Product search
The shop search box and ranking behind it, matching what a shopper types to the products they are most likely trying to find.
See it
What it is
Product search turns a shopper's words into a ranked product list. It matches more than titles: category, brand, description, attributes, SKU, synonyms, and common misspellings can all help. Services such as Algolia and Elasticsearch combine text matching with business signals like availability and popularity.
Reach for it as soon as the catalog is too large to scan from a menu. Useful store search includes query suggestions, typo tolerance, highlighted matches, facets, and a results page whose query and filters live in the URL.
Gotcha: relevance cannot be judged from a few happy-path searches. Log zero-result queries and refinements, then add synonyms or catalog data for the language shoppers actually use. Never let popularity bury an exact SKU or product-name match, and push unavailable items down rather than returning a first page full of dead ends.
Ask AI for it
Build product search with Algolia. Configure searchableAttributes so SKU and title rank ahead of brand, category, attributes, and description; enable typoTolerance; add explicit synonym sets; and use customRanking to prefer in-stock products before sales count. Add a debounced autocomplete with query suggestions and highlighted matches, plus a full results page with facets and the query stored in the q URL parameter. Track submitted queries, clicked result positions, and zero-result queries without recording customer email addresses or other free-form personal data.