An entity is a uniquely identifiable object that a language model recognises as a distinct “thing” and links to other things: the brand CodaAI, the product Digital Visibility Audit, the city of Gütersloh, the term grounding. Entities are the nodes in a model’s knowledge. Entity density describes how many such nodes occur in a section of text — particularly in the first paragraphs of a page.
How do entities work in AI systems?
During training, a model learns which names occur together: which brand belongs to which industry, which product to which manufacturer, which person to which company. The more frequently and consistently an entity is described in connection with a topic, the more stable that link is in the model knowledge. In web search, in turn, entities help to classify a passage: a section with concrete product, brand and technical names is easier for a system to assign to a topic than a section made of generic terms.
Why do entities matter for AI visibility?
“Brands matter more than websites”: the AI system does not care whether information about a brand comes from its own website or from third-party sites. What counts is a consistent picture of the entity across all sources. A brand that is established as an entity is mentioned even when its website does not appear in grounding. A brand the model does not know as an entity does not appear in brand-neutral answers — no matter how good the website is.
What does this mean for your website?
Increase the entity density in the first paragraphs: place topic-relevant names, terms and references early, and use concrete product, tool and brand names instead of vague generic terms. Name your own brand and your products explicitly. And make sure the description of your brand is the same on every platform (consistent brand description) — the entity emerges from the sum of the sources, not from one.