Vector Search
A retrieval method that finds content based on semantic similarity, represented as numerical embeddings, rather than exact keyword matching.
What it means
Vector search converts text into numerical embeddings that capture meaning, so a query for "warm winter jacket" can retrieve a product described as "insulated cold-weather coat" even without shared keywords. It's the retrieval mechanism underneath most RAG systems and AI search platforms. A product catalog that's been embedded and indexed for vector search is retrievable by semantic intent, not just literal keyword overlap, which matters because buyer queries to AI platforms are rarely worded the way product titles are.
Why it matters for Shopify
Shopify merchants whose product content is embedded and indexed for vector search can be retrieved by AI platforms even when a buyer's query doesn't share exact keywords with their product titles or descriptions.