Section 09 of 12 / Learn and assess
Scorecard
Decide how to compare suppliers on evidence, and which gates override a high score.
- Pages
- 16
- Sources cited
- 8
- Evidence date
- Oct 20, 2026
Scorecard
Score up to three providers on evidence
- Scope completeness and correct uncertainty 20
- Source traceability and reviewer usability 15
- Net estimator effort 15
- Addenda and revision control 10
- Pricing, units and labour controls 10
- Security and data handling 10
- Integration and recovery 7
- Operations, support and change testing 5
- Complete cost and commercial clarity 5
- Exit and portability 3
Weights total 100 of 100.
| Criterion | |||
|---|---|---|---|
| Mandatory gates 6 gates: any failure overrides the total | |||
| Authority to process the dataApproved document use and complete data pathIf unresolved: No confidential uploads | |||
| Accountable releaseNamed technical/commercial approvers and demonstrated permissionsIf unresolved: No live quotation or price update | |||
| Critical scope handlingAgreed test outcomes; no unresolved critical failure in the accepted holdout setIf unresolved: Remediate and retest | |||
| Source and revision integrityMaterial outputs traceable to the correct issueIf unresolved: No acceptance of the affected task | |||
| Recoverable live integration, if includedFailed/duplicate transfer, reconciliation and recovery demonstratedIf unresolved: Remain offline or in sandbox | |||
| Sustainable operationNamed owner, support and usable exit outputIf unresolved: No routine production release | |||
| Weighted criteria Rate each 0 to 5. An unverified capability starts at 0. | |||
| Scope completeness and correct uncertaintyTest: Reference-output comparison and omissions log | |||
| Source traceability and reviewer usabilityTest: Reviewer completes a source-check exercise | |||
| Net estimator effortTest: Paired/comparable trial log with denominator | |||
| Addenda and revision controlTest: Deliberately revised tender case | |||
| Pricing, units and labour controlsTest: Stale-offer, unit and margin-exception cases | |||
| Security and data handlingTest: Data-flow review and deletion/access test | |||
| Integration and recoveryTest: Sandbox fault tests and field-ownership map | |||
| Operations, support and change testingTest: Support runbook and change exercise | |||
| Complete cost and commercial clarityTest: Normalised proposal and acceptance/payment schedule | |||
| Exit and portabilityTest: Export/rebuild exercise and contract clauses | |||
| Results | |||
| Total out of 100Weight × rating ÷ 5, summed | 0.0/ 100 Evidence coverage 0% (weight rated 2 or above) 0 to less than 50: Evidence insufficient: clarify or stop 6 of 6 gates not demonstrated The total is provisional until every gate has evidence. | 0.0/ 100 Evidence coverage 0% (weight rated 2 or above) 0 to less than 50: Evidence insufficient: clarify or stop 6 of 6 gates not demonstrated The total is provisional until every gate has evidence. | 0.0/ 100 Evidence coverage 0% (weight rated 2 or above) 0 to less than 50: Evidence insufficient: clarify or stop 6 of 6 gates not demonstrated The total is provisional until every gate has evidence. |
Rating anchors and gate definitions: rating scale · gates · weights · how the total is calculated · bands.
The weights, rating scale and decision bands are proposed editorial criteria, not observed supplier scores or a validated procurement standard.
Questions in this section
- 01How should we compare suppliers without inventing a winner?Apply the same evidence requirements and workload to every option, including improving your current tools.Read →
- 02What must be decided before any score is calculated?Define the task, document families, approved outputs and operating boundary.Read →
- 03Which gates override a high total?Six mandatory gates that override a high total, with the evidence each needs and what follows if it is unresolved.Read →
- 04How is a criterion rated?A 0–5 rating scale, from no usable evidence to repeated evidence across the agreed variation, applied to every criterion.Read →
- 05What are the proposed weighted criteria?Ten weighted criteria totalling 100, led by scope completeness and source traceability, each with what good and weak evidence look like.Read →
- 06How is the total calculated?Each criterion contributes its weight multiplied by its rating over five, and the ten contributions sum to a total out of 100.Read →
- 07What do the proposed bands mean?Four bands, from insufficient evidence to a candidate for final acceptance review, always subject to the gates.Read →
- 08How do we test scope and source evidence?Test scope and source evidence with a missing schedule, a blank quantity, a duplicate tag and a changed controls requirement.Read →
- 09How do we test pricing and labour controls?Use an authorised or synthetic estimate with a supplier quote expiring before the proposed customer validity period.Read →
- 10How do we test security and operating responsibility?Request a complete application data path rather than only the underlying cloud service’s policy.Read →
- 11How do we prevent commercial bias?Use the same case, workload and clarification questions for each candidate.Read →
- 12What should the completed scoring sheet contain?The fields a completed scoring sheet must record, from candidate and version to the evidence behind each rating.Read →
Reference
Research quality
Self-assessed scores out of 10, not supplier ratings or external validation.
Detailed evidence anchors, mathematical scoring and non-negotiable gates make this usable for procurement. It remains an editorial framework, not a supplier league table or assurance certificate.
Read the assessment →