- Salesforce stock fell nearly 3% on September 17, 2026, despite Agentforce ARR growing over 200% in consecutive quarters, investors are pricing the cost of reaching $63 billion in revenue by FY2030, not just the growth rate.
- A February 2026 survey of 500 finance leaders by DoiT and Sapio Research found 79% of enterprises experienced AI cost overruns in the past year, while The Futurum Group’s own September 2026 survey separately found 46.9% running over their AI budget, making budget accountability the central friction between AI vendors and their customers.
- Forrester’s May 2026 research found fewer than a third of decision-makers could name specific financial outcomes from their AI investments, a gap that pushed enterprise AI spending deferrals into 2027.
Salesforce’s stock dropped nearly 3% the day after Dreamforce 2026 Investor Day, even as Agentforce ARR had just posted back-to-back quarters above 200% growth. The gap between those two data points is the clearest summary of where enterprise AI sits: vendors are posting real numbers, and the market still does not trust the economics.
Big Targets, Cautious Market
At Dreamforce 2026, Salesforce presented new offerings including AIforce and Koa, its initial CRM reasoning model, and reiterated its target of over $63 billion in revenue by fiscal year 2030, a goal management ties explicitly to its “Agentic Enterprise” strategy. The underlying growth numbers are not modest. Agentforce ARR grew 205% in Q1 FY27 and over 240% in Q2, contributing to Q2 revenue of $11.3 billion, an 11% year-over-year increase. Constant currency remaining performance obligation growth came in at 14% year-over-year. Those figures prompted a wave of analyst price target increases. Salesforce shares still fell nearly 3% in midday trading on September 17, 2026.
Where the ROI Numbers Get Uncomfortable
The cool market reception reflects a wider problem with AI budget accountability. According to a February 2026 survey of 500 finance leaders by DoiT and Sapio Research, 79% of enterprises experienced AI cost overruns in the past year. The Futurum Group’s own September 2026 survey put the figure at 46.9% running over budget — either way, most AI spending is running above plan, not producing documented returns against it. Enterprise AI deployments frequently struggle to connect activity metrics to operating outcomes, and that disconnect is now showing up in procurement decisions.
Forrester’s May 2026 research found enterprises were pushing a significant portion of planned AI spending to 2027, largely because fewer than a third of decision-makers could identify specific financial outcomes from their AI investments. The accountability gap has real consequences at the extreme end: one company spent $500 million in a single month on AI services without adequate financial controls. That case is an outlier, but it illustrates the direction the gap points when left unaddressed.
The Pace Debate, Live on Stage
Dreamforce itself became a public forum for disagreement about how fast frontier AI should move. Anthropic’s Dario Amodei advocated for a slower pace of development. Nvidia’s Jensen Huang argued for speed. Marc Benioff acknowledged the uneven adoption picture, noting that employees often use AI in their personal lives while their workplaces lag behind. He said models alone cannot run the enterprise, pointing to AIforce as the connective layer between probabilistic AI outputs and deterministic corporate data systems.
Salesforce’s Q2 FY27 result was one of its stronger growth quarters in recent years. The stock’s reaction to Dreamforce suggests investors are weighing the cost of reaching $63 billion against a backdrop where, by Futurum’s count, most enterprises are already over budget on AI and struggling to show what they got for it.
