Entity clarity, explained.
Before an assistant recommends a business, it has to be certain which business it is talking about. Entity clarity is how certain you let it be.
What an “entity” is to a machine
To a retrieval system your business is a bundle of facts that should all point at the same thing: a name, an address, a phone number, a website, a category, a set of services, an area served. When those facts agree across your site and the sources that mention you, the system forms a single, confident entity. When they disagree — two spellings of the name, an old address on one directory, a category that does not match your services — it forms two fuzzy entities, or one it does not trust.
Why fuzzy entities do not get recommended
An assistant that is unsure whether “Marisol Café” in the news article is the same as “Café Marisol” on the review site will not merge what it knows. Each half is thinner than the whole, and neither clears the confidence bar. The assistant names a competitor whose facts line up, and it never tells you why.
The five things that make an entity clear
- One canonical name, used identically on your site, in schema, in directories and in your own social profiles.
- A stable address and phone in plain text, in schema, and matching everywhere.
- A one-sentence definition — “Café Marisol is a specialty coffee bar in Ruzafa, Valencia, open since 2011” — near the top of your home page and repeated in your About page and schema description.
- A category that matches your services. If your schema says “Restaurant” and your pages say “coffee roaster”, pick the one buyers ask for.
- Links between your own properties. Your site links to your social profiles; they link back. Schema
sameAslists them all.
Common ways clarity breaks
- A rebrand that left old names live on directories and in old articles.
- A move, with the old address still cached in the largest directory.
- A second location with the same name and no distinguishing city in the name or schema.
- A site that renders name and address only in an image or a map embed.
- Another business with the same name in another city, and nothing on your pages that says which one you are.
How the report measures it
Our crawler extracts every name, address, phone and category it finds on your site and compares them with the values in your schema and the values the assistants used when they described you. Disagreements are listed as entity issues with the exact strings that conflict, so the fix is mechanical.