• 3 min read
Salesforce says Agentforce now generates $1.2B in annual recurring revenue, but a partner survey finds subdued demand and weak customer readiness.
Image: TechRadar Salesforce says its Agentforce platform has reached roughly $1.2 billion in annual recurring revenue, but customer adoption has not matched that figure. A TD Cowen survey found subdued demand among partners, while customers remain concerned about data readiness and the maturity of agentic AI deployments.
Annual recurring revenue can include early pilots and trial-driven spending, while the survey suggests many customers have not committed to broad production rollouts. Partners may be helping companies experiment with Agentforce without seeing the sustained deployment activity or commercial expansion that would signal durable demand.
Only one-third of surveyed partners said there was strong current interest in Agentforce. Another 56% expected interest to develop later but said customers need time for their initiatives to mature. 11% reported little immediate interest.
Recommended reading
GitHub outage hits APIs, Copilot and Actions
Sergey Kuznetsov • • 4 min read
The same survey found that the share of Salesforce partners meeting or exceeding their commercial targets fell to 33% this quarter, down from 43% last quarter. TD Cowen characterized adoption as “subdued,” suggesting enterprise customers are not yet moving rapidly from AI demonstrations to scaled agent deployments.
Data quality and agent maturity remain barriers #
The survey describes operational obstacles rather than purely financial ones. Customers reportedly question whether their data is sufficiently prepared for agents and whether Agentforce is mature enough for larger deployments. An agent that acts across customer records and business workflows depends on reliable, accessible data; experiments can proceed with limited scope, but broader automation raises the cost of incomplete or poorly structured information.
That helps explain why Salesforce can report revenue while partners describe weak momentum. A company can pay for access, consulting or a limited proof of concept before deciding that an agent is reliable enough to handle more consequential work. The reporting does not establish how much of Agentforce’s $1.2 billion figure comes from pilots, production workloads or expansion within existing accounts, so the revenue number cannot by itself establish large-scale customer adoption.
The timing also puts pressure on Salesforce’s claim that agentic AI is becoming its next major growth engine. Agentforce was introduced roughly two years before these findings, yet customers are still described as evaluating the product’s maturity and working through basic data-readiness issues. The findings do not show that the platform has failed. They do show that the conversion from interest to repeatable enterprise use is taking longer than the strongest AI-growth narrative implies.
A broader industry pattern may be contributing to the hesitation. Gartner predicted in 2025 that more than two in five agentic AI projects would be canceled by the end of 2027, citing problems such as unclear business value. That forecast does not measure Agentforce specifically, but it puts the survey in context: companies are still sorting experiments from deployments that justify ongoing spending.
Salesforce still sees AI as its main growth opportunity #
Salesforce continues to present its transition toward an AI-first company in optimistic terms. After what the company described as an “outstanding” quarter, CEO Marc Benioff positioned agentic AI as a major source of future customer growth.
“The biggest growth opportunity for [the company’s] customers.”
The contrast is straightforward: Salesforce reports a $1.2 billion Agentforce run rate, while only one-third of surveyed partners report strong current interest and the proportion hitting commercial targets has dropped by 10 percentage points from the previous quarter. The unresolved question is whether those pilots mature into production deployments before customers conclude that the data work, product limitations and uncertain business value outweigh the promise of autonomous software.
Marcus Vance Enterprise Editor
Marcus follows the money. He covers enterprise software, cloud architecture, and the tectonic shifts in Big Tech strategy. He translates dense earnings calls and complex M&A activity into actionable insights about where the industry is actually heading. If a tech giant makes a silent pivot, Marcus is usually the first to notice.