1Q26: Memo Update No. 1 · Results Review

Palantir Technologies (PLTR US)

Memo Update No. 1 · 1Q26 Results Review · 8 May 2026

I. Thesis Refresher

The central question under evaluation is whether Palantir's operational orientation — distinct in architecture from the analytical workloads that define the hyperscaler stack — produces commercial economics that compound durably rather than cyclically. The primer concluded that the conditions for that compounding were present, anchored by the bootcamp-led acquisition motion, the use-case proliferation mechanism within existing accounts, the IDIQ-led forward visibility in government, and the financial profile that distinguishes Palantir from any reasonable peer comparison.

The investment horizon remains 36–48 months. The watch conditions established in the primer specify what would substantiate or undermine that compounding case across the period: net dollar retention sustained above 130%, US Commercial growth above 60%, IDIQ-to-task-order conversion at observable velocity, capital allocation discipline against a now-substantial cash position, stock-based compensation discipline as the company scales, and competitive positioning as the hyperscalers extend their orchestration products toward operational territory. International commercial recovery beyond the UK and key-person or governance events sit alongside as monitoring conditions.

This memo covers the 1Q26 reporting cycle, drawing on the earnings release of 5 May 2026, the accompanying call transcript, supplementary disclosures from AIPCon 9 in March 2026, federal contract activity recorded on USASpending during the period, and competitor disclosures from Microsoft, Amazon, and Snowflake. It is the first quarterly observation against the watch condition framework. The reading is, on balance, materially favourable across the quantitative conditions and unresolved on the qualitative ones — but a single quarter is directional rather than dispositive, and the language used in the sections that follow is calibrated accordingly. No refinements to the analytical frame are warranted from this period. The watch conditions hold as written.

II. 1Q26 in Brief

1Q26 was the highest-growth quarter in Palantir's reporting history, with total revenue of $1.633bn growing 85% year-on-year against a comparable that itself reflected the company's transition into the AI platform era. Adjusted operating margin reached 60.2% and adjusted free cash flow margin 56.6%. Management's composite Rule of 40 score reached 145%, a combination of growth and margin that does not have a contemporary public software comparable at this revenue scale.

The segment cut frames the analytical reading that follows. The US business reached $1.282bn, or 78.5% of total revenue, growing 104% year-on-year. US Commercial grew 133% on a reported basis (143% adjusted for a commercial-to-government program transition that contributed approximately $20–25m of segment reclassification in the quarter). US Government grew 84% to $687m, the strongest quarterly growth rate the segment has produced since the IDIQ-driven forward visibility framework became the dominant feature of its economics. International grew 37% in aggregate to $351m, with International Government at 51% growth and International Commercial at 26%. The latter is both an improvement on the 2% growth recorded across FY25 and the source of an analytical tension addressed in Section IV.

Forward visibility metrics moved in the same direction as the headline numbers. Net dollar retention reached 150%, with the CFO noting that the trailing-twelve-month methodology does not yet fully capture the underlying acceleration. Remaining performance obligations grew 134% year-on-year to $4.5bn. Total customer count reached 1,007 (+31% year-on-year), with US Commercial customer count at 615 (+42%).

Management raised FY26 revenue guidance by $436m at the midpoint, from $7.21–7.23bn to $7.650–7.662bn. The composition of the raise is informative: $374m, or 86%, was attributed to US Commercial (raising the floor on that segment from $2.85bn to $3.224bn), with the remaining ~$62m distributed across the three other segments. Adjusted free cash flow guidance for FY26 was set at $4.2–4.4bn, implying sustained mid-50s margin for the year.

III. Watch Condition Assessment

Cluster 1 — Commercial momentum

Net dollar retention cleared the 130% watch condition threshold by twenty points, with the CFO's note that the trailing-twelve-month methodology lags the underlying run-rate functioning as a useful caveat — it implies the in-period retention dynamic is meaningfully stronger than the headline figure conveys. US Commercial growth at 133% reported (143% adjusted) is more than double the 60% threshold and reflects the bootcamp-led acquisition motion the primer described, now compounding through use-case proliferation within the existing customer base. Customer count growth of 42% in US Commercial alongside the 47 deals of at least $10m closed in the quarter indicates the growth is broadening across the customer base rather than concentrating in a small number of large expansions.

The relevant forward-quality signal is duration-weighted TCV bookings growth at 135%. Customers are committing to longer contracts at a faster rate than they are committing to short-cycle expansions, a pattern more consistent with structural adoption than cyclical buying behaviour. The compounding mechanism the primer described is operating as specified.

Cluster 2 — Government conversion

The forward visibility framework the primer described — IDIQ ceiling as optionality, remaining deal value as contracted commitment, recognised revenue as realisation — operated visibly during the quarter at all three layers. Total remaining deal value reached $11.8bn, up 98% year-on-year, the most direct quantitative evidence of ceiling-to-commitment conversion since the framework became material. US Government segment revenue of $687m, up 84%, is the realisation expression of the same dynamic. Both numbers, read against the $12.3bn IDIQ ceiling carried into the period, indicate that conversion is occurring at a pace materially faster than the watch condition required for affirmation.

Three program-level developments substantiate the conversion mechanic. The Maven Smart System contract ceiling was expanded from approximately $480m to $1.3bn through 2029, covering all five US combatant commands and all military services, with usage of the system having doubled across the four months ending March 2026. TITAN, having delivered its prototype vehicles in 1Q25, transitioned formally into production during the quarter as a program of record with associated production-level task orders. Federal civilian footprint broadened beyond the previously disclosed $300m USDA IDIQ, with a new $86.3m DHS Enforcement Removal Operations BPA call obligated on 30 April and continued execution under existing IDCs at Treasury/IRS, NIH, and the VA. ShipOS, while not driving the headline numbers, was disclosed at AIPCon 9 to be in use across eighteen major suppliers in the naval shipbuilding industrial base.

International Government grew 51% to $172m, anchored by the renewal of the UK MOD enterprise agreement at £240.6m across three years from 1 April 2026. The renewal was a direct award (non-competed) on technical-difficulty-of-switching grounds and explicitly mandates interoperability with NATO and AUKUS partners — embedding Palantir within a multi-country strategic architecture rather than a single-country contract.

The analytically distinctive feature of the period is that ceiling expansion occurred concurrently with conversion. The Maven re-ceiling from $480m to $1.3bn and the new $300m USDA IDIQ replenish optionality faster than recognised revenue depletes it, leaving the forward visibility framework structurally intact even as commitment and realisation accelerate. This is the conversion mechanism behaving as the primer specified rather than the optionality being run down to fund the current quarter's growth.

Cluster 3 — Financial profile

Stock-based compensation reached $202m in absolute terms, up from $155m a year prior, but compressed to 12.4% of revenue from 17.5% — the lowest ratio in the company's reporting history and well below scaled software peer norms. Diluted share count grew 0.7% year-on-year, consistent with the dilution discipline the primer flagged as structurally distinctive. Sales and marketing expense compressed to 19.6% of revenue from 26.7% (a 710 basis point decline), and research and development to 9.9% from 15.3%, the first single-digit print as a public company. Both ratio compressions reflect operating leverage: absolute spend grew (S&M +35%, R&D +19%) but revenue grew faster.

The forward question is whether this leverage extends as the company invests against its product roadmap. Management guided to absolute expenses ramping through the remainder of FY26 to support technical hiring and the AIP Agent Engine SDK build-out, characterised by the CFO as continued investment in the product pipeline. Adjusted free cash flow margin of 56.6% sets the realised expression of the leverage at a level that remains materially distinct from the closest data infrastructure peer. Snowflake, by contrast, reported FY26 free cash flow margin of 25.5% on its 4Q26 print — a meaningful improvement on the 17% historical comparable but a gap that remains roughly twofold.

Cluster 4 — Capital allocation

Cash, cash equivalents, and short-term investments grew to $8.0bn from $7.2bn at year-end FY25, an increase consistent with the quarterly free cash flow generation rate. No new buyback authorisation was announced, and the buyback terminated in January 2026 with $112m executed against a $1.0bn programme remains the most recent capital allocation decision the company has communicated. No M&A activity, dividend signal, or articulated framework for deploying the cash position emerged during the period. The watch condition the primer set — that capital allocation discipline would be tested against the size of the position — was not engaged in 1Q26.

One rhetorical shift is worth noting. Karp characterised the stock as "somewhat undervalued" during the call, language meaningfully different from the philosophical resistance to buybacks that previously dominated his commentary. The remark was not a commitment and should not be read as one, but it is the first directional movement at the rhetorical layer in several quarters. Whether it precedes a corresponding movement at the policy layer is a question for subsequent prints.

Governance and insider activity

Insider transactions during the quarter were concentrated in February and March and predominantly executed through 10b5-1 plans, with the executive sales on 20 February (Karp, Cohen, Sankar, Glazer) accompanied by option exercises that resulted in net increases to ownership for Karp and Cohen. Thiel's 2 March sale of 2.0m shares ($289.7m, ~12.7% of direct holdings) scaled in absolute terms with the share price rather than departing from prior pattern. No leadership changes occurred. One disclosure-quality observation: the company has discontinued aggregate bootcamp count reporting in favour of sales productivity framing — a methodology shift recorded here as a matter of governance, with the substantive read on bootcamp motion health supplied by customer count growth and deal velocity.

IV. Competitive and Ecosystem Developments

The competitive picture moved during the quarter in ways that warrant careful reading rather than declarative judgement. The primer's load-bearing claim — that Palantir's operational orientation is structurally distinct from the analytical workloads that define the hyperscaler stack — was not falsified by competitor disclosures during the period, but it was tested at the rhetorical and product layer in ways the primer did not fully anticipate. The honest framing is that competitive pressure is rising, the architectural distinction remains intact at the workload-deployment layer, and the next two to three quarters will determine whether the convergence we are observing translates into actual workload competition.

The hyperscaler vocabulary shift

Microsoft's 3Q26 commentary, delivered by Satya Nadella in late April, marked a deliberate vocabulary shift toward the operational layer. The framing of an "agentic computing era," the introduction of Agent 365 as a "control plane" governing fleets of agents at enterprise scale, and the positioning of the Foundry Agent Service for "durable stateful agents" capable of long-running workflows are language constructions that previously sat closer to Palantir's positioning than Microsoft's. Adoption metrics support the framing: 35,000 paid Fabric customers (+60% YoY), Copilot Studio active in roughly 90% of the Fortune 500, and tens of millions of agents managed through Agent 365.

Amazon's 1Q26 disclosures move in the same direction. Bedrock Managed Agents, launched in preview in collaboration with OpenAI, is being positioned around "stateful runtimes" as the orchestration primitive for the next generation of AI workloads — the most aggressive operational positioning AWS has produced to date.

What these moves do not yet demonstrate is the extraction of a real operational workload from Palantir. The Fortune 500 Copilot adoption is dominated by productivity use cases adjacent to the M365 footprint rather than the mission-critical operational deployments where Palantir does its differentiated work. AWS Bedrock is positioned as infrastructure for customers and partners to build agents on, rather than as a vertically integrated operational layer for specific workflows. The architectural distinction the primer drew is bending under rhetorical and product pressure but has not yet broken at the workload layer. The moat is under siege rather than breached, and competitors are constructing their own positions in the architectural mould Palantir established rather than dismantling it.

Snowflake's operational pivot

The more surprising read of the period was Snowflake's 4Q26 disclosures, which represent a more concrete extension into operational territory than either hyperscaler produced. The general availability of Snowflake Postgres marks the company's entry into the transactional database market. The $600m acquisition of Observe extends Snowflake into the IT operations market with AI-powered observability built natively on the platform. Snowflake Intelligence, the company's agentic platform, scaled to over 2,500 accounts in 4Q26 (doubling sequentially), with 9,100 accounts now using Snowflake AI products. The Sanofi reference deployment — AI-powered workflows replacing traditional software systems for licence and invoice management — is a direct competitive analogue to the kind of operational replacement Palantir does at scale. Snowflake Postgres and the Observe acquisition are nonetheless infrastructure-and-tooling moves rather than vertically integrated operational deployments. Snowflake is moving up the stack but has not yet reached the deployment layer where Palantir does its differentiated work. Worth monitoring as the most concrete competitive pivot in the period.

The Accenture partnership

The Accenture Palantir Business Group, formally launched in 2Q26 (Accenture's reporting calendar), represents a more substantive ecosystem development than was visible at AIPCon 9 in March. Accenture has positioned the group as a strategic ecosystem investment, supported by two acquisitions during the period — Deco in the UK for defence and public sector, and Ranger Data in the US for cross-industry implementations — and approximately 2,000 Palantir-skilled professionals operating under a Forward Deployed Engineer model that mirrors Palantir's own delivery approach.

The strategic significance is that Accenture provides distribution and implementation capacity in geographies and use cases where Palantir's own FDE bandwidth is constrained. This is the most plausible structural answer to the international commercial gap addressed below — the markets where Karp says Palantir has "no time, no energy" are precisely the markets where Accenture's existing global footprint can carry deployment.

The international commercial tension

International Commercial growth of 26% reads positive against the FY25 base of 2%, but the management posture during the call was the opposite of strategic re-engagement. Karp characterised European commercial markets as a "waste of time machine" where Palantir would not invest sales effort, framing the deprioritisation as deliberate rather than circumstantial. The headline reads better than the underlying signal: the watch condition framed international commercial recovery beyond the UK as the test of whether the gap was structural or cyclical, and the commentary indicates the gap is hardening into strategy rather than softening. The Accenture partnership is the clearest mechanism by which international commercial revenue could reaccelerate without Palantir's own re-engagement, though its contribution will not be visible in the headline numbers for several quarters.

Management tone

Karp's commentary during the call was unusually combative in register — "AI slop," "witchcraft dance," critics who "disagree with the West being strong" — and Sankar's framing of "tokens as the new coal" with AIP as "the steam engine" extends the same posture into the technical commentary. The register is meaningfully different from the conventional CEO-investor mode. Whether the posture compounds into operational consequences (recruitment, customer perception, regulatory exposure) is a question for subsequent quarters; in 1Q26 it is a feature of the disclosure environment rather than a thesis-relevant development.

V. Thesis Standing

The thesis is affirmed. The four quantitative watch conditions in Section III — net dollar retention, US Commercial growth, IDIQ-to-task-order conversion, and SBC discipline alongside operating leverage — were cleared with margin, in several cases by a substantial one. The IDIQ conversion mechanism the primer described as the most thesis-distinctive forward-visibility framework operated visibly during the period, with ceiling expansion occurring concurrently with conversion rather than the optionality being depleted to fund current-period growth. The financial profile that distinguishes Palantir from any reasonable peer comparison remains in place; the gap to the closest data infrastructure peer narrowed but is still measured roughly twofold.

Two conditions are unresolved. Capital allocation against the now-$8.0bn cash position remains the genuinely open question, with the period's only movement at the rhetorical layer rather than the policy layer. Competitive positioning was tested by hyperscaler vocabulary and product moves toward operational territory and by Snowflake's more concrete pivot via Postgres and Observe; the architectural distinction the primer drew remains intact at the workload-deployment layer but is bending under rhetorical and product pressure. International commercial weakness, recorded in the primer as a monitoring condition, is hardening into deliberate strategy rather than softening — a development the Accenture partnership may eventually offset but will not in the next several quarters.

A single quarter is directional rather than dispositive. 1Q26 did not falsify any element of the thesis and substantiated several of its load-bearing claims, but the deceleration scenarios contemplated by the primer require multiple quarters to play out and were never going to be tested by a single print. The honest reading is that the strong tailwind observable in the quarter is consistent with the compounding mechanism the primer described, with judgement on durability deferred to subsequent observations.

VI. What to Watch

The 2Q26 earnings release, expected in early August, will be the first read on whether the trajectory established in 1Q26 holds across multiple quarters rather than reflecting a single high-growth print. The specific items that matter most are NDR direction (whether the 150% print compresses, holds, or extends), US Commercial growth (whether the bootcamp-displaced acquisition motion sustains above 100%), and the disclosure of IDIQ ceiling movement — particularly any quantification of how much of the carried $12.3bn ceiling has converted into RDV during the quarter and how much new ceiling has been added. The 10-Q footnotes are likely to be more informative than the press release on the conversion mechanic.

Capital allocation framework articulation, or its continued absence, is the running test of the unresolved condition in Cluster 4. Karp's "somewhat undervalued" framing in 1Q26 was the first directional movement at the rhetorical layer in several quarters; whether that compounds into a corresponding policy-layer movement (a new buyback authorisation, an articulated framework for the cash position, or any signal of M&A intent) is the question for the period ahead. The absence of any such movement after multiple subsequent prints would be its own observation.

Hyperscaler 2Q disclosures — Microsoft, AWS, and Snowflake — will be the next read on whether the rhetorical and product convergence observed in 1Q26 translates into actual workload competition. The specific evidence that would alter the read is the disclosure of named operational workload migrations away from Palantir, or named customer wins by hyperscaler agentic platforms in deployment categories where Palantir has previously been the incumbent. Absent that evidence, the architectural distinction the primer drew remains intact and the competitive observation stays in monitoring rather than thesis-relevant.

The Accenture Palantir Business Group's contribution to international commercial growth is a longer-dated item but worth flagging for the canon. Visible reacceleration in International Commercial growth alongside continued management deprioritisation of those markets would be the empirical signature of partnership-led distribution working. Conversely, continued single-digit International Commercial growth despite the Accenture expansion would indicate the structural answer to international weakness is not coming through that channel either.

The thesis asks whether Palantir's operational orientation can sustain compounding economics that no analytical-stack peer has produced. 1Q26 supplied the strongest single-quarter substantiation of that question available since the company began reporting in current form. The next four prints will determine whether the substantiation holds.