Score evidence, not presentation. Use this common scale:
| Rating | Meaning |
|---|---|
| 0 | Tested and unusable for the stated requirement |
| 1 | Materially below requirement |
| 2 | Partially meets requirement; significant limitation remains |
| 3 | Meets the pre-agreed minimum acceptance conditions |
| 4 | Exceeds those conditions with demonstrated operational benefit |
| 5 | Materially exceeds them, with repeatable evidence across relevant test cases |
| U | Unknown or not tested; not equivalent to a demonstrated failure |
Weighted points = criterion weight × rating ÷ 5.
Total score = sum of weighted points, out of 100.
For a provisional total, give U no points while retaining the full denominator. Label the result a lower-bound score, show the unknown weights separately, and do not treat it as a final ranking. Never inflate a score by removing untested criteria from the denominator.
Record an evidence label alongside each rating:
- Observed: inspected or reproduced in an authorised club test.
- Documented: supported by the relevant contract, specification or reviewed assurance document.
- Supplier-claimed: asserted but not verified for the club’s scope.
- Unknown: evidence unavailable or not tested.
Use observed evidence for operational performance and the relevant documented evidence for contractual requirements. A supplier’s public coverage page alone cannot earn an observed-coverage score. A successful export alone cannot earn a licence-fit score.
How should reviewers resolve disagreement?
Have two reviewers score independently before discussion. Reconcile differences greater than one rating point by returning to the evidence and acceptance condition; retain both initial ratings and the final rationale.
For numerical measures, define score boundaries before the trial. For judgement-based measures, require an evidence reference and an explanation of why the answer is sufficient to change—or support—the next recruitment action.
What decision bands should the club use?
These are proposed decision rules, not market benchmarks:
| Result | Recommended action |
|---|---|
| Any gate fails | Reject the proposed scope or renegotiate and retest it, regardless of score |
| Any gate is on Hold, or a criterion remains U | Continue diligence; no final approval |
| 80–100, all gates pass | Strong candidate for approval, subject to the budget, loss register and implementation plan |
| 70–79, all gates pass | Negotiate, narrow scope or run a targeted follow-up before committing |
| Below 70 | Do not approve on the present evidence; consider a different scope or the baseline option |
Also require at least 3/5 for decision contribution, coverage and accuracy, plus any other criterion designated mandatory in the brief. Meeting minimum conditions everywhere produces 60/100: this framework deliberately requires more than basic adequacy for a strong purchasing recommendation.