You signed the procurement agreement that treats every post-go-live data mapping change as an out-of-scope variation. The clause now forces your team to fund the real integration work from operating budgets while the vendor walks away with the milestone payment.
Consultants who wrote the original requirements document earn their next engagement by defending the original scope rather than revealing what the production feeds actually contain. Their incentive sits with the next statement of work, not with the records that never matched the demo schema.
Payment schedules tied to user-acceptance testing reward the moment the test environment passes, not the moment the live ERP feed stops rejecting 12% of transaction types. Once the money has cleared, the same team that approved the contract must now explain the new line item for external mapping labour.
The contract language that penalises scope creep also penalises discovery. Any internal engineer who logs an unmapped field after cut-over triggers the variation process that the procurement team fought to keep under 10% of contract value.
Vendors know the pattern. They staff the sales phase with integration architects and replace them post-signature with delivery managers whose KPI is milestone completion, not data fidelity. The gap appears only when the first monthly reconciliation fails.
Your finance team tracks the platform licence cost but treats the hidden mapping hours as general overhead. The result is an operating model that hides the true cost of ownership inside salary lines until the next budget round.
Rewrite the next agreement so that integration testing continues for 90 days after first production use and any schema mismatch found in that window stays inside the original price. Without that clause the same capture repeats on the following platform purchase.