Last reviewed June 2026
A buyer-side playbook for Salesforce's consumption-priced AI: how Agentforce charges per conversation and per action, how Data Cloud burns credits, how Einstein is bundled into the platform, and how to cap and govern it all before you sign.
Salesforce has moved its AI stack, Agentforce, Data Cloud, and Einstein, onto consumption pricing, where the meter runs on conversations, actions, and credits rather than on named users. The bill is no longer a predictable per-seat line; it is a variable cost that scales with usage you do not fully control.
This playbook is written for the buyers who sign and govern that commitment, not for the channel that sells it. Every section ends with the buyer move and the contractual guardrail that keeps a consumption model from running away.
Procurement and vendor-management leads sizing a Salesforce AI commitment.
CRM and platform owners deploying Agentforce or Data Cloud.
Finance teams forecasting consumption-based AI cost.
Sourcing teams preparing a Salesforce renewal with AI in scope.
Across more than 500 enterprise engagements, buyers we advise have negotiated over $2.4 billion in software contracts, with average savings of 38 percent and average audit claim reductions of 72 percent.Atonement Licensing engagement record
Related resources: read the full playbook on the Salesforce AI Pricing Playbook page, then see our Salesforce Negotiation Services and the Salesforce Licensing Guide.
Instant access to the full playbook. No sales calls.
Weekly Oracle, Microsoft, SAP, and cloud licensing intelligence for enterprise buyers.
Our advisors represent buyers directly. Confidential assessment within one business day.