Vector Search
Vector search is a retrieval method that finds passages by meaning rather than keywords, comparing the numeric embedding of a query against the embeddings of indexed content to surface the closest matches.
Also known as: semantic search
Vector search finds content by meaning, not exact words. It represents both the question and every indexed passage as embeddings — vectors positioned by meaning — and retrieves the passages whose vectors sit closest to the query's. This semantic matching is the retrieval backbone of modern answer engines.
For AEO, vector search is why you optimize for clear ideas rather than keyword density. A passage that expresses one concept cleanly and completely lands in a precise spot in vector space and gets matched to the questions it truly answers; a vague, padded, or multi-topic passage embeds into a muddy region and matches nothing well. Writing one self-contained idea per passage — the extractability discipline — is, in effect, optimizing for vector search.
Example. A page titled "trimming an overgrown hedge" can be retrieved by vector search for the query "how do I cut back a bush that got too big," because the meanings are near each other in vector space even though the wording differs completely.
Relevant pillar
Related terms
- EmbeddingsEmbeddings are numerical representations of text that capture its meaning, letting AI systems find passages that are semantically related to a query even when they share no exact keywords.
- RerankingReranking is a second pass in retrieval where an initial set of candidate passages is reordered by a more precise relevance model, deciding which few actually make it into the AI's answer.
- RAG (Retrieval-Augmented Generation)RAG is the technique behind most AI answer engines, where the model first retrieves relevant documents from the live web or an index and then generates an answer grounded in what it found.