Salesforce is having the worst year of the megacap software cohort — down 31.9% year to date, down 36% over twelve months — while its AI agent platform grows 169%. That contradiction is the whole stock right now, and the next earnings print is the referee.
| Metric | Value |
|---|---|
| YTD performance | -31.9% |
| 12-month | -36.0% |
| Agentforce ARR | $800M, +169% YoY |
| Cumulative deals | 29,000+ (+50% QoQ) |
| Agentic work units | 2.4B (+57% QoQ) |
| Q1 FY27 consensus | $3.12 EPS / $11.05B revenue |
| FY27 guide | $45.8–46.2B (+10–11%) |
Why it moved
The bear case crushing CRM all year is existential: if AI agents do the work, why pay for seats? Salesforce's answer is to sell the agents itself — and Agentforce at $800 million ARR growing 169% is a genuinely fast enterprise ramp. The market's problem is scale: $800 million against a $46 billion revenue base rounds to nothing yet. So the stock trades on a question the income statement cannot answer for several more quarters: is Agentforce a new growth engine, or a margin defense against its own disruption?
What it means for you
Watch deal count, production accounts and ARR trajectory — not the headline beat. A 34.3% non-GAAP margin and 10-11% organic growth means the base business is a cash machine trading at a discounted multiple; any believable evidence that agents add to rather than cannibalize seat revenue re-rates the stock quickly.
Bottom line: CRM down 32% is where contrarian setups usually live — but I want the Agentforce momentum confirmed one more quarter before calling the bottom. The downside from broken AI narratives is always bigger than the consensus models.