Pay for outcomes

Hourly billing rewards effort. AI should be paid for outcomes.

One of the quieter shifts inside AWS’s delivery model — visible in both the Professional Services transformation and the new Forward Deployed Engineering motion — is commercial, not technical: fixed-price engagements tied to production-deployed business outcomes, instead of time and materials.

It sounds like a procurement detail. It changes everything about incentives.

Under time-and-materials, every efficiency you gain shrinks your own invoice. Automation is a threat. Long timelines are, quietly, fine.

Under outcome pricing, the equation flips. Every agent that accelerates delivery makes the team better, not smaller. Speed becomes profit instead of lost billings. And the customer buys the only thing they ever actually wanted: a working system, not a burn rate.

Matt Garman said it well: “Agents are the way that most enterprises are going to get most of their value out of AI.” If that’s true — and everything I see in the field says it is — then the commercial model has to reward deployed value, or the incentives fight the technology.

Founders selling AI into enterprises: watch this shift closely. The companies that price outcomes will out-position the ones that price seats and hours. Not eventually. Now.