Anchor the scale
Describe what weak, acceptable, and strong evidence looks like for each measure. Record the period and sample behind an assessment. Leave an item unscored when the evidence is missing.
Review disagreement
If two reviewers score the same evidence differently, discuss the definition before averaging their answers. The useful outcome is a clearer decision, not a more impressive-looking number. Keep your personal assessment separate from public customer reviews.
Keep an evidence trail
For each assessment, record the requirement, the supporting example or document, the date, and the unresolved questions. Distinguish a published capability from an account-specific demonstration. Mark missing information as unknown. Where you use a numerical score, explain the meaning of the scale and the reason for the weight. Your notes should make it possible to revisit a conclusion when the proposal or operating conditions change.
Define a test before recording a score
Write the expected result, the sample, and the evidence you will accept. Keep a failed requirement separate from an unanswered question. Use the same test across providers where their proposed scopes are comparable. If the test changes, record why; otherwise an improved score may reflect an easier question rather than a better fit.
Use this question in your next meeting:
What would someone need to see to reach the same conclusion?
Go deeper: Build a 3PL scorecard that can survive a second look. The companion article includes an editable worksheet and links to provider profiles.
Further reading
Shopify: fulfilling orders. Background on fulfillment workflows and working with fulfillment services.