Rate suitability and evidence separately. A well-documented bad term is still bad. An attractive but unverified promise is still unverified.

What does each 0–5 rating mean?

Rating Meaning
U — unknown Material information is missing or contradictory. Use zero points provisionally, but label it U rather than alleging failure.
0 — unacceptable Evidence confirms that the proposal cannot meet the relevant requirement.
1 — weak A significant deficiency is acknowledged, with no adequate remedy.
2 — partial Some requirements are addressed, but important elements remain incomplete or unsupported.
3 — adequate The buyer-specific arrangement meets the table’s acceptable standard in writing.
4 — strong It exceeds the adequate standard and material supporting evidence has been checked.
5 — demonstrated It meets the strongest anchor, with relevant completed-work or operational evidence corroborating the promise.

Use whole-number ratings. Give 2 or 4 only when the evidence genuinely lies between the adjacent anchors; do not use decimals to manufacture precision.

What evidence limits apply?

For each criterion, record both the source origin—supplier-stated, buyer-provided, independently checked or unresolved—and the following evidence level:

Level Evidence available Maximum rating
E0 No usable evidence or an unresolved material contradiction U
E1 General website copy, sales assurances, generic reviews or promotional photographs 2
E2 Buyer-specific written specification, appointment plan, quotation or remedy commitment 3
E3 E2 plus relevant corroboration: inspected work, attributable production records, a confirmed partner arrangement or a checked customer case 5

Apply the level to the weakest material element of that criterion, not its most impressive attachment. A confirmed appointment does not corroborate cutter competence; a mill catalog match does not authenticate the length of cloth used in the suit.

Reviews are discovery material, not a substitute for this evidence. The FTC advises examining the source, recency and breadth of reviews because reviews can be deceptive or manipulated. That supports using several evidence types rather than awarding points for stars alone. (consumer.ftc.gov)

What are the formulas?

Weighted points for a criterion = criterion weight × rating ÷ 5Total score = sum of the ten weighted-point results

Also report:

  • Documented coverage: sum of the weights of criteria supported at E2 or E3, expressed as a percentage of the 100-point framework.
  • Corroborated coverage: sum of the weights supported at E3.
  • Unknown weight: sum of the weights marked U.
  • Mandatory-check status: pass, hold or fail for each of the six checks.

Do not remove unknown criteria and rescale the remainder to 100. That would make an incomplete proposal appear stronger. Equally, do not describe low coverage as proof of poor workmanship: it describes what the buyer can currently substantiate.

What decision follows from the result?

Result Recommended action
Any mandatory check fails or remains unresolved Do not pay until the plan changes or the uncertainty is resolved
Below 60 Do not fund this proposal as written
60–74.9 Seek targeted improvements, then rescore
75–84.9 Eligible for a controlled pilot if all additional conditions below are met
85–100 Stronger documented candidate; still approve only a pilot for a new supplier relationship

Additional conditions: documented coverage must be at least 80%; criteria 1, 2, 3, 7 and 8 must each score at least 3; no mandatory buyer requirement may remain unknown. A high cloth or styling score cannot compensate for weak fitting, remedies, process, cost control or scheduling.

These are conservative editorial thresholds. They should be recalibrated when the desk has actual commission outcomes—not presented as statistically validated cutoffs.