Salesforce's $63B FY30 Target Hides a 50% ARR Test for Agentforce — and $300M a Year Already Going Out the Door to Anthropic

The $63B FY30 number did not appear from nowhere. Salesforce first set a $60B+ target at its October 2025 investor day, framing it as a 10%+ organic FY26–FY30 CAGR. Five months later, on February 25, 2026, alongside the Q4 FY26 print, management pushed the target to $63B and pointed to one specific reason: the November 18, 2025 close of the $8B Informatica acquisition. The Delta between the two targets — about $3B of incremental FY30 revenue — is essentially the inorganic layer.

On the organic line, the slope has barely moved. From the $41.525B FY26 base, $60B implies ~9.7% CAGR and $63B implies ~11.0% CAGR — a 130-basis-point step-up, of which the entire uplift is Informatica. The organic growth the company is asking investors to underwrite is essentially what Salesforce already delivered in FY26 (10% reported, 9% organic).

Strip out the core subscription business — running at ~12% Y/Y in Q2 FY27 — and Salesforce needs Agentforce + Data 360 + Informatica to contribute the entire delta between FY26 and FY30. The math is unforgiving. To deliver $63B against a $46.4B FY27 guide, the company must add ~$16.6B in three years. Subscription at ~11% handles ~$15B of that; the remaining ~$1.6B-per-year burden falls on Agentforce, Data 360 and Informatica.

Agentforce + Data 360 ARR exited Q2 FY27 at $3.9B, up >210% Y/Y. To plausibly contribute another ~$5B of annualized revenue by FY30, that ARR pool needs to reach ~$15–18B — i.e., a roughly four-fold scale-up, or about 50% CAGR. That is a far cry from the current 200%+, but it is also far from impossible: the FY26 → FY27 trajectory already validates a step-down from triple-digit to roughly 100% growth, and Microsoft Azure just printed 43% growth off a $100B annualized base.

The Informatica acquisition closes one part of the gap. The $8B deal that closed in November 2025 contributed $440M to Q2 FY27 subscription revenue — about 4% of the quarter — and management's FY27 guide includes 'slightly above 4 points' of Informatica contribution on the subscription line. At a similar contribution rate annualized, Informatica is plausibly a $1.5–2B revenue line by FY30, which absorbs roughly half of the ~$3B target uplift.

Marc Benioff told the All-In podcast in May 2026 that Salesforce is on track to spend ~$300M on Anthropic tokens in 2026, mostly for internal AI-coding work inside Slack and engineering. That figure is striking for two reasons. First, it is roughly 20% of the Agentforce ARR ($1.5B) Salesforce booked against external customers in the same window — a glimpse of how much of the AI-CRM revenue line recycles back to the model provider. Second, it is already large enough to be a line item: at FY30 ARR scale of $15B+ for Agentforce + Data 360, even a comparable 20% passthrough implies $3B+ a year of inference cost.

There is an offsetting financial asset. Bloomberg reported in June 2026 that Salesforce's equity stake in Anthropic had marked to roughly $5B. Q2 FY27 reported a $2.53-per-share adjusted gain from strategic investments — almost entirely the Anthropic mark — which contributed meaningfully to the quarter's adjusted EPS beat of $4.30 vs. $3.27 consensus. The setup is unusual: Salesforce is both a $300M-a-year customer of Anthropic and a multi-billion-dollar equity holder, which both de-risks the supply relationship and creates accounting noise that can flatter reported growth.

Anthropic is private, so the supply-side moat does not appear in this article as a linkable security. What does appear is the competitive math: every additional Agentforce dollar of ARR carries an embedded inference cost that flows to a third-party model provider, while every dollar of Microsoft Copilot revenue is largely captured in-house on Azure.

The $63B FY30 ask is also a competitive ask, because every other enterprise-software platform is moving on the same AI-agent prize. Microsoft reported Q4 FY26 revenue of $90B (+18% Y/Y) with Azure and other cloud services up 43% — the Azure run rate crossed $100B annualized for the first time — and Copilot Studio is embedded directly into the existing Microsoft 365 enterprise customer base. Copilot is not just a CRM play; it is the OS-layer agent surface that touches the same procurement budgets Salesforce is chasing.

ServiceNow is the closer comparable. Q2 2026 (calendar) subscription revenue grew 24.5% Y/Y to $3.877B, total revenue hit $3.987B, and ServiceNow AI crossed $1B in annual contract value during the quarter — with management explicitly guiding to $1.5B by year-end 2026. That is roughly two-thirds of Salesforce's Agentforce ARR pace, achieved on a base one-third the size and at a 24%-growth corporate trajectory that is more than double Salesforce's subscription growth. ServiceNow's go-to-market in workflow and IT is narrower than CRM, but the agent attach rate is at parity.

Oracle is the wildcard on the data side. Q4 FY26 cloud revenue rose 47% to $9.9B, with Cloud Infrastructure (IaaS) up 93% and Fusion cloud applications up 10%; the company's remaining performance obligations ballooned to $638B. Oracle is selling both the AI training infrastructure and the agent runtime that competes with Salesforce's Informatica-fed Data 360.

Below the revenue line, Salesforce is engineering EPS growth through capital returns rather than operating leverage. The board authorized a $50B aggregate repurchase program in February 2026, of which $25B was launched as an accelerated share repurchase (ASR) in March 2026 — funded by $25B of newly issued senior notes across eight tranches. Final ASR settlement is expected in October 2026 and is set to retire at least 14% of shares outstanding.

That arithmetic matters because diluted share count has already moved. Per the FY26 income statement, diluted weighted-average shares were 956M; by Q2 FY27 the count had compressed to 821M, a 14% reduction in twelve months. With Salesforce guiding to ~$15B in current-year free cash flow and committing to return 190% of FY27 free cash flow, the buyback cadence is unlikely to slow even as growth decelerates — which means FY30 EPS growth can outrun FY30 revenue growth by 4–5 points without any underlying acceleration.

Operating margin, meanwhile, is broadly flat. FY26 closed at 34.1% non-GAAP; Q2 FY27 printed the same 34.1%; the FY27 guide is 34.3%. Management has not publicly restated a specific FY30 operating-margin number since the original October 2025 'Profitable Growth Framework' — but Investor's Business Daily and several sell-side models extrapolate to ~40% by FY30. Achieving 600 bps of margin expansion over four years on top of a $20B revenue add would itself be a stretch goal.