDE): unlike vendors that rely on customer teams or systems integrators, Palantir deploys its ownengineers on its own stack directly into a customer's production environment[16]. Since 2023, five-day AIP boot campshave become the funnel — building a real workflow on the customer's data, then converting to a paid, expanding deal[15][25].

A healthcare company completed two boot camps with us last summer and signed a $96 million deal with us before the end of the year.
Palantir management · Q4 2025 earnings call · Feb 2, 2026 · source

The mechanism shows up in the retention numbers: a utility customer expanded from $7M in annual contract value in Q1 2025 to $31M by year-end[17], and company-wide net dollar retention reached 150% in Q1 2026[49] — meaning existing customers alone grew spend by half.

Revenue mix: still a government company, increasingly a commercial one

Government and commercial were roughly balanced in FY2025 — 54% / 46% — but the growth is lopsided toward U.S. commercial, which grew 109% for the year and 133% in Q1 2026[9][40][48]. The mix is shifting under the surface even as the headline split looks even.

  • Government54%
  • Commercial46%

FY2025 10-K segment split[9]. Geographically, 74% U.S. / 26% international.

The economics, and the asterisk

The headline economics are strong: ~82% gross margins, a 32% GAAP operating margin and ~$2.1B of free cash flow in FY2025[46][41]. The most-cited asterisk is stock-based compensation, which was $684M in FY2025; critics argue it flatters non-GAAP margins and steadily dilutes shareholders, a concern amplified by heavy insider selling (covered in Risks)[19]. Palantir counters that it is now firmly GAAP-profitable even after expensing SBC[41].

Why the model is a strength

  • Boot camps compress the sales cycle and FDEs drive rapid expansion, producing 150% net dollar retention[25][49].
  • ~82% gross margins and ~$2.1B FCF show the model converts to cash, not just bookings[46][42].
  • a16z argues services here are a means to product adoption, not the revenue itself — unlike consulting[27].

Why the model is questioned

  • Short-sellers call it "locked-in consulting wrapped in software" that doesn't compound like a pure-software flywheel[29].
  • The FDE model is people-intensive; a16z itself warns copycats risk "thousands of bespoke deployments that are impossible to maintain"[28].
  • $684M of SBC remains a real dilution and margin-quality question for skeptics[19].
🧭
The synthesis: the numbers say "high-margin software"; the delivery model says "high-touch." Both are true. The unresolved question is whether the FDE/boot-camp engine keeps scaling efficiently as Palantir adds hundreds of commercial customers — the operating-leverage bet that the valuation depends on.
Competitive Landscape & Positioning

A 'category of one' — or surrounded?

Palantir insists it has no true peer. Look closer and it competes with different rivals on each front: defense consultants and primes in government, and the entire cloud-data-and-AI complex in commercial.

Five Forces2×2 positioning

Palantir calls itself "an n of 1"[20] — and on the full platform-plus-government-depth combination, that is roughly fair. But it faces real competition on each axis: Booz Allen, Leidos and the primes in government[21], and Databricks, Snowflake, Microsoft and the hyperscalers in commercial[22][23].

Who Palantir competes with

In defense and government, the named competitors are services-and-integration players — Booz Allen Hamilton, Leidos, CACI — plus the traditional primes (RTX, Northrop Grumman); Palantir differentiates with a single productized platform and forward-deployed engineers rather than staff-augmentation consulting[21]. In commercial data and AI, the alternatives are Databricks, Snowflake, Microsoft Fabric, Alteryx and C3.ai; some customers report a lower total cost of ownership on those platforms and prefer them to Palantir's more "ontology-heavy" approach[22]. Layered on top, hyperscalers like Microsoft are embedding AI into tools enterprises already pay for, pressuring all standalone enterprise-AI vendors[23].

Palantir's own framing is that none of these is a true substitute for an end-to-end operational system. The evidence is genuinely mixed: it is the only scaled player combining a unified platform, deep government accreditation and embedded delivery — but "no exact comparable" is not the same as "no competition."

Porter's Five Forces: the AI-software arena

Click a force for the rating and its basis. Ratings are this study's analysis of the cited evidence.

Enterprise & government AI software
Competitive rivalryHigh pressure. Crowded commercial arena (Databricks, Snowflake, Microsoft Fabric, C3.ai) plus a rising defense-tech field (Anduril); "category creator" does not guarantee category ownership[22][14].

Positioning: platform breadth vs. government depth

Hover or tap a company. Placements are this study's qualitative read of the cited evidence, not precise coordinates.

Point toolFull platformCommercial-focusedGovernment / defense-focusedPalantirSnowflakeDatabricksMicrosoftC3.aiBooz AllenAnduril

Palantir: End-to-end platform (Gotham/Foundry/AIP) with deep government accreditation and a large commercial business — the one player high on both axes.

Is "n of 1" defensible?

The case that it's genuinely differentiated

  • No other scaled vendor combines a unified operational platform, FedRAMP High/IL5–6 accreditation and embedded engineers[26][16].
  • Government entrenchment (a $10B Army agreement, expanding Maven) is extremely hard for commercial rivals to match[11][13].
  • Model-agnostic AIP lets it ride, rather than fight, the foundation-model race[4].

The case that it's surrounded

  • On each individual front there are strong, often cheaper, rivals (Databricks/Snowflake in data, Booz Allen/primes in gov)[22][21].
  • Hyperscalers bundle AI into tools customers already buy, attacking the commercial value proposition[23].
  • The defense-tech wave is minting direct rivals like Anduril, so the government moat is not static[14].
🧭
The synthesis: "category of one" is closest to true for the combination Palantir offers, weakest when you decompose it. The competitive question is less "who is exactly like Palantir?" and more "can good-enough, cheaper alternatives satisfy most buyers most of the time?"
Strategy & Moats

The ontology, the engineers, and the clearances

Palantir's strategy is to be the operational layer organizations run on. Its claimed moats — the ontology, forward-deployed engineers, and government accreditation — are real; whether they're durable against hyperscalers and copycats is the live question.

Ontology · FDEs · FedRAMP HighSWOT

The stated strategy is simple: land via boot camps, expand via the ontology, entrench via clearances. The moat is strongest where all three combine — high-stakes, regulated, data-heavy operations. The bear case is that the engineer-heavy delivery doesn't scale and the data layer gets commoditized[28][22].

Stated strategy vs. revealed strategy

What Palantir says: it is building the software organizations use to turn data and AI into decisions, and boot camps are the go-to-market engine for that[25]. What the numbers reveal: the strategy is working fastest in U.S. commercial (revenue +133% in Q1 2026) while government provides the durable, high-ceiling base[48][11]. The two reinforce each other — credibility earned in intelligence and defense de-risks Palantir for regulated commercial buyers.

The three claimed moats

1. The ontology. Palantir describes its ontology as a "digital twin of the organization," binding together data objects, their relationships, and the actions/permissions that operate on them[24]. Once a customer's operations are modeled this way, switching means rebuilding the model — a genuine source of switching cost.

2. Forward-deployed engineers. Palantir embeds its own engineers in customer production; a16z argues that, unlike consulting, these services exist to drive product adoption rather than to be the revenue[27]. The flip side: the model is people-intensive, and a16z itself warns that copying only the embedded-engineer part yields "thousands of bespoke deployments that are impossible to maintain"[28].

3. Government accreditation. Palantir holds FedRAMP High plus DoD Impact Level 5 and 6 authorizations across its full suite, clearing it for the most sensitive unclassified workloads[26]. Accreditation, clearances and incumbency are slow and expensive for rivals to replicate — the clearest, most defensible moat of the three.

SWOT

Strengths

  • Re-accelerating, profitable growth: FY2025 +56%, 36% GAAP net margin, 127% Q4 Rule of 40[39][42]
  • Deep government entrenchment: $10B Army agreement, ~$1.3B Maven, FedRAMP High/IL5–6[11][26]
  • High switching costs once the ontology is embedded; 150% net dollar retention[24][49]

Weaknesses

  • People-intensive FDE delivery raises scaling and margin-mix questions[28]
  • Customer concentration: top three customers were 16% of revenue[62]
  • $684M of stock-based compensation and ongoing dilution[19]

Opportunities

  • Large commercial TAM (~$395B) with U.S. commercial growing >100%[10][48]
  • Defense-software budget tailwinds and procurement consolidation[11][13]
  • Model-agnostic AIP positions it to monetize whichever LLMs win[4]

Threats

  • Hyperscalers bundling AI into existing enterprise tools[23]
  • Commoditization by Databricks/Snowflake; some buyers cite lower TCO[22]
  • Valuation, reputational and governance overhangs (see Risks)[50][64]

How durable is the moat?

Durable

  • Accreditation + incumbency in classified/regulated settings is slow and costly to replicate[26].
  • The ontology creates real switching costs; retention of 150% shows expansion, not churn[24][49].
  • a16z frames the FDE model as a genuine product-adoption flywheel, not billable hours[27].

Erodable

  • Short-sellers argue the business is "locked-in consulting wrapped in software" that doesn't compound[29].
  • The data/integration layer is exactly what hyperscalers and lakehouse vendors are commoditizing[22][23].
  • People-intensive delivery may cap operating leverage as commercial scales[28].
🧭
The synthesis: the government-accreditation moat looks the most durable; the ontology moat is real but contestable; the FDE moat is the most debated. The valuation (next sections) implicitly bets that all three hold and compound for years.
Peer Comparison & Benchmarking

Fast and profitable — and priced like nothing else

Against data-and-AI platforms and enterprise-software peers, Palantir's growth-plus-profitability profile is genuinely top-tier. Its valuation multiple is in a category by itself.

6 named peerslatest reported FY

On fundamentals Palantir sits near the top: 56% growth with a 36% GAAP net margin is rare[39][41]. Only private Databricksgrows faster (>65%) and it is not yet as profitable[33]. But on valuation — ~42× forward sales vs. ServiceNow's ~ — Palantir is an outlier[37].

The benchmark table

Latest reported fiscal year per company; figures rounded. Margins are GAAP unless noted. Multiples are third-party and move daily; treat them as directional.

CompanyRevenue (latest FY)GrowthGross marginGAAP profitable?Valuation note
Palantir$4.48B (FY2025)[39]+56%[39]~82%[46]Yes — $1.63B NI[41]~$341B cap; ~42× fwd sales[73][37]
Databricks (private)~$5.4B run-rate[33]>65%[33]n/dFCF-positive, not GAAP-stated[33]$134B private valuation[33]
Snowflake$4.47B product (FY2026)[32]+29%[32]~75% (non-GAAP)[32]No (non-GAAP only)[32]~7× out-year sales[67]
ServiceNow$13.28B (FY2025)[35]+21%[35]~79%*Yes — $1.75B NI[35]~6× fwd sales[37]
Datadog$3.43B (FY2025)[38]+28%[38]~80%*Slim GAAP op. loss[38]high-growth SaaS multiple
Booz Allen Hamilton$11.2B (FY2026)[36]~flat/declining[36]services (low)Yes (services)[36]low single-digit P/S
C3.ai$0.25B (FY2026)[34]−36% (fell)[34]31%[34]No — $470M loss[34]AI-software cautionary case[34]

"n/d" = not disclosed comparably. Snowflake reports non-GAAP product gross margin; Booz Allen is a services business with structurally lower gross margins. * ServiceNow and Datadog gross margins (~79–80%) are industry-typical software figures shown for context, not separately cited here.

Growth: top-tier, but not alone

Latest-FY revenue growth (%)
Databricks
>65%
Palantir
+56%
Snowflake
+29%
Datadog
+28%
ServiceNow
+21%
C3.ai
−36%

C3.ai's revenue actually fell ~36%; its bar is shown at a floor for visibility[34]. Sources: Palantir[39], Databricks[33], Snowflake[32], Datadog[38], ServiceNow[35].

Valuation: the real outlier

Sales multiple vs. peers (× revenue; forward / out-year, mixed bases)
Palantir
~42×
Cloudflare
16×
CrowdStrike
13×
Snowflake
ServiceNow

Palantir ~42× forward sales vs. ServiceNow ~6×[37]; Jefferies cites Cloudflare 16×, CrowdStrike 13×, Snowflake 7× on out-year revenue[67]. Bases differ slightly between sources, so read this as the shape of the premium, not a precise ranking.

📊
What the comps show: Palantir earns a premium on its rare growth-plus-profitability profile — but the size of the premium (roughly 6–7× the multiple of a profitable, 21%-growing ServiceNow) is what the bull and bear cases fight over in Forward View.
Financials & Growth

The numbers are excellent. The multiple is the argument.

Palantir's financials flipped from cash-burning to cash-generating, then re-accelerated. Almost no one disputes the fundamentals; the entire bear case lives in the valuation paid for them.

FY2025 & Q1 2026GAAP profitable since 2023

FY2025: revenue $4.475B (+56%), GAAP net income $1.625B (36% margin), free cash flow ~$2.1B, and a Q4 Rule of 40 of 127%[39][41][46][42]. Q1 2026 then accelerated to +85% with a 145% Rule of 40[48][49]. The catch: the stock trades near 70× salesand 150× earnings[50][72].

$4.475B
FY2025 revenue
+56% YoY [39]
$1.625B
FY2025 GAAP net income
36% margin [41]
~$2.1B
FY2025 free cash flow
~82% gross margin [46]
150%
net dollar retention (Q1'26)
up 1,100 bps QoQ [49]

From losses to real profit

Palantir lost $520M in 2021 and $374M in 2022, turned GAAP-positive in 2023 ($210M), then scaled net income to $462M (2024) and $1.625B (2025)[45][41]. That inflection — profitability arriving just as growth re-accelerated — is what changed the story.

Palantir GAAP net income (US$B), FY2021–FY2025
20212022202320242025

GAAP net income by fiscal year[45][41].

The growth re-acceleration

Growth had decelerated to the high-teens by 2023; then AIP demand reversed it. Full-year growth went 17% → 29% → 56% (2023–2025), and Q1 2026 reached 85% as total U.S. revenue grew 104% — the first time it crossed 100%[44][39][48]. U.S. commercial is the engine: +109% in FY2025 and +133% in Q1 2026[40][48]. Total remaining deal value reached $11.8B (Q1 2026), giving visibility into future quarters[49].

The bull's favorite stat: a 127% Rule of 40 in Q4 2025 (rising to 145% in Q1 2026; the full-year 2025 score was 106%) — growth plus margin far above the 40% benchmark that defines a healthy software company[42][49].

The valuation, stated plainly

Entering 2026 Palantir traded at a price-to-sales ratio of 100 — described as "unprecedented." Even after a ~34% drawdown it sat at a P/S of 67 and a P/E of 153, versus the S&P 500's P/S of 3.6 and P/E of 31[50]. By early June 2026, with the stock near $160.65, the forward P/E was ~145× and price-to-sales ~71.8×[72]. The fundamentals are excellent; the price assumes they stay that way for a very long time.

The fundamentals case

  • 56% growth with a 36% GAAP net margin and ~$2.1B FCF is a genuinely rare profile at this scale[39][41][46].
  • Acceleration, not deceleration: Q1 2026 +85%, NDR 150%, $11.8B remaining deal value[48][49].
  • Rule of 40 of 127% (Q4'25) to 145% (Q1'26) sits far above the 40% software benchmark[42][49].

The valuation case

  • ~70× sales / ~150× earnings vs. the market's ~3.6× / ~31× — extreme even versus elite SaaS[50].
  • $684M of stock-based compensation tempers margin quality and dilutes holders[19].
  • "No margin for error": at this multiple, a single soft quarter can re-rate the stock sharply[72].
🧭
The synthesis: separate the business from the stock. The business is one of the best growth-and-profit stories in software. Whether the stock is attractive depends almost entirely on what multiple you believe such a business can sustain — the subject of the Forward View.
Risks, Ethics & Controversies

The contested half of the story

No major software company is followed by as much ethical, governance and valuation controversy as Palantir. This section presents the criticisms — each attributed — alongside Palantir's responses, and lets you weigh them.

Ethics · governance · valuationBoth sides, attributed

The same government work that anchors Palantir's demand — immigration enforcement, battlefield AI, national-health data — draws sustained protest, divestment and former-employee dissent[51][57][53]. Palantir consistently responds that it operates lawfully, as a data processor whose customers control their data[59][61]. Both the contracts and the criticism are real.

Ethics & civil liberties

Immigration enforcement. Per the ACLU, ICE entered a $30Mcontract with Palantir in April 2025 to build "ImmigrationOS," upgrading ICE's case-management system for deportation operations; the ACLU argues the campaign's "cruel, brutal, and lawless character inevitably carries over" to companies that facilitate it[51][52]. In May 2025, thirteen former Palantir employees signed an open letter saying the company's founding principles "have now been violated"[53].

Battlefield AI.Palantir's Maven Smart System is designed to "autonomously detect, tag and track objects or humans of interest" from surveillance imagery[54] — a capability critics frame as automating targeting, and which the company and CEO frame as deterrence. Alex Karp has invoked the claim that the West rose through "superiority in applying organized violence"[55], and defends Palantir's defense work in blunt terms.

something really bad is going to happen to you and your friends and your cousins and your bank account and your mistress and whoever was involved.
Alex Karp · CEO, Palantir, on deterring adversaries · 2025 · source

Israel & divestment. Norway's Storebrand Asset Management divested roughly $24M from Palantir, citing concern that its sales to Israel for use in occupied Palestinian territories risked violating international humanitarian law[57]. Karp has been openly proud of supporting Israel.

Surveillance & health data. A New York Times report described Palantir helping consolidate federal data into a searchable system; Palantir called the article "blatantly untrue" and said it is "not a vendor on any master database project to unify databases across federal agencies"[58][59]. In the UK, the Financial Times reported NHS England gave Palantir staff "unlimited" access to patient data in a data-integration tenant[60]; Palantir UK replied that it acts "exclusively as a data processor" and never uses NHS data for its own purposes[61].

Governance & concentration

Palantir's 10-K flags that its multi-class structure and Founder Voting Trust concentrate voting power with Cohen, Karp and Thiel[63]; a shareholder suit alleged the Class F structure made the founders "emperor for life," fixing their voting power at roughly 49.99% untethered from equity — though the founders dismissed the claim as "a raft of hyperbole," noting the structure was approved by investors at the 2020 IPO and is permitted under Delaware law[64]. Customer concentration is disclosed: the top three customers were 16% of revenue in 2025[62]. And insider selling is heavy — a roughly 9.3-to-1 sell-to-buy ratio over three months, with Thiel selling more than 2 million shares on a single day in March 2026[65]. Palantir characterizes such sales as pre-arranged diversification; critics, including Jim Chanos, note the optics of executives selling while "ridicul[ing] the skeptics"[30].

Valuation & market risk

The most quantifiable risk is the multiple itself. RBC's Rishi Jaluria rates the stock Underperform with a $90 target, arguing the price embeds "years of flawless execution"[66]; Jefferies (Underperform, $70) says "multiple downside outweighs fundamental upside"[67]. At ~70× sales, even excellent results can disappoint a price that assumes perfection.

Weighing it

Why the risks may be manageable

  • Palantir consistently positions itself as a lawful data processor; customers, not Palantir, own and control the data[59][61].
  • Government entanglement also means durable, mission-critical demand and high barriers to displacement[11][26].
  • Insider sales are largely pre-arranged plans, and the company is now solidly GAAP-profitable[65][41].

Why the risks may bite

  • Reputational exposure is real: protests, a $24M divestment, and an open letter from former staff[57][53].
  • Founder-controlled voting and 16% customer concentration limit external checks and add fragility[64][62].
  • At ~70× sales, valuation alone is a standalone risk regardless of the ethics debate[66][67].
⚖️
How to read this section:the criticisms are attributed to named sources (ACLU, former employees, Storebrand, the FT/NYT, RBC, Jefferies, Chanos), and Palantir's rebuttals are quoted alongside them. Whether you view the government work as responsible mission-fulfilment or as ethically fraught is exactly the judgment this case study leaves to you.
Forward View

Three scenarios, not a prediction

Palantir's future hinges on a few measurable variables. Rather than pick a winner, here are the bull, base and bear paths, what would trigger each, and what to watch — so you can weigh them yourself.

Bull / base / bearAnalyst targets $40–$255

Wall Street's price targets span $40 to $255 — one of the widest disagreements on any mega-cap[76][75]. That spread is the story: the fundamentals are agreed, the right multiple is not. The decisive variables are U.S. commercial growth, the durability of the moat, and whether the AI-software premium holds.

The disagreement, in price targets

Analyst price targets vs. ~$161 price (Jun 2026)
BofA
$255
Citi
$235
Wedbush
$230
Morgan Stanley
$205
Consensus
$174
Price (Jun'26)
$161
RBC
$90
Jefferies
$70
Citron
~$40

Bull targets: BofA $255, Citi $235, Wedbush $230, Morgan Stanley $205[75][74]. Consensus ~$173.76[77]. Bear: RBC $90[66], Jefferies $70[67], Citron ~$40[76]. Price as of June 3, 2026[72].

The three scenarios

Bull

$230–255 (Wedbush / BofA / Citi)

U.S. commercial growth stays above 100%, AIP becomes "the enterprise AI standard," and government scales on the $10B Army agreement and an expanding Maven. Palantir grows into a trillion-dollar AI platform and the premium multiple holds[74][75][11].

Watch: U.S. commercial growth, net dollar retention, Rule of 40 staying >100%.

Base

~$174 consensus (29 analysts, Buy)

Growth stays strong but normalizes from triple digits; margins and FCF compound. The stock roughly tracks fundamentals as the multiple slowly compresses toward (still-rich) software norms — the 29-analyst consensus target sat near $173.76 with a Buy rating[77].

Watch: Whether the multiple compresses faster than earnings grow.

Bear

$40–90 (Citron / RBC / Jefferies)

Any growth wobble, government-budget shift, or AI-cycle cooling triggers a sharp de-rating from ~70× sales. Critics argue "multiple downside outweighs fundamental upside," and even at OpenAI's 17× sales Palantir would be ~$40[67][66][76].

Watch: A single soft quarter, a guidance cut, or a sector-wide AI re-rating.

The questions that decide it

  • Does U.S. commercial stay above ~100% growth? The whole re-acceleration thesis rests on it[48].
  • Does the moat compound or get commoditized? Ontology + clearances vs. hyperscalers and lakehouse vendors[24][22].
  • Does the AI-software premium hold? A sector-wide re-rating would hit the highest multiple hardest[67].
  • Does government demand and reputation stay favorable? Budgets, administrations and the ethics debate all feed in[11][51].
🧭
The honest conclusion: Palantir is, by the fundamentals, one of the strongest software businesses of its generation — andpriced such that being right about the business is not enough; you also have to be right about the multiple. Reasonable, well-informed investors land in all three scenarios above. This case study's job is to make the variables legible, not to choose for you.
Methodology & Limitations

How this was made, and where it may be wrong

A research compilation is only as good as its honesty about its own limits. Here is the method, the framework set, and the claims to treat with caution.

As of June 4, 2026Neutral compilation

Method

Research proceeded by fan-out web search across ten question areas and direct fetching of primary and reputable secondary sources. Palantir's own SEC filings, earnings releases, earnings-call transcripts and official blog were preferred, followed by reputable secondary press (Fortune, Bloomberg Law, DefenseScoop, Al Jazeera, The Motley Fool) and named data providers (StockAnalysis, TIKR). Every URL cited on the Sources page was opened and read during research; no link was reconstructed from memory. Each claim was transcribed into a structured manifest tagged with a tier (1–3), a confidence level and a stance — 77 sourcesin all (16 Tier-1, 50 Tier-2, 11 Tier-3; stance mix 28 supporting / 21 critical / 28 neutral, all English-language as befits a U.S.-headquartered company). The load-bearing figures are Palantir's FY2025 revenue, GAAP net income, free cash flow and Rule of 40; the Q1 2026 acceleration; the segment and geographic mix; the major government contracts; and the valuation multiples that frame the debate.

Frameworks used

The analysis applies the Pyramid Principle for answer-first synthesis, Porter's Five Forces to read industry structure, peer benchmarking against Snowflake, Databricks, ServiceNow, Datadog, Booz Allen and C3.ai, a SWOT to organize internal and external factors, a 2×2 positioning map of platform breadth versus government depth, and bull/base/bear scenarios for the forward view. BCG growth-share, Ansoff and the McKinsey 7S model were deliberately skipped because the clean, non-decorative data they require was not available here.

Disclosed vs. estimated

Disclosed figures are those Palantir itself reports — revenue, segment and geographic mix, GAAP net income, margins, free cash flow, remaining deal value, customer counts and guidance. Several Palantir primary pages (SEC EDGAR, Business Wire, investors.palantir.com) returned bot-blocking responses to the automated fetcher, so a number of Tier-1 figures are sourced from the sameofficial release text mirrored on Yahoo Finance or transcribed in The Motley Fool's earnings-call transcript; the figures cross-check consistently across sources[39][43]. Market-size (TAM), valuation multiples, private-company figures (Databricks) and price targets are third-party and move daily; they are labeled as estimates throughout[10][33][50].

⚠️
Where this case study may be wrong
  • Some primary pages blocked automated fetching. Palantir IR/SEC/Business Wire returned 403s; affected Tier-1 figures were corroborated via mirrored official text and the earnings-call transcript, and may resolve in a browser[39][48].
  • Valuation multiples are point-in-time and inconsistent across sources (forward vs. trailing, this-year vs. out-year sales). We show ranges and label the basis rather than a single number[50][72][67].
  • Peer comparison mixes bases. Snowflake reports non-GAAP product margins, Databricks is private, Booz Allen is a services business — the table is directional, not strictly like-for-like[32][33][36].
  • Positioning-map and chart coordinates are analytical illustrations based on the cited evidence, not precise data points.
  • Contested ethics claims rely partly on advocacy and press reporting. ICE/ImmigrationOS, the NHS "unlimited access" and federal "master database" items are attributed to the ACLU, FT and NYT respectively; Palantir disputes several, and its rebuttals are quoted alongside[51][58][59][60].
  • Insider-selling figures come from a secondary tracker and reflect a three-month window, not a full audited picture[65].
  • This is point-in-time. Figures are as of June 4, 2026; the commercial ramp, government budgets, the AI-software multiple and the ethics debate are all moving[71].

Neutrality & independence

This is a compilation, not an argument: each section deliberately pairs the case for and the case against, so supporting and critical evidence sit side by side and you can reach your own conclusion. Critical claims are attributed to named sources; Palantir's responses are quoted wherever available. The study is not affiliated with Palantir, and it is point-in-time as of June 4, 2026.

🧭
This case study is independent and not affiliated with, sponsored by, or endorsed by Palantir Technologies Inc. It is for informational and educational purposes only and is not investment, legal, or financial advice. All trademarks belong to their owners.
Sources

Full bibliography

Every load-bearing claim on this site links here. Each source was fetched during research; grouped by section, with tier, stance, and confidence shown.

77 sources16 Tier-150 Tier-211 Tier-3
📊
Stance mix: 28 supporting · 21 critical · 28 neutral. Tiers:Tier-1 = primary (Palantir SEC filings, earnings releases & earnings-call transcript, Palantir's own blog, U.S. Army / S&P / Nasdaq press releases, Palantir docs); Tier-2 = reputable secondary (Fortune, Bloomberg Law, DefenseScoop, Al Jazeera, The Motley Fool, Benzinga, Britannica, StockAnalysis); Tier-3 = tertiary/advocacy, used for context or attributed sentiment (ACLU, Wikipedia, market-sizing firms, insider-tracking sites). All sources are English-language (U.S. company).

Company & Timeline

  1. Palantir was founded in 2003 by Peter Thiel, who brought on Alex Karp as CEO, PayPal engineer Nathan Gettings, and Stanford students Joe Lonsdale and Stephen Cohen.

    In 2003, Peter Thiel, cofounder of PayPal (PYPL), incorporated Palantir Technologies, bringing on board Stanford Law School colleague Alex Karp as CEO, PayPal engineer Nathan Gettings, and Stanford University students Joe Lonsdale and Stephen Cohen.

    https://www.britannica.com/money/Palantir-Technologies-Inc
  2. Palantir's first major breakthrough came in 2004 with early backing from In-Q-Tel, the venture-capital arm of the CIA; the cofounders aimed to apply post-9/11 counter-terrorism data techniques.

    Thiel funded its initial launch, but the company's first major breakthrough came in 2004 when it secured early backing from In-Q-Tel, the venture capital arm of the U.S. Central Intelligence Agency (CIA).

    https://www.britannica.com/money/Palantir-Technologies-Inc
  3. After the September 11, 2001 attacks, Palantir's cofounders set out to bolster U.S. counter-terrorism efforts.

    After the terrorist attacks of September 11, 2001, Palantir's cofounders hoped to bolster the United States' counter-terrorism efforts.

    https://builtin.com/articles/what-is-palantir
  4. [4]Palantir Technologies — WikipediaTier 3neutralHigh confidence

    Gotham, released in 2008, is Palantir's defense and intelligence software; Foundry serves commercial and civil-government sectors; in April 2023 Palantir launched its Artificial Intelligence Platform (AIP), integrating large language models into private networks.

    Released in 2008, Palantir Gotham is Palantir's defense and intelligence software. ... Palantir Foundry is a software platform offered for use in commercial and civil government sectors. ... In April 2023, the company launched Artificial Intelligence Platform (AIP), which integrates large language models into privately operated networks.

    https://en.wikipedia.org/wiki/Palantir_Technologies
  5. Palantir went public via direct listing on the NYSE under ticker PLTR on September 30, 2020; the NYSE set a reference price of $7.25, valuing it at about $16 billion fully diluted, and the listing raised no primary capital.

    On a fully diluted basis, which Palantir says represents 2.2 billion shares total according to its most recent S-1 filing, the company is valued at $16 billion. ... as it is pursuing a direct listing, it will raise no primary capital as part of its debut.

    https://techcrunch.com/2020/09/29/palantir-reference-price/
  6. Palantir was added to the S&P 500 effective prior to the open of trading on Monday, September 23, 2024, replacing American Airlines.

    Palantir Technologies Inc. (NYSE: PLTR), Dell Technologies Inc. (NYSE: DELL), and S&P MidCap 400 constituent Erie Indemnity Co. (NASD:ERIE) will replace American Airlines Group Inc. (NASD:AAL), Etsy Inc. (NASD:ETSY) and Bio-Rad Laboratories Inc. (NYSE:BIO) in the S&P 500 ... effective prior to the open of trading on Monday, September 23

    https://press.spglobal.com/2024-09-06-Palantir-Technologies,-Dell-Technologies,-and-Erie-Indemnity-Set-to-Join-S-P-500-Others-to-Join-S-P-MidCap-400-and-S-P-SmallCap-600
  7. Palantir was added to the Nasdaq-100 Index effective prior to market open on Monday, December 23, 2024.

    The following three companies will be added to the Index: Palantir Technologies Inc. (Nasdaq: PLTR), MicroStrategy Incorporated (Nasdaq: MSTR), and Axon Enterprise, Inc. (Nasdaq: AXON). ... which will become effective prior to market open on Monday, December 23, 2024.

    https://www.globenewswire.com/news-release/2024/12/14/2997099/0/en/Annual-Changes-to-the-Nasdaq-100-Index.html
  8. [8]Palantir Technologies — WikipediaTier 3neutralHigh confidence

    Palantir moved its headquarters from Palo Alto to Denver, Colorado in August 2020.

    In August 2020, Palantir Technologies moved its headquarters to Denver, Colorado.

    https://en.wikipedia.org/wiki/Palantir_Technologies
  9. Per Palantir's FY2025 10-K, the company derived 54% of revenue from government and 46% from commercial customers, 74% from the U.S. and 26% abroad, and had 954 customers as of December 31, 2025.

    54% from government customers and 46% from commercial customers ... 74% of our revenue from customers in the United States, and 26% from those abroad ... As of December 31, 2025, we had 954 customers

    https://www.stocktitan.net/sec-filings/PLTR/10-k-palantir-technologies-inc-files-annual-report-74b188995295.html

Market & Industry Structure

  1. The global big-data-analytics market was estimated at about $394.7 billion in 2025, projected to reach roughly $1,176.6 billion by 2034 (≈12.8% CAGR) — a third-party market-sizing estimate.

    The global big data analytics market size was valued at USD 394.70 billion in 2025. The market is projected to grow from USD 447.68 billion in 2026 to USD 1,176.57 billion by 2034, exhibiting a CAGR of 12.80%

    https://www.fortunebusinessinsights.com/big-data-analytics-market-106179
  2. In July 2025 the U.S. Army awarded Palantir an Enterprise Agreement capped at $10 billion over up to 10 years, consolidating 75 contracts (15 prime, 60 related) into one.

    Through this EA, the Army consolidates 75 contracts, comprised of 15 prime contracts and 60 related contracts, into a single contract ... not to exceed the $10 billion cap.

    https://www.army.mil/article/287506/u_s_army_awards_enterprise_service_agreement_to_enhance_military_readiness_and_drive_operational_efficiency
  3. In May 2025 the Pentagon raised the ceiling on Palantir's Maven Smart System contract by $795 million, lifting it to nearly $1.3 billion through 2029; Maven had more than 20,000 active users across 35+ military tools, more than doubling since January.

    Pentagon leaders opted to boost the existing contract ceiling for Palantir Technologies' Maven Smart System by $795 million ... to nearly $1.3 billion through 2029 ... more than 20,000 active Maven users across more than 35 military service and combatant command software tools

    https://defensescoop.com/2025/05/23/dod-palantir-maven-smart-system-contract-increase/
  4. The Pentagon is moving to formalize Palantir's Maven Smart System into a broader, department-wide capability with wider adoption across the services.

    The Pentagon is moving to expand that designation into a broader, department-wide program with wider adoption across the services.

    https://www.military.com/feature/2026/03/22/pentagon-expands-palantirs-role-ai-contract.html
  5. A software-first defense-tech wave is challenging legacy primes; peer Anduril — founded by a former Palantir engineering director — is valued at roughly $31 billion and projects 2026 revenue of $4.3 billion, up from $2.2 billion in 2025.

    The company confirmed to Fortune that its revenue projections for 2026 are $4.3 billion, up from $2.2 billion in 2025.

    https://fortune.com/2026/05/06/anduril-ceo-brian-schimpf-defense-tech-military-pentagon-palmer-luckey/

Business Model & Economics

  1. [15]Palantir Technologies — WikipediaTier 3neutralHigh confidence

    Palantir sells three platforms — Gotham (intelligence/defense), Foundry (commercial and civil government), and AIP (launched April 2023, integrating LLMs into private networks) — and onboards prospective customers via five-day boot camps.

    Palantir offers five-day boot camps to onboard prospective customers

    https://en.wikipedia.org/wiki/Palantir_Technologies
  2. Unlike typical software vendors that rely on customer teams or partners to implement, Palantir deploys its own 'forward-deployed engineers' on its own stack directly into customer production.

    While software companies typically rely on customer teams or partners to implement, Palantir deploys its own people, on its own stack, directly into production

    https://www.everestgrp.com/palantir-inside-the-category-of-one-forward-deployed-software-engineers-blog/
  3. AIP boot camps drive a land-and-expand motion: a healthcare company completed two boot camps and signed a $96 million deal within the year; a utility expanded from $7M ACV in Q1 2025 to $31M ACV by year-end.

    A healthcare company completed two boot camps with us last summer and signed a $96 million deal with us before the end of the year ... A utility company expanded from $7 million ACV in Q1 2025 to $31 million ACV by year-end

    https://www.fool.com/earnings/call-transcripts/2026/02/02/palantir-pltr-q4-2025-earnings-call-transcript/
  4. Palantir pricing is opaque and custom: the platform is accessible via private pricing only, negotiated by use case, data volume, users and customization, with no meaningful public price list.

    Palantir Platform is accessible via private pricing only, with the public price being a placeholder and actual payment potentially different depending on many factors

    https://linkgo.dev/faq/the-pricing-structure-for-palantir
  5. Stock-based compensation was $684 million in FY2025 (down ~1% from $692M in 2024); critics note it remains a meaningful source of dilution alongside heavy insider selling.

    FY2025 annual result came in at $684.0 million, down 1.1% from the prior year

    https://businessquant.com/metrics/pltr/share-based-compensation

Competitive Landscape & Positioning

  1. Palantir positions itself as having no true peer; in February 2026 the company described itself as 'an n of 1,' the only company exclusively focused on scaling operational leverage from advancing AI models.

    We are an n of 1 ... the only company 'choosing to exclusively focus on scaling the operational leverage made possible by the rapid advancements of AI models'

    https://fortune.com/2026/02/02/palantir-record-earnings-stock-up-8-percent-alex-karp-n-of-1/
  2. In defense and government, Palantir's named competitors include Booz Allen Hamilton, Leidos, CACI, RTX and Northrop Grumman; Palantir differentiates via forward-deployed engineers and a single unified platform rather than consulting.

    Booz Allen Hamilton – 'serves as a consultant, helping design, build, and deploy analytics and AI tools' ... Leidos – 'delivering data integration, analytics, and AI tools to defense and intelligence agencies'

    https://www.fool.com/investing/how-to-invest/stocks/palantir-competitors/
  3. [22]Top 5 Best Palantir Competitors in 2026 — IBTimesTier 3criticalMedium confidence

    In data/AI platforms, Databricks, Snowflake, Microsoft Fabric, Alteryx and C3.ai are cited as Palantir alternatives; some commercial customers report lower total cost of ownership than Palantir and prefer less 'ontology-heavy' platforms.

    Customers praise its performance on massive datasets and lower total cost of ownership compared with Palantir in many commercial use cases ... key advantages over more rigid, ontology-heavy platforms

    https://www.ibtimes.com.au/top-5-best-palantir-competitors-2026-led-databricks-snowflake-microsoft-fabric-data-ai-platforms-1865435
  4. Analysts note hyperscalers such as Microsoft are embedding AI across tools enterprises already pay for (Azure, Office 365, Dynamics, GitHub Copilot), pressuring standalone enterprise-AI vendors.

    The 46% revenue plunge and the subsequent 26% workforce cut are indicators of a company that grew too fast on the back of AI hype and is now facing the harsh reality of enterprise budget tightening.

    https://www.financialcontent.com/article/finterra-2026-2-27-the-c3ai-crisis-analyzing-the-46-revenue-plunge-and-the-path-to-2027

Strategy & Moats

  1. [24]Ontology Overview — Palantir Foundry documentationTier 1supportingHigh confidence

    Palantir's core technical idea is the Ontology — an operational layer it describes as a 'digital twin of the organization' combining semantic elements (objects, properties, links) and kinetic elements (actions, functions, security).

    the Ontology serves as a digital twin of the organization, containing both the semantic elements (objects, properties, links) and kinetic elements (actions, functions, dynamic security) needed to enable use cases of all types.

    https://www.palantir.com/docs/foundry/ontology/overview
  2. Palantir's stated go-to-market strategy for AIP is boot camps, which it says deliver real workflows on customer data in days; nearly 300 organizations had used AIP within five months of its 2023 launch.

    We will continue investing meaningfully in boot camps as our go-to-market strategy for AIP. ... nearly 300 distinct organizations have used AIP since our launch just five months ago.

    https://www.constellationr.com/blog-news/insights/palantirs-commercial-business-scales-help-ai-boot-camps
  3. A regulatory-entrenchment moat: Palantir holds FedRAMP High plus DoD Impact Level 5 and 6 authorizations across its full product suite, enabling the most sensitive unclassified government workloads.

    full suite of products, including the AIP, Apollo, Foundry, Gotham, FedStart, and Mission Manager ... FedRAMP Moderate and Department of Defense Impact Level 5 and 6 (DoD IL5 and IL6) ... the most sensitive unclassified workloads

    https://www.investing.com/news/stock-market-news/palantir-shares-surge-on-fedramp-high-authorization-93CH-3751722
  4. Andreessen Horowitz argues that, unlike consulting firms, Palantir's forward-deployed services are a means to drive product adoption rather than the primary revenue stream.

    Unlike consulting firms, services are a means to drive product adoption, not the primary revenue stream.

    https://a16z.com/the-palantirization-of-everything/
  5. Disconfirming view: a16z warns the forward-deployed-engineer model can fail to scale — copying only the embedded-engineer part yields thousands of bespoke deployments that are impossible to maintain, and an expensive services business carrying a software multiple.

    most companies copying the aesthetic are setting themselves up to become expensive services businesses with a software valuation multiple ... if you only copy the embedded-engineer part, you end up with thousands of bespoke deployments that are impossible to maintain

    https://a16z.com/the-palantirization-of-everything/
  6. Disconfirming view: short-seller Citron Research argues Palantir is essentially 'locked-in consulting wrapped in software' that lacks the self-reinforcing flywheel of an OpenAI.

    OpenAI is a self-reinforcing growth engine, while Palantir is essentially locked-in consulting wrapped in software.

    https://fortune.com/2025/08/21/why-palantir-stock-falling-openai-bubble-citron-shortseller-andrew-left/
  7. Disconfirming view: short-seller Jim Chanos criticized Palantir management for ridiculing skeptics about the stock price while executives kept selling material amounts of shares.

    Palantir management continues to ridicule the skeptics on their stock price…while they keep selling material amounts of shares

    https://www.benzinga.com/markets/equities/25/03/44369830/jim-chanos-slams-palantir-management-for-continuing-to-ridicule-skeptics-while-executives-sell-328-million-worth-of-stock

Peer Comparison & Benchmarking

  1. Palantir's Q1 2026 revenue was $1.633 billion, up 85% year-over-year, with GAAP net income of $871 million (a 53% net margin).

    first-quarter revenue coming in at $1.633 billion, up 85% from the same quarter a year ago ... GAAP net income was $871 million, a 53% GAAP net margin

    https://www.tikr.com/blog/palantir-q1-2026-earnings-u-s-revenue-crosses-100-growth-for-the-first-time
  2. Peer (data/AI platform): Snowflake's FY2026 product revenue was $4.472 billion, up 29% YoY, with non-GAAP product gross margin around 75% and non-GAAP operating margin of 11% — but it remains GAAP-unprofitable.

    full-year product revenue reaching $4.472 billion, up 29 ... Non-GAAP product gross margin held steady at 75% in Q4 ... non-GAAP operating margin expanded to 11%

    https://www.investing.com/news/company-news/snowflake-q4-fy2026-slides-30-revenue-growth-margin-expansion-ahead-93CH-4526231
  3. Peer (data/AI platform, private): Databricks surpassed a $5.4 billion revenue run-rate growing >65% YoY in Q4, with a $1.4B AI run-rate and positive free cash flow, at a $134 billion valuation.

    crossed a $5.4 billion revenue run-rate ... >65% year-over-year growth during its Q4 ... Crossing $1.4 billion revenue run-rate for its AI products ... at a $134 billion valuation

    https://www.databricks.com/company/newsroom/press-releases/databricks-grows-65-yoy-surpasses-5-4-billion-revenue-run-rate
  4. Peer (AI-software cautionary case): C3.ai's FY2026 total revenue FELL to $250.3 million (from $389.1M) with a $470.4 million GAAP net loss and a 31% GAAP gross margin.

    Total Revenue was $250.3 million ... GAAP net loss per share was $(3.35) ... GAAP gross profit was $77.4 million, representing 31% gross margin

    https://www.stocktitan.net/sec-filings/AI/8-k-c3-ai-inc-reports-material-event-b190625cd7e2.html
  5. Peer (enterprise SaaS): ServiceNow's FY2025 total revenue was $13.278 billion, up ~21% YoY, with GAAP net income of $1.748 billion.

    Revenue ... $13,278 [million] ... Revenue Growth (YoY) 20.89% ... Net Income ... $1,748 [million]

    https://stockanalysis.com/stocks/now/financials/
  6. Peer (defense/government consulting): Booz Allen Hamilton's FY2026 revenue (ended March 2026) was about $11.2 billion, down ~6% YoY from $11.98 billion — a contrast with Palantir's government-segment growth.

    Revenue ... $11,217 [million] ... YoY Growth -6.37% ... [prior year] $11,980 [million]

    https://stockanalysis.com/stocks/bah/financials/
  7. Palantir's valuation dwarfs faster-cheaper peers: it traded at roughly 42× forward sales versus ServiceNow's ~6× as of late May 2026.

    trades at a much frothier forward P/S of 42 ... ServiceNow ... trades at a forward price-to-sales ratio of 6

    https://www.fool.com/investing/2026/05/29/palantir-trades-42-times-forward-sales-servicenow/
  8. Peer (observability SaaS): Datadog's FY2025 revenue was $3.43 billion, up 28% YoY, with a roughly 80% GAAP gross margin but a slim GAAP operating loss of $(44) million.

    Revenue was $3.43 billion, an increase of 28% year-over-year ... GAAP operating loss was $(44) million; GAAP operating margin was (1)%

    https://www.stocktitan.net/news/DDOG/datadog-announces-fourth-quarter-and-fiscal-year-2025-financial-buqhnpplsjrr.html

Financials & Growth

  1. Palantir's FY2025 revenue grew 56% year-over-year to $4.475 billion.

    Full year 2025 revenue grew 56% year over year to $4.475 billion.

    https://www.fool.com/earnings/call-transcripts/2026/02/02/palantir-pltr-q4-2025-earnings-call-transcript/
  2. Palantir's Q4 2025 revenue grew 70% YoY to $1.407 billion; FY2025 U.S. commercial revenue grew 109% YoY to $1.465 billion (Q4 U.S. commercial +137% to $507 million).

    The company reported fourth quarter net income of $609 million ... on revenue of $1.407 billion, up 70% from a year ago. ... Palantir's US commercial revenue was $507 million, up 137% from a year ago.

    https://www.constellationr.com/insights/news/palantir-delivers-strong-q4-sees-2026-us-commercial-revenue-surge-ahead
  3. Palantir's FY2025 GAAP net income was $1.625 billion (a 36% margin); GAAP income from operations was $1.414 billion (a 32% margin).

    Full year GAAP net income was $1.625 billion, representing a 36% margin. ... Full year GAAP operating income was $1.414 billion, representing a 32% margin.

    https://www.fool.com/earnings/call-transcripts/2026/02/02/palantir-pltr-q4-2025-earnings-call-transcript/
  4. Palantir generated about $2.13 billion in cash from operations in FY2025 and ended the year with $7.2 billion in cash, equivalents and short-term U.S. Treasuries; its Rule of 40 score was 127% in Q4 (106% for the full year).

    For the full year, we generated $2.13 billion in cash from operations. ... Our combined revenue growth and adjusted operating margin accelerated to 127% in the fourth quarter ... Our full year Rule of 40 score was 106%. ... We ended the quarter with $7.2 billion in cash, cash equivalents, and short-term US treasury securities.

    https://www.fool.com/earnings/call-transcripts/2026/02/02/palantir-pltr-q4-2025-earnings-call-transcript/
  5. Palantir's Q4 2025 net dollar retention was 139% (up 500 bps); customer count grew 34% YoY to 954, U.S. commercial customers grew 49% to 571, and total remaining deal value reached $11.2 billion (up 105% YoY).

    Net dollar retention was 139%, an increase of 500 basis points from last quarter ... Customer count grew 34% year over year ... to 954 customers ... Our US commercial customer count grew to 571 customers ... 49% year over year ... $11.2 billion in total remaining deal value, an increase of 105% year over year

    https://www.fool.com/earnings/call-transcripts/2026/02/02/palantir-pltr-q4-2025-earnings-call-transcript/
  6. [44]Palantir (PLTR) Financials — StockAnalysis.comTier 2neutralHigh confidence

    Palantir's annual revenue trajectory was $1,542M (FY2021, +41%), $1,906M (FY2022, +24%), $2,225M (FY2023, +17%), $2,866M (FY2024, +29%), and $4,475M (FY2025, +56%).

    FY 2021: $1,542M (41.11% growth) ... FY 2022: $1,906M (23.61% growth) ... FY 2023: $2,225M (16.74% growth) ... FY 2024: $2,866M (28.79% growth) ... FY 2025: $4,475M (56.18% growth)

    https://stockanalysis.com/stocks/pltr/financials/
  7. [45]Palantir (PLTR) Financials — StockAnalysis.comTier 2supportingHigh confidence

    Palantir reached GAAP profitability in FY2023: net income went from -$520M (2021) and -$374M (2022) to +$210M (2023), then $462M (2024) and $1,625M (2025).

    FY 2021: -$520.38M (loss) ... FY 2022: -$373.71M (loss) ... FY 2023: $209.83M (turned positive) ... FY 2024: $462.19M ... FY 2025: $1,625M

    https://stockanalysis.com/stocks/pltr/financials/
  8. [46]Palantir (PLTR) Financials — StockAnalysis.comTier 2supportingHigh confidence

    Palantir's FY2025 gross margin was about 82% and free cash flow was roughly $2,101 million.

    Gross Margin FY 2025: 82.37% ... Free Cash Flow FY 2025: $2,101M

    https://stockanalysis.com/stocks/pltr/financials/
  9. [47]Palantir (PLTR) Financials — StockAnalysis.comTier 2neutralHigh confidence

    Palantir's FY2024 revenue was $2.866 billion, up ~29% year-over-year.

    FY 2024: $2,866M (28.79% growth)

    https://stockanalysis.com/stocks/pltr/financials/
  10. Palantir's Q1 2026 revenue grew 85% YoY to $1.633 billion; U.S. revenue grew 104% to $1.282 billion (the first time U.S. revenue exceeded 100% growth), with U.S. commercial up 133% and U.S. government up 84%.

    Revenue grew 85% year-over-year and 16% quarter-over-quarter to $1.633 billion ... U.S. revenue grew 104% year-over-year and 19% quarter-over-quarter to $1.282 billion ... U.S. commercial revenue grew 133% year-over-year ... U.S. government revenue grew 84% year-over-year

    https://www.stocktitan.net/sec-filings/PLTR/8-k-palantir-technologies-inc-reports-material-event-e4ff3e917756.html
  11. [49]Palantir Q1 2026 Earnings — TIKRTier 2supportingHigh confidence

    Palantir's Q1 2026 GAAP net income was $871 million (53% margin); adjusted FCF was $925 million (57% margin); the Rule of 40 score reached 145% and net dollar retention rose to 150%.

    GAAP net income was $871 million, a 53% GAAP net margin ... Net dollar retention of 150%, up 1,100 basis points sequentially ... Total remaining deal value of $11.8 billion and RPO of $4.5 billion

    https://www.tikr.com/blog/palantir-q1-2026-earnings-u-s-revenue-crosses-100-growth-for-the-first-time
  12. Critical view: entering 2026 Palantir carried a price-to-sales ratio of 100 (called 'unprecedented'); even after a ~34% drawdown it traded at a P/S of 67 and a P/E of 153, versus the S&P 500's P/S of 3.6 and P/E of 31.

    Entering 2026, Palantir had a price-to-sales ratio (P/S) of 100. This is an unprecedented valuation figure ... Palantir now trades at a P/S ratio of 67 and a P/E ratio of 153 ... The S&P 500 trades at a P/S ratio of just 3.6 and a price-to-earnings ratio (P/E) of 31

    https://www.fool.com/investing/2026/05/08/palantir-just-crushed-earnings-so-why-is-the-stock/

Risks, Ethics & Controversies

  1. ICE entered a $30 million contract with Palantir in April 2025 to build 'ImmigrationOS,' upgrading ICE's case-management system to support deportation operations.

    ICE entered a $30 million contract with Palantir in April 2025 to upgrade the ICM with those capabilities.

    https://www.aclu.org/news/privacy-technology/palantir-deportation-roundup
  2. The ACLU argues the cruel character of the deportation campaign 'inevitably carries over' to companies like Palantir that facilitate it.

    the stench of the campaign's cruel, brutal, and lawless character inevitably carries over to any company that, like Palantir, participates in facilitating those constitutional and human rights violations.

    https://www.aclu.org/news/privacy-technology/palantir-deportation-roundup
  3. In May 2025 thirteen former Palantir employees signed an open letter saying the company's founding principles 'have now been violated,' warning its software could facilitate authoritarian uses.

    These principles have now been violated, and are rapidly being dismantled at Palantir Technologies and across Silicon Valley.

    https://www.wgbh.org/news/2025-05-05/former-palantir-workers-condemn-companys-work-with-trump-administration
  4. Palantir's Maven Smart System is designed to autonomously detect, tag and track objects or humans of interest from surveillance imagery — capability critics view as battlefield-AI risk and the company frames as deterrence.

    autonomously detect, tag and track objects or humans of interest from still images or videos captured by surveillance aircraft, satellites and other means.

    https://defensescoop.com/2025/05/23/dod-palantir-maven-smart-system-contract-increase/
  5. [55]Alex Karp's War for the West — The New RepublicTier 2criticalHigh confidence

    Critics characterize CEO Alex Karp's worldview as militaristic; he has invoked Samuel Huntington's claim that the West rose through 'superiority in applying organized violence.'

    The rise of the West was not made possible 'by the superiority of its ideas or values or religion … but rather by its superiority in applying organized violence.'

    https://newrepublic.com/article/191786/alex-karps-war-west-palantir
  6. [56]Alex Karp's War for the West — The New RepublicTier 2supportingHigh confidence

    Karp defends Palantir's military work in deterrence terms, arguing adversaries who threaten Americans should fear severe consequences.

    something really bad is going to happen to you and your friends and your cousins and your bank account and your mistress and whoever was involved.

    https://newrepublic.com/article/191786/alex-karps-war-west-palantir
  7. Norway's Storebrand Asset Management divested roughly $24 million from Palantir, citing concerns that its sales to Israel for use in occupied Palestinian territories risked violating international humanitarian law.

    excluded Palantir Technologies Inc. from our investments due (to) its sales of products and services to Israel for use in occupied Palestinian territories.

    https://www.calcalistech.com/ctechnews/article/hkfpidjgje
  8. Palantir flatly denied New York Times reporting that it was helping build a federal 'master database,' calling the article 'blatantly untrue' and saying it never collects data to unlawfully surveil Americans.

    The recently published article by the New York Times is blatantly untrue. Palantir never collects data to unlawfully surveil Americans

    https://www.snopes.com/news/2025/06/13/citizens-palantir-surveillance-database/
  9. In its official June 2025 rebuttal, Palantir said it is 'not a vendor on any master database project to unify databases across federal agencies' and that it does not own or control customer data.

    Palantir is not a vendor on any master database project to unify databases across federal agencies.

    https://blog.palantir.com/correcting-the-record-responses-to-the-may-30-2025-new-york-times-article-on-palantir-55b60ae107da
  10. The Financial Times reported NHS England had allowed Palantir employees 'unlimited' access to patient data in a data-integration tenant, fuelling UK privacy backlash over the Federated Data Platform.

    NHS England had allowed Palantir employees 'unlimited' access to patient data

    https://www.aljazeera.com/news/2026/5/12/palantir
  11. Palantir UK denies misusing NHS data, saying it acts exclusively as a data processor and never uses patient data for its own purposes.

    in no way uses patient data, or any NHS data, for its own purposes. Palantir acts exclusively as a data processor

    https://www.aljazeera.com/news/2026/5/12/palantir
  12. [62]Palantir (NYSE: PLTR) 2025 10-K — StockTitanTier 1neutralHigh confidence

    Palantir's FY2025 10-K discloses customer concentration: its top three customers together accounted for 16% of revenue in 2025 (17% in 2024).

    Our top three customers together accounted for 16% and 17% of our revenue for the years ended December 31, 2025 and 2024, respectively.

    https://www.stocktitan.net/sec-filings/PLTR/10-k-palantir-technologies-inc-files-annual-report-74b188995295.html
  13. [63]Palantir (NYSE: PLTR) 2025 10-K — StockTitanTier 1neutralHigh confidence

    Palantir's 10-K lists its multi-class structure and Founder Voting Trust as a risk factor that concentrates voting power with founders Cohen, Karp and Thiel and their affiliates.

    the multi-class structure of our common stock, the Founder Voting Trust Agreement, and the Founder Voting Agreement concentrate voting power with certain stockholders

    https://www.stocktitan.net/sec-filings/PLTR/10-k-palantir-technologies-inc-files-annual-report-74b188995295.html
  14. A shareholder suit alleged Palantir's Class F super-voting structure made the founders 'emperor for life,' fixing their voting power at ~49.99% untethered from equity; the founders called the allegations 'a raft of hyperbole,' noting the structure was approved by investors at the 2020 IPO and is permitted under Delaware law.

    this structure made them the company's "emperor for life." ... The defendants contend ... the class F shares were "approved by a majority of Palantir's other investors when the company went public" ... dismissed the allegations as "a raft of hyperbole," comparing their structure to Ford Motor Company's longstanding 40% voting control provision for the Ford family.

    https://www.fa-mag.com/news/peter-thiel--palantir-co-founders-slam--emperor-for-life--claims-63261.html
  15. Insider selling has drawn scrutiny: the insider sell-to-buy ratio over three months sat at 9.3-to-1, with director Peter Thiel selling more than 2 million shares on a single day in March 2026 (at $140.97–$146.80), and Karp and director Stephen Cohen also selling.

    The selling-to-buying ratio over the last three months sits at 9.3 to 1 ... director Peter Thiel offloading more than 2 million shares on a single day in March at prices between $140.97 and $146.80 ... Karp and director Stephen Cohen joined the parade in February

    https://247wallst.com/investing/2026/05/14/palantir-insiders-are-selling-9-shares-for-every-1-they-buy-here-is-the-defense-stock-smart-money-is-buying-instead/
  16. Bearish view: RBC Capital's Rishi Jaluria rates PLTR Underperform with a $90 target (May 2026), arguing multiples 'price in years of flawless execution' and that PLTR screens as the most expensive name in RBC's software coverage.

    RBC Capital ... Underperform ... Price Target: $90 ... Multiples price in "years of flawless execution"

    https://www.benzinga.com/trading-ideas/movers/26/05/52875641/palantir-stock-is-either-a-bargain-or-a-bubble-depending-on-who-you-ask
  17. Bearish view: Jefferies rates PLTR Underperform ($70 target), saying 'multiple downside outweighs fundamental upside'; the stock peaked near 73× calendar-2027 revenue before falling to ~31×, still a premium to peers like Snowflake (~7×).

    We believe multiple downside outweighs fundamental upside ... Late 2025 peak: 73x calendar 2027 revenue ... Current: approximately 31x ... Cloudflare (16x), CrowdStrike (13x), Snowflake (7x)

    https://www.proactiveinvestors.com/companies/news/1086983/palantir-technologies-valuation-concerns-weighing-on-stock-despite-strong-fundamentals-jefferies-says-1086983.html

Forward View

  1. Palantir soared roughly 340% in 2024, making it the top-performing stock in the S&P 500 that year.

    shares of data analytics firm Palantir Technologies ... soared by 340% ... making it the top performer in the S&P 500 index

    https://www.fool.com/investing/2025/01/16/palantir-was-top-performing-stock-in-sp-500-in/
  2. Palantir remained a market leader in 2025: by mid-June 2025 it was the best-performing S&P 500 stock, up more than 89% year-to-date.

    up more than 89% in 2025 ... the stock is now the best-performing S&P 500 stock

    https://finviz.com/news/81652/palantir-defies-bears-leads-sp-500-in-2025
  3. With Q4 2025 results, Palantir issued FY2026 revenue guidance of $7.182–$7.198 billion (≈61% growth), with U.S. commercial revenue guidance of at least 115% growth (in excess of $3.144 billion) — about 15% above the ~$6.22B Street estimate.

    Full-year 2026 revenue guide: $7.182 to $7.198 billion, implying 61% growth. That compares to a Street estimate of $6.22 billion, representing 15% upside to consensus ... The company closed $4.26 billion in total contract value, up 138% year-over-year

    https://247wallst.com/investing/2026/02/03/here-is-why-palantir-is-up-11-after-earnings/
  4. After Q1 2026, Palantir raised FY2026 revenue guidance to $7.650–$7.662 billion (~71% growth) and U.S. commercial revenue guidance to in excess of $3.224 billion (at least 120% growth).

    We are raising our revenue guidance to between $7.650 – $7.662 billion ... We are raising our U.S. commercial revenue guidance to in excess of $3.224 billion, representing a growth rate of at least 120%

    https://www.stocktitan.net/sec-filings/PLTR/8-k-palantir-technologies-inc-reports-material-event-e4ff3e917756.html
  5. As of June 3, 2026 PLTR traded around $160.65 with a 52-week range of $118.93–$207.52, a forward P/E of ~145× and a price-to-sales ratio of ~71.8× — leaving, per the author, 'no margin for error.'

    Current Price: $160.65 ... 52-Week Range: $118.93 to $207.52 ... Forward P/E: 145x ... Price-to-Sales Ratio: 71.83 ... no margin for error

    https://247wallst.com/investing/2026/06/03/prediction-palantir-will-trade-at-this-price-in-two-years/
  6. Palantir's market capitalization was approximately $340.9 billion as of June 3, 2026.

    Palantir has a market cap or net worth of $340.9 billion as of June 3, 2026.

    https://stockanalysis.com/stocks/pltr/market-cap/
  7. Bull view: Wedbush's Dan Ives rates PLTR Outperform with a $230 target, calling it the 'Messi of AI' and expecting it to join the trillion-dollar market-cap club.

    Wedbush - Dan Ives ... Outperform ... Price Target: $230 ... Called Palantir the "Messi of AI" and expects it to join the trillion-dollar market-cap club

    https://www.benzinga.com/trading-ideas/movers/26/05/52875641/palantir-stock-is-either-a-bargain-or-a-bubble-depending-on-who-you-ask
  8. Bull view: Citigroup's Tyler Radke upgraded PLTR to Buy with a $235 target (Jan 2026), and Morgan Stanley's Sanjit Singh raised his target to $205, calling Palantir 'the enterprise AI standard.'

    Tyler Radke (Citigroup) ... Upgraded to Buy ... Target price: $235/share ... Sanjit Singh (Morgan Stanley) ... Raised target to $205/share ... "emerging as the enterprise AI standard...hard to find a better fundamental story"

    https://www.fool.com/investing/2026/01/14/palantir-stock-investors-great-news-wall-street/
  9. Bear view: short-seller Citron Research (Andrew Left) argued in August 2025 that even at OpenAI's 17× sales multiple Palantir would be worth only about $40/share.

    even at a 17x sales multiple, OpenAI has the highest multiple of any scaled SaaS stock in the world, and that number in itself is extreme

    https://fortune.com/2025/08/21/why-palantir-stock-falling-openai-bubble-citron-shortseller-andrew-left/
  10. As of the May 2026 'bargain or bubble' debate, the 29-analyst consensus price target was about $173.76 with a consensus rating of Buy, even as trailing P/E sat near 161× and forward P/E near 98×.

    P/E Ratio (trailing): 161x ... P/E Ratio (forward): 98x ... Average Price Target: $173.76 (29 analysts) ... Consensus Rating: Buy

    https://www.benzinga.com/trading-ideas/movers/26/05/52875641/palantir-stock-is-either-a-bargain-or-a-bubble-depending-on-who-you-ask