Autocomplete

Inline suggestions that complete a search query as the shopper types, surfacing products, categories, or popular searches before they hit Enter.

Autocomplete (also called typeahead or instant suggestions) reduces the friction of searching by predicting the most likely query completions as a shopper types. Good autocomplete pulls from at least three sources: previously successful searches across the store, current product titles and categories, and a popularity signal so trending queries appear first.

On ecommerce sites, autocomplete materially affects conversion: shoppers who interact with a suggestion convert 2–6× more often than shoppers who type the full query themselves. The mechanic also lets stores recover from typos before they happen — a half-typed “sneek” can still surface “sneakers” if your suggestion index is fuzzy-tolerant.

Common gotchas: serving stale suggestions after a catalog refresh, ignoring per-language indexes for multilingual stores, and over-indexing on raw popularity (which buries newer products). Most search platforms now combine BM25 lexical matching with embedding similarity so semantic near-misses (“running shoes” → “trainers”) still surface.

Related terms