Cost-migration test
A check that traces every apparent AI saving across budget owners to see whether the cost was removed or merely moved into review, exceptions, recovery, governance or adoption.
When one part of an AI workflow gets cheaper, the useful question is: where did the cost go? The cost-migration test asks it systematically. For any change — a cheaper model, a new routing rule, a larger automation scope — it follows the money across teams and budget lines to see whether the cost disappeared or just changed address.
A cheaper model can lower the inference bill while raising human review time, retries, escalations, failure handling or governance effort. If those costs land in a different budget than the saving, nobody sees the trade-off and the local win gets credited as a real one.
The rule of thumb
Do not credit a local saving before checking the costs it displaced.
How to run it
Record before, after, delta and owner for each cost dimension, and keep the result in a cost-migration ledger. Compare the totals against the cost per accepted outcome rather than the cost of a single call. This also makes total cost of ownership attributable before anyone tries to optimise it.
Read more in A cheaper AI model can move the cost instead of removing it.
Related terms
Cost-migration ledger
A small record attached to a meaningful AI change that lists, for each cost dimension, the before, after, delta and owner, so shifted costs stay visible across budgets.
Total cost of ownership
The full cost of running an AI system across its life: model and inference spend plus integration, human review, exceptions, recovery, governance and maintenance — not just the price per call.
Cost per accepted outcome
The total cost of getting one AI-assisted result that is actually accepted and used, counting model spend, retries, human review, exceptions and recovery — a truer measure than cost per call.