| Pitfall | Better practice |
|---|---|
| Supplier chooses every demonstration player | Use the club’s authorised sample and a separate discovery brief |
| Same source appears in several products | Map source lineage and distinguish evidence from interpretation |
| Only successful tasks are timed | Record failures, corrections and verification effort |
| Overall coverage hides a priority-market failure | Report competition-season and critical-player exceptions |
| A polished report outranks a failed hard gate | Apply gates before weighted preferences |
| A technically successful export is treated as permission | Maintain a separate clause-linked rights register |
| Historical validation uses future information | Require as-of-date inputs or label the limitation |
| Low usage is assumed to mean cancellable spend | Test dependencies and contractual commitments |
| Group ownership is assumed to permit sharing | Name the authorised entities and recipients in the agreement |
| Staff time saved is called a cash saving | Show capacity and avoidable expenditure separately |
| DIY becomes an unbounded data-platform project | Stop at the decision scope; approve any build separately |
The decisive question is not “Which supplier has the most features?” It is “Which approved configuration answers our recruitment questions with sufficient evidence, manageable effort and acceptable cost?” A well-supported decision to keep the current stack is a successful outcome.