Salesforce just delivered one of the most impressive demonstrations of AI monetization any enterprise software company has ever produced — 22,000 Agentforce deals in a single quarter, $1.8 billion in combined ARR from its AI platform, and Q4 revenue of $11.18 billion — and the stock dropped 5% in premarket trading. If that doesn’t tell you everything about where we are in the AI cycle, nothing will.
The company’s fiscal Q4 2026 results, reported Tuesday evening, painted a picture of a business successfully transitioning into the “agentic AI” era. Revenue grew 11.7% year-over-year, hitting the top end of guidance. Non-GAAP earnings per share came in at $3.04. And the Agentforce platform — Salesforce’s bet that autonomous AI agents, not chatbots, represent the future of enterprise software — showed staggering adoption. The platform processed 11.14 trillion tokens in Q4 alone, a number that says customers aren’t just testing this stuff. They’re running it at production scale.
So why is the stock getting punished? Simple: fiscal 2027 revenue guidance fell short of analyst expectations. Wall Street wanted a bigger number for next year, and Salesforce didn’t deliver it. This is the same pattern that’s crushed software stocks across the board in 2026 — CRM is down 32.7% year-to-date and 42.1% over the past 12 months. The market is treating every enterprise software company as guilty until proven innocent on the AI disruption question.
The irony is thick. Salesforce is arguably the company best positioned to answer that disruption question. CEO Marc Benioff pivoted toward AI back in 2014 — years before most competitors even had an AI strategy. The company’s 150,000-plus enterprise customers, including 90% of the Fortune 500, have spent decades embedding their business logic and data into the Salesforce ecosystem. That’s a moat that Claude or ChatGPT can’t replicate with a few prompts, no matter what the bears say.
Dan Ives at Wedbush called the selloff overblown ahead of earnings, arguing that Salesforce’s installed base creates unique data advantages that generic AI models simply can’t match. He appears to be right about the fundamentals — but right about the fundamentals and wrong about the stock has been the story of 2026 for every SaaS company not named Palantir.
What makes this moment particularly interesting for traders is the valuation. At roughly 35 times trailing earnings, Salesforce is now trading at a massive discount to peers like Shopify (above 80x) and ServiceNow (above 100x). If you believe agentic AI is real — and 22,000 deals in one quarter suggests it very much is — then CRM at these levels starts looking like one of the more compelling setups in enterprise tech.
The broader question Salesforce’s earnings raise is philosophical: what happens when a company executes beautifully but the market has already decided the narrative? Right now, the answer is that execution doesn’t matter until sentiment turns. When it does turn, the stocks that were actually building real AI revenue — not just talking about it — will be the ones that rip hardest. Salesforce just proved it belongs in that conversation. The market isn’t listening yet.