Salesforce just reported a quarter that should have been a mic drop. Revenue grew 10% to $11.2 billion. Operating margins hit 34.2%. Their AI agent platform — Agentforce — is on a $800 million annual run rate, up 169% year over year. And the company authorized a $50 billion stock buyback, one of the largest in software history. The stock initially fell 5% after earnings, then ripped 5% higher the next day. Welcome to the most confused market in a decade.
The Agentforce numbers are genuinely impressive. Total AI and Data Cloud annual recurring revenue hit $2.9 billion, tripling year over year. Billings accelerated to 18% growth — outpacing revenue — which typically signals a coming acceleration in reported sales. Management is so confident they’ve tightened the timeline for top-line acceleration to the second half of fiscal 2027, with full-year revenue guided to $45.8-$46.2 billion.
So why the schizophrenic market reaction? Because there’s a philosophical war happening in software right now, and Salesforce is ground zero. The bull case: AI agents make CRM software more valuable, not less. Every company needs to manage customer relationships, and AI makes those tools smarter, faster, and stickier. Morningstar still rates it a wide-moat business with a $300 fair value — well above the current price.
The bear case: AI agents could eventually replace the humans who use CRM software. If you don’t need as many salespeople, do you need as many Salesforce seats? That existential fear hammered the entire software sector in 2025, and it hasn’t let up. Block’s 40% workforce cut this week — with Jack Dorsey explicitly citing AI — only added fuel to the fire.
Here’s what the numbers actually show: the seat-based model isn’t dying — it’s evolving. Salesforce’s AI revenue is additive, not cannibalistic. Companies are buying Agentforce on top of their existing CRM seats, not instead of them. Free cash flow margins exceeded 31% last year, and the company’s sitting on $9.6 billion in cash. The $50 billion buyback is management putting their money where their mouth is.
Morningstar’s bear case math is telling: even with zero growth beyond year 10, the stock is worth $240 — not far from where it trades today. Meaning the market is pricing in a near-worst-case scenario for one of the most entrenched enterprise software companies on the planet. For patient investors who believe enterprise software survives the AI transition, that’s either a massive opportunity or a value trap. The next two quarters should tell us which.
