Salesforce explores outcome-based AI pricing beyond traditional user licences
Salesforce is rethinking how it charges for AI products because agents and APIs perform work without individual human users, undermining its traditional per-seat licensing model. It is exploring outcome-based pricing, such as charging for customer-service cases resolved, alongside consumption-based credits and bundled contracts.
The company is testing three main approaches: seat licences, Flex Credits based on usage, and fees linked to measurable outcomes. Salesforce says consumption revenue may take three to five years to become significant, while customers face uncertainty over annual agreements, all-you-can-eat deals, Agentic Enterprise License Agreements and Salesforce Commit contracts, which allow spending to shift between products and usage.
- Salesforce is considering charging for AI outcomes rather than user seats.
- Flex Credits and bundles will cover harder-to-measure AI usage.
- Customers face growing uncertainty over future contract costs.
Both sides, in good faith
The strongest fair case each way — we don't pick a winner.
The case for
Outcome-based pricing aligns the interests of vendor and customer more fairly than traditional licensing, as Salesforce is compensated only when its AI agents deliver measurable results rather than for unused capacity. This model encourages genuine product quality and innovation, since revenue depends on demonstrated value. For customers, it eliminates waste of paying for software that sits idle, instead tying costs directly to business impact and allowing them to scale expenses proportionally to actual value received.
The case against
The shift to outcome-based pricing introduces significant uncertainty for enterprises accustomed to predictable annual budgets, making financial planning substantially more difficult and exposing them to potential cost escalation. Vendors retain the power to define what constitutes an "outcome" and how it is measured, creating asymmetric information that could disadvantage customers and lead to unfavourable terms, particularly once organisations have grown dependent on the software. The proliferation of pricing models risks administrative complexity and confusion.
AI Health Mental Health Technology
Read the full article at the source →
Originally published by The Register as “Salesforce wants to charge for AI outcomes, but first it needs to figure out how”.