An entity is a distinct, uniquely identifiable thing, whether a person, place, company, product, or abstract concept, that a search engine can recognise on its own, separate from the words used to name it. Entities carry attributes and relationships, and they are the building blocks of Google's Knowledge Graph and of modern AI answers.
Old-school search matched strings of characters: type "jaguar" and the engine looked for that exact sequence of letters. Since the 2012 launch of the Knowledge Graph, Google works with things instead. It understands that "jaguar" could mean the animal, the car marque, or the NFL team, and it uses surrounding context to resolve which entity you mean. This is why two pages can rank for the same query using completely different vocabulary; they describe the same underlying entity.
Large language models and AI Overviews lean heavily on entity understanding to decide what is true and what to cite. A brand that exists as a well-defined entity is far more likely to be pulled into an AI answer than one the model has never resolved.
Take "Apple." The bare word is ambiguous, but Google holds Apple Inc. as an organisation founded in 1976, headquartered in Cupertino, connected to product entities like the iPhone. A smaller business earns the same treatment by building consistent signals: a Wikidata or Wikipedia reference, matching name, address, and phone details everywhere it appears, and Organization structured data, which is often the trigger for a Knowledge Panel.
Entity building sits at the heart of semantic search and increasingly of AI-driven SEO, where being understood matters as much as being crawled.