What a selection scorecard should do
The scorecard should apply the same criteria, scales and weights to every eligible supplier. It helps reviewers compare evidence consistently and understand which factors drive the result.
The scorecard should not decide eligibility. A failed mandatory requirement remains a gate, not a low score.
Design each criterion before scoring begins
For every criterion, define the business meaning, preferred direction, evidence source, scoring scale, weight and owner. A five-point scale is not useful if reviewers interpret “3” differently. Where possible, anchor scores to observable evidence.
| Score | Meaning | Evidence example |
|---|---|---|
| 1 | Materially below requirement | Documented gap with no acceptable mitigation |
| 3 | Meets requirement | Current evidence supports the stated requirement |
| 5 | Materially exceeds requirement | Evidence shows a relevant advantage beyond the baseline |
Use evidence, not familiarity
A known supplier can receive an inflated score when reviewers substitute past impressions for current evidence. Record the source and date of the evidence. Treat missing data according to an approved rule; do not silently assume the average score.
A high score can still be an unacceptable recommendation
A scorecard can obscure critical failures if every issue is converted into points. Keep compliance failures, sanctions concerns, missing approvals and other non-negotiable conditions visible. Similarly, a weighted score should not hide a large residual risk that management would not accept.
Selection scorecard versus supplier performance scorecard
The selection scorecard compares potential suppliers before award. The performance scorecard monitors an active supplier after award. Selection criteria may include quoted cost, capacity and implementation readiness; performance measures may include actual delivery, defects, service and corrective-action closure.
Five review questions
- Do all reviewers understand the scale in the same way?
- Are mandatory requirements separated from weighted preferences?
- Do weights total 100% and reflect the category?
- Can every score be traced to evidence?
- Does the recommendation still show residual risk and award conditions?