Ecommerce Search Recovery Checklist | Scouty

This is the checklist a Scouty specialist works through during an expert-led Search Audit, condensed into a self-serve guide. It will not catch every issue. It will catch most of the ones that actually move conversion.

Run it once. Then run it monthly.

Phase 1. Triage

1. Pull the failed-search list

• Export the top 50–100 zero-result queries by volume for the last 30 days.
• Export the top 50 queries with results but zero clicks.
• Note seasonality. Some queries are seasonal noise; flag them.

2. Sample your storefront

• Pick five queries from each list above and run them on your live store.
• Take screenshots. You will refer back to these.
• Note response time, result quality, and obvious failures.

3. Estimate impact

• For each failed query, estimate session-to-cart value if it had worked.
• Use even rough averages. A directional estimate is fine.

Phase 2. Root-cause each failure

For every failed query, classify into one of these buckets:

• Synonym mismatch. Shopper says X, catalog says Y.
• Attribute gap. Searched attribute is not indexed.
• Typo / stem. Stemming or typo tolerance is off.
• Out-of-stock or unpublished. Product hidden by stock policy.
• Tokenization. “BMW-E92” vs “BMW E92” vs “BMWE92.”
• Genuine catalog gap. You don’t sell it. (Document for sourcing.)

Phase 3. Fixes

4. Synonyms

• For high-volume mismatch queries, add a synonym rule.
• Decide one-way vs two-way explicitly.
• Don’t delete existing rules without checking impact.
• Group rules by category to keep coverage manageable.

5. Attributes

• Index the attributes your shoppers actually search by (size, color, material, model, fitment).
• Make sure variants surface their own attributes, not just the parent product.
• Check that “internal” attributes (SKUs, internal codes) are searchable when they should be.

6. Typo and stemming

• Confirm typo tolerance is enabled.
• Spot-check ten common typos in your category.
• Test stem cases (“hike,” “hiker,” “hiking”).

7. Stock policy

• Decide explicitly: do out-of-stock items show in results? Where?
• If hidden, are restock notifications offered?
• If shown, is stock status clear in the result card?

8. Tokenization

• Test SKU and model variants with and without separators.
• If your engine doesn’t handle alternates, add synonyms.

Phase 4. Filters and facets

9. Filter audit

• List the filters available on your top three category pages.
• Compare to what shoppers actually search for.
• Are critical filters missing? Add them.
• Are unused filters cluttering? Remove or hide.

10. Filter usage

• What percentage of search sessions apply at least one filter?
• If under 20%, look at filter visibility, defaults, and value ordering.

11. Filter quality

• Are filter values consistent (e.g., “S, M, L” vs “Small, Medium, Large”)?
• Are filter counts accurate?
• Do filters interact correctly (selecting “blue” + “size 9” returns sane results)?

Phase 5. Merchandising

12. Boost rules

• Are your top sellers boosted on relevant queries?
• Are seasonal items boosted seasonally?
• Are out-of-stock items buried?

13. Pin rules

• Do top brand searches pin the right hero product?
• Are pins reviewed quarterly to avoid stale top results?

14. Campaign rules

• If you run promotions, are search results aware of them?
• Do campaign products appear in autocomplete?

Phase 6. Discovery surfaces beyond keyword

15. Semantic / use-case search

• Run five “use case” queries (e.g., “for sweaty feet”, “for sensitive skin”).
• Are relevant products surfaced? If not, semantic search is probably the lift.

16. Image / similarity search

• In fashion, home, parts, beauty: is search-by-image available?
• If not, would your imagery support it (clean primary shots, variant coverage)?

17. Document search

• Are PDFs, manuals, and spec sheets searchable?
• Are they linked to products?
• Are page-level snippets shown?

18. AI Q&A

• Do you have a use case for grounded product Q&A or guided buying?
• If yes, are sources indexed and citations surfaced?

Phase 7. Measurement

19. Track the impact

• After applying a fix, watch the same query for 30 days.
• Did zero-result rate drop? Did clicks rise? Did carts rise?

20. Build the routine

• Schedule a weekly review (top 10 issues).
• Schedule a monthly deep dive (full checklist).
• Schedule a quarterly review with merchandising stakeholders.

When to ask for help

If you’ve run this checklist and you still see a meaningful zero-result rate, persistent low search-to-cart, or you’re trying to decide whether semantic, visual, document, or AI search is worth turning on, the next step is to bring an expert in.

Request a free expert-led Search Audit. A Scouty specialist will run a manual review of your storefront, sample queries, and add-on fit and send you a short report.