TL;DR
Salesforce has closed 29,000 deals with Agentforce and reports $800 million in ARR, but its stock is down 30 percent in 2026 amid the SaaSpocalypse selloff. The Williams-Sonoma, UChicago Medicine, and SharkNinja demos turned out to be works in progress rather than live implementations.
Salesforce has a problem that no amount of marketing can solve. The company has built its entire narrative around Agentforce, its AI agent platform, and the numbers look impressive on paper: 29,000 closed deals, $800 million in annual recurring revenue, and a roadmap that promises to replace entire categories of human work. But Wall Street doesn’t believe it, and the gap between what Salesforce shows on stage and what customers actually use is growing wider.
The action tells the story. Salesforce shares fell nearly 21 percent in 2025 and have fallen another 30 percent so far in 2026. The drop follows a broader sell-off in software as a service companiesan event that the market has called SaaSpocalypse. Approximately $285 billion in SaaS market cap evaporated in a single 48-hour period in February. The logic is simple: if one AI agent can do the work of ten employees, why would a company pay for ten positions?
Salesforce has tried to get ahead of that issue by positioning itself as the company that sells agents instead of seats. CEO Marc Benioff has called Agentforce a “Digital labor platform.In earnings calls, the company cites the 29,000 deals and the ARR figure as evidence that companies are buying.
The 💜 of EU technology
The latest rumors from the EU tech scene, a story from our wise founder Boris and some questionable AI art. It’s free, every week, in your inbox. Register now!
The problem is that prominent examples continue to crumble under scrutiny. At Dreamforce, Salesforce demonstrated a Williams-Sonoma AI agent called Olive that was supposed to act as an agent sous chef, helping customers plan meals and find products. In practice, Olive struggled with specific questions and recommendations. The agent’s most advanced capabilities were described using future tense, “soon you will be able to” instead of functions that were active.
A similar pattern appeared at the University of Chicago Medicine. Salesforce featured the hospital system as a flagship implementation of Agentforce for Health. The reality was more modest: UChicago Medicine’s first AI agent was launched in a web chat to handle basic questions like parking directions and clinic availability. More ambitious features, including voice-based patient assistance, were still in development.
SharkNinja, the maker of Shark vacuum cleaners and Ninja kitchen appliances, was another major customer. Salesforce said the company would use Agentforce to optimize customer service. Bloomberg reported a 20 percent reduction in support calls as part of the pitch. But the deployment described was forward-looking and it was hoped that the agents “guide customers through the purchasing process” and “manage returns,“It is not a report on the results already achieved.
This is important because Salesforce is not the only company exaggerate AI capabilities. Apple agreed to pay $250 million in May to settle a class-action lawsuit that alleged it had overstated what Apple Intelligence and a smarter Siri would offer when it launched the iPhone 16. The settlement covered allegations that the company’s marketing went far beyond what the technology could do at launch.
Salesforce’s financial track record adds another layer. Revenue growth has slowed from about 25 percent a few years ago to about 10 percent in fiscal 2026, when the company reported $41.5 billion in total revenue. It’s still a great business and the company delivered a strong fourth quarter with 12 percent growth. But the slowdown is exactly what investors fear when they hear that AI agents will reduce the number of human users who need software licenses.
The company has attempted to address the pricing issue. Agentforce uses a consumption-based model rather than traditional per-seat pricing, and charges by what Salesforce calls “agent work units.” It has consumed almost 20 trillion tokens and converted them into more than 2.4 billion such units. The central bet is whether that model can grow fast enough to offset the structural threat to seat-based revenue.
Smaller customers illustrate both the promise and the cost. The City of Kyle, Texas, deployed Agentforce to run its 311 service, handling more than 12,000 resident requests since March 2025 with nearly 90 percent first-call resolution. Bloomberg reported that the city doubled its spending on Salesforce to $300,000. For a rapidly growing municipality, that may be a reasonable investment. For enterprise customers doing the same calculation at scale, the economics are less clear.
The competitive pressure is real. SAP presented its Autonomous Company with more than 200 AI agents and an Anthropic partnership in Sapphire 2026. ServiceNow, Google and Microsoft are creating agent platforms. The question is no longer whether AI agents will reshape enterprise software, but whether Salesforce can maintain its position as the market reprices around it.
Benioff has responded with characteristic confidence, announcing a new revenue target of $60 billion for fiscal 2030. He has also committed $50 billion in share buybacks, a signal to investors that the company believes its shares are undervalued. The transformation of Slack into an agent platformWith more than 30 new AI capabilities and a required bundle with every new Salesforce account starting this summer, it’s part of that push.
None of this resolves the central tension. Salesforce is asking customers to pay for a future its own demos haven’t yet delivered, while asking investors to trust that consumption-based AI revenue will replace the seat-based model the company built. The 29,000 offers are real. The $800 million in ARR is real. But he agent AI market It rewards results, not ads, and the gap between the two is where Salesforce’s credibility will be tested.
