Use this worksheet to connect what shoppers expected with what they could find, choose and receive. Review one product family at a time; do not combine unrelated problems into one score.
Return cohort definition
- Product family and variants: [details]
- Original shipment dates included: [start and end]
- Return observation cutoff: [date]
- Has the normal return window elapsed? [yes/no; explain exclusions]
- Units shipped in the cohort: [A]
- Units from that cohort returned by the cutoff: [B]
- Unit return rate: [B divided by A multiplied by 100; undefined if A is zero]
- Separately tracked outcomes: [cancellations, refunds without returns, exchanges]
Hypothetical example: 18 returned units from 200 shipped units gives 9%. This is a calculation example, not a benchmark. A return can be processed in a different month from the original order.
Reason and remedy log
| SKU | Customer reason | Supporting detail | Preventable cause to investigate | Owner | Change | Review date |
|---|---|---|---|---|---|---|
| [SKU] | [reason] | [ticket or inspection] | [hypothesis] | [name] | [action] | [date] |
Check fit, compatibility, description accuracy, packing, damage, picking errors and product quality. Preserve the customer's wording; do not assume a selected reason proves the root cause.
Five discovery tasks
Test with actual customer language on mobile and using a keyboard. Use only compatibility claims verified against product specifications.
| Task | Search term or starting collection | Expected useful result | Observed result | Severity | Fix |
|---|---|---|---|---|---|
| Find a compatible product | [term] | [result] | [result] | [impact] | [action] |
| Choose a size | [term] | [result] | [result] | [impact] | [action] |
| Compare similar options | [term] | [result] | [result] | [impact] | [action] |
| Find a refill | [term] | [result] | [result] | [impact] | [action] |
| Recover from no results | [term] | [helpful next step] | [result] | [impact] | [action] |
- Filters use complete and consistent product attributes.
- Active filters are visible and removable.
- Back navigation preserves a useful browsing state.
- Sold-out variants and delivery expectations are clear.
- Comparison information uses matching units and criteria.
- Synonyms connect genuinely related terms.
Before-and-after review
Record dates, counts and changes in traffic or product mix. Compare mature return cohorts. Track purchase completion, support contacts and contribution alongside the target metric. A small sample supports a hypothesis, not a guaranteed lift.
- Primary issue and evidence: [details]
- One change and release date: [details]
- Primary measure and denominator: [details]
- Return costs included without double counting: [details]
- Customer experience guardrail: [details]
- Outcome and next decision: [details]
Read the returns guide and search and merchandising guide before interpreting the results.
Want help applying this to your store?
Take the free store assessment for a suggested next step, or book a 15-minute call to talk it through.