Scorecard · No. 05 / 17
How do you calculate the score?
Last updated
Award points for demonstrated capability and executable commitments, not presentation quality.
Use one evidence scale across all criteria
| Rating | Meaning | Evidence standard |
|---|---|---|
| U | Unknown | Requested evidence has not been supplied. Count as zero provisionally, but preserve the U label. |
| 0 | Absent or unacceptable | Evidence confirms that the requirement is missing or contradicted. |
| 1 | Asserted | Sales statement, generic assurance or unsupported case-study headline. |
| 2 | Partially specified | Some useful detail, but a material dependency, definition or acceptance test is missing. |
| 3 | Operationally credible | Complete method or plan, named owner and objective acceptance test; supporting artefacts are inspectable. |
| 4 | Demonstrated | A relevant trial or work sample shows the proposed method functioning, including limitations and exceptions. |
| 5 | Buyer-verified | The buyer or independent reviewer has checked relevant outputs, recalculated results or verified delivery against the acceptance test. |
Weighted points = criterion weight × rating ÷ 5. Add the points across all 12 criteria. Retain one decimal place for calculation, but do not interpret a one-point difference as meaningful precision.
Also report:
- Evidence coverage: sum of weights for criteria not marked U, out of 100. This measures completeness, not truth or confidence.
- Gate status: six separate results.
- Critical-floor status: criteria 2, 3, 4, 5, 6 and 11 must each reach at least 3/5 before pilot approval.
- Unresolved dependencies: owner, required evidence and deadline.
Have the SEO or analytics lead and the budget owner score independently. Reconcile differences of two or more rating points against the evidence. Do not simply average away a methodological disagreement.