Before an engine can say anything about you, it has to decide which company your site is about.
When a site makes that hard, the answer describes somebody else, hedges, or says nothing.
Where identity breaks
An engine builds an identity from the name in the markup, the footer, the about page, the profiles the site links to, and what other sites call you. When those agree, the answer is easy. When they do not, the engine picks one, hedges, or leaves you out.
The disagreement is rarely deliberate: a legal name in the markup, a trading name in the header, an old name in the footer copyright, and a fourth form on a profile nobody has edited.
The four checks
Fifteen of the hundred points, tied for heaviest in the rubric with retrieval and trust. A citation needs somebody to attribute the claim to.
- EN-01, organization identity consistency. Whether the site names itself the same way everywhere. This check finds the four-names problem.
- EN-02, product and service identity. Whether what you sell is named consistently, and whether each name means one thing.
- EN-03, people and authorship identity. Whether the people the site credits exist in a form a machine can read. We failed this one on our own site, and the finding is published.
- EN-04, external identity reconciliation. Whether the profiles you link to agree with the site. Linking a stale profile endorses what it says.
What this is not
We are not a knowledge-panel service and submit nothing on your behalf. No form makes an engine know you, and nobody can sell access to a knowledge graph.
What you control is whether every source an engine can read agrees. That is a finite, checkable job, and it is what this dimension measures.
What your site says about you
The free scan reads the identity your markup and head tags declare and returns it quoted from your own pages, without judgement. Anything wrong or missing in it is the finding.
The entity clarity dimension is published in full, and getting cited covers what identity is for once an engine has settled it.