Semantic search · Search by meaning
Semantic search seeks to understand the meaning and intent of a query rather than match exact words. It generally relies on embeddings and is the basis of modern AI engines.
Standalone definition · 30 words
How does it work?
Query and documents are converted into vectors, and the closest are returned, even with no word in common. It is often combined with lexical search (BM25) in hybrid systems, because the two cover different cases.
How does it differ from vector search?
Semantic search is the goal, understanding meaning. Vector search is one of the means, through embedding proximity. Engines may also rely on query rewriting or reranking models.
What does it change for content?
Google writes that its systems understand synonyms and general meanings, and that you need not capture every phrasing variation. A clear, complete, well-structured piece beats an accumulation of neighboring keywords.
Frequently asked questions
Does semantic search eliminate keywords?
No. Exact terms keep a role, notably for proper names and references, hence hybrid systems.
How do you audit semantic coverage?
By comparing your pages’ embeddings with those of target queries and spotting missing or diluted topics.
Related terms
Further reading
- Profound vs Otterly vs Ahrefs Brand Radar: query fan-out comparison
- ← Back to the GEO glossary (114 terms)