Salesforce moves to add usage-based billing to its Agentforce push, targeting monetization of AI assistants

Salesforce is taking a step aimed at turning enterprise demand for AI assistants into a more directly metered revenue stream. In a move described in a recent market report, the cloud software company said it is adding usage-based billing to its Agentforce platform, a pricing change intended to better match how customers consume AI in day-to-day workflows.

Agentforce is Salesforce’s umbrella for AI agents designed to help businesses act inside tools customers already use, such as sales, service, and marketing systems. Rather than limiting AI to a flat per-seat or bundled subscription feature, usage-based billing ties charges to measurable activity, such as how often an agent runs or how frequently AI functions are invoked.

From a customer standpoint, usage-based billing can look like a closer fit for AI adoption patterns. AI use in enterprises often ramps unevenly, with experimentation and then selective scaling. A metered approach can allow buyers to expand AI usage without committing to the same fixed level of cost in advance, at least in theory.

For Salesforce, the shift is also about aligning financial incentives with the way customers evaluate AI. AI features are frequently tested through pilots and then refined as organizations determine where the assistants deliver measurable productivity gains. Usage-based pricing can give Salesforce room to monetize incremental adoption, rather than relying solely on broader subscription expansion.

The announcement in the market report emphasizes the pricing mechanics, saying the deal adds usage-based billing to the Agentforce platform. Beyond that headline, Salesforce did not provide in the cited report the full commercial terms, including the contract structure, customer rollout timeline, or the exact measurement units the company plans to bill against.

Salesforce’s broader strategy has been to treat AI agents as a central layer across its customer relationship management and customer service products. If agents become embedded in routine enterprise tasks, pricing can become a competitive differentiator, particularly as rival software vendors race to offer agentic capabilities. A metered model also changes how software companies forecast revenue, because it can be more sensitive to real consumption patterns than seat-based licensing.

There is, however, a caveat: the available reporting does not detail how Salesforce plans to reconcile usage-based billing with existing subscription tiers, nor does it clarify what happens when customers already have commitments tied to traditional licensing arrangements. Companies adopting AI frequently care about predictability and budgeting, so the internal logic of Salesforce’s pricing rollout, and any safeguards or caps, would be important to watch.

Looking ahead, the key question for investors and enterprise customers is whether usage-based billing accelerates AI monetization without slowing adoption. Salesforce will likely face scrutiny around how quickly the new approach can be deployed across customer bases, how usage will be measured in practice, and whether customers view metering as a fair reflection of AI value or as a source of billing complexity.