Search & Merchandising Signals: Facets, Ranking & Recommendations

September 11, 2025 9 min read

Turn clean attributes into high-relevance search: facet governance, ranking signals (text, popularity, margin), boosts, synonyms and recommendations.

Objective: use normalized product data to power search that understands users and merch rules.

1) Facets that guide, not confuse

2) Ranking signals

3) Synonyms & spelling

4) Boosting & demotion

5) Recommendations

6) Measurement

Outcome

Clean attributes + curated facets + measured boosts = faster findability and higher revenue per visit.

What's next

Next up: Catalog Health Metrics — measuring data quality, coverage and staleness to prioritize fixes.

Share LinkedIn X/Twitter