Semantic search
Semantic search is the ability of a search engine to understand the meaning and intent behind a query, not just match its exact words.
Early search engines matched keywords. Semantic search understands that ‘cheap flights to the Big Apple’ means low-cost flights to New York, and that a page about ‘reducing churn’ is relevant to ‘keeping customers’.
In more detail
Semantic search relies on language models, entities and the relationships between concepts. It lets engines rank pages that answer the question well even when they do not repeat the exact phrase. It also means stuffing a keyword into a page does little; covering the topic thoroughly, with related concepts and clear answers, does far more.
Why it matters
Write for the topic and the intent, not the keyword string. Semantic understanding is also the foundation of AI search, so content that covers a subject fully and clearly performs well in both. See topical authority.
A page titled 'How to keep subscribers from cancelling' ranks for 'reduce subscription churn' even though it never uses the word 'churn', because it covers onboarding, pricing changes and cancellation flows that clearly match the intent.