What: A practical framework for measuring the true cost and value of enterprise AI now that consumption pricing has changed the economics. I introduce a four-layer total-cost-of-ownership model — licenses, readiness, management, and consumption — and the shift from measuring usage to measuring value.
Why now: As organizations move from AI experiments to production, "people are using it" no longer answers the question leaders actually face: is it worth it? Consumption-priced agents stack variable costs on top of licenses, and most success metrics track only the cheapest, most visible layer. This is the gap between AI ambition and proven value.
How: A real-world walkthrough mapping an enterprise AI scenario across all four cost layers, plus the Knowing→Doing→Value→Spend chain — evidence-based, vendor-agnostic, and grounded in over a decade modeling enterprise technology cost and value at Microsoft.
Takeaways: Attendees leave able to retire broken usage metrics, apply a four-layer TCO model to any AI investment, and build a value narrative a CFO will actually act on.