Brand entity

A brand entity is the record that a machine’s knowledge representation uses to treat a brand as a single, uniquely identifiable object — rather than just a string of characters that happens to appear in text.

When a search engine or a large language model encounters “Acme Corp,” it first has to map that string to a specific thing. This step is called entity linking. When the name is unique, it happens automatically. When it isn’t, the system picks a default based on the strength of evidence in its data, and that choice decides who gets credit for everything said about that name afterward.

This is a common problem for companies with generic or shared names, and it is especially common in China, where many companies share the same short trade name across different cities and industries.

What an entity is made of

• Unique identifiers. Full registered legal name, business registration number (in China, the unified social credit code), official domain.
• Distinguishing attributes. Founding date, registered location, business scope, founders.
• Relationships. Products, subdomains, social profiles, links from press coverage.

The job of these attributes is to provide distinguishing features. A description like “a company that does AI” cannot disambiguate anything, because it is equally true of the other company with the same name.

How to get the entity established correctly

• Create a clear entity page on your site. State in one sentence at the top which organization this is, then list the key facts.
• Mark it up with structured data. Use the Organization type to structure those fields, and list every domain and profile that belongs to the same organization in sameAs.
• Say explicitly which same-name organizations you are not.

That last step is unusually valuable. What a model needs is distinguishing features, and “X is not Y” supplies one directly. It also gives any answer that retrieves the page something solid to rely on. The effect is measured by name confusion rate.

Consistency matters as much as presence. If your website, your LinkedIn page, a directory listing and a press release describe your business scope in four different ways, you are giving the model four slightly different entities to reconcile. Keep one canonical set of facts and use it everywhere.

How it relates to knowledge graphs

A knowledge graph is a collection of entities and the relationships between them. Building a brand entity is essentially submitting a self-consistent, verifiable record to the various knowledge graphs out there — the search engines’, the ones implicit in model training data, and third-party databases.

Third-party evidence is what makes that record believable: independent sources that state the same facts and link back to the same domain.

Why it matters for AI visibility

Entity work sits underneath everything else in generative engine optimization (GEO). Answer engine optimization (AEO) can make your content easy to extract, but if the model maps your name to someone else, the extracted answer goes to them. That is why an AI visibility audit checks the entity layer first.