Also known as: Entity, Named Entity, NER, Named entity
An entity is a concrete, uniquely identifiable concept from the real world: a person (Albert Einstein), a place (Berlin), an organisation (Google), a product (iPhone 15), a term (quantum mechanics). Unlike pure keywords, entities have a unique identity that is independent of how they are spelled: ”Apple Inc.”, ”Apple Computer” and ”Apple” as a company all refer to the same entity, whose Wikidata ID is Q312. Search engines and LLMs increasingly work with entities rather than keywords — entity understanding is a basic prerequisite for semantic search and AI answers.
Named Entity Recognition is the automatic detection of entities in free text — for example: ”Steve Jobs founded Apple in Cupertino” → three entities recognised (person, organisation, place) and linked to their knowledge base IDs. NER models are part of search engine pipelines (Google), LLM pipelines (ChatGPT, Claude) and content analysis tools. If your own website names entities clearly and consistently, NER can recognise them reliably and link them to the correct knowledge graph entry.
Three core practices: (1) Explicitly name all important entities on the topic — anyone writing about ”search engine optimisation” should mention Google, Bing, Schema.org, Wikidata, PageRank and similar central entities. (2) Schema.org markup with sameAs links the named entities to Wikidata QIDs. (3) Consistent spellings — clearly mark ”SEO” and ”search engine optimisation” as synonyms instead of mixing different terms for the same thing. The effect: AI engines recognise the thematic completeness and authoritativeness of the page.
A systematic method: take the top three competitors for your own money keyword, run NER over their texts and generate a list of all recognised entities. Then analyse your own page with the same NER. Entities that appear with several competitors but are missing from your own page are the entity gaps — closing them (a short explanation, a Schema.org sameAs link) measurably increases thematic authority and the chances of citation in AI answers.
Example: A website on the topic of ”page speed” had 90% topical relevance, but the entity analysis showed that competitors consistently mention Lighthouse, CrUX, the Web Vitals JS Library and WebPageTest — while the site itself only mentioned ”Google PageSpeed Insights”. After adding these missing entities with short explanations and schema sameAs links, the page ranked in the top 10 for six additional long tail keywords within eight weeks and appeared consistently as one of the cited sources in ChatGPT answers about ”measuring page speed”.
Entity gap analysis in the GEO audit
Free SEO & GEO Check
SEO score, AI visibility and citability of your website in 30 seconds — no registration required.
Register for free, get 10 credits and start right away.
Register now