Investors who buy Salesforce (CRM +3.20%) today are really betting on whether CEO Marc Benioff's "agentic enterprise" vision sticks, not just whether the next quarter beats expectations.
His conviction isn't just bravado; it's tied to a specific bet that artificial intelligence (AI) will become digital labor within Salesforce rather than a commodity feature that would make its platform irrelevant.
That makes the stock a higher‑risk, higher‑reward hold or accumulation story, not a sleepy blue chip.
Image source: Getty Images.
When Benioff pushes back on fears that AI will kill Salesforce, he's arguing that the company has rebuilt itself around the idea of the agentic enterprise -- a world where AI agents do work and humans provide judgment. On Salesforce's own materials, you can see the evolution: Einstein introduced predictive AI, then generative capabilities layered on top, and now Agentforce is presented as agentic AI that plans and executes multistep tasks grounded in unified customer data.
In its fiscal 2026 press release, Salesforce describes itself as "the operating system for the Agentic Enterprise" and introduces Agentic Work Units, a way of counting tasks completed by AI agents, with 2.4 billion of those units already delivered. To me, that's a quiet but important signal: Salesforce is trying to measure and sell AI as labor, not just as another checkbox feature.
NYSE: CRM
Key Data Points
Some aspects to watch out for
Wall Street's anxiety still has teeth. The company itself, in that same release, lists "uncertainties regarding AI technologies and their integration into our product offerings" among key risks. Salesforce is pushing a transition from simple seat‑based pricing toward consumption metrics tied to AI work. That may be smart over a decade, but in the near term, it muddies visibility for investors who grew comfortable tracking users and subscription revenue.
There's also execution risk: Salesforce is simultaneously ramping Agentforce, expanding Data 360, and weaving AI through Slack and other acquired platforms. Even in its own language, the business acknowledges dependence on external model providers and a shifting regulatory backdrop. A strategy this ambitious leaves less margin for error.
At the same time, the internal signals around AI look more substantial than hype. Salesforce reports Agentforce annual recurring revenue (ARR) of $800 million, up 169% year over year, with 29,000 deals closed and Agentforce accounts in production up nearly 50% quarter over quarter. Those numbers are still small relative to total revenue, but they show customers paying directly for AI agents rather than passively accepting them.
Data 360 ingestion, hitting 112 trillion records in fiscal 2026, more than doubled the prior year, suggesting customers are consolidating more of their customer data and workflows on Salesforce's rails. The more intelligence moves to where work happens, the more valuable that data gravity becomes.
So should investors bet on Benioff's conviction? To me, Salesforce today is best used as a deliberate, sized wager on a specific AI thesis -- that large enterprises will demand trusted, governed AI labor deeply integrated with their customer data. If you buy that premise and can tolerate volatility while this agentic model matures, the current skepticism‑driven pullback looks like a reasonable entry point.
If you think AI will be commoditized and lighter tools will win, this is a stock to approach cautiously or avoid.
