Rubric v0.8 · 4 checks · 15 of 100 points
Does an answer engine know which company your site is about?
An engine has to decide which organization a page is about before it can recommend that organization. Conflicting names and missing contact facts force it to guess, and it may guess wrong.
Why this dimension exists
An answer engine cannot recommend an organization it cannot identify. It first has to decide which company a page is about, and whether it is the same company as on the other twelve pages.
People resolve this without noticing. A reader who sees a trading name in the header, a legal entity in the footer, another spelling on the invoice and a fourth variant on LinkedIn sees one organization. A retrieval system may see four. It has no reason to assume the strings co-refer, and where facts conflict it picks one or declines to name you.
The four checks cover the organization, its offerings, its people, and whether the identity the site asserts matches what the rest of the web asserts. Sites control that last one least and neglect it most.
For somebody asking why an answer engine does not seem to know their company, the same ground is covered <a href="../../entity-recognition/">without the scoring</a>.
The guide for this dimension.
- Entity recognition: how an answer engine works out which organization a site belongs to, and where that breaks.
The 4 checks
None of these is deterministic. Every one is a reading rather than a measurement, so the free scan names them as not covered rather than guessing.
EN-01 — Organization identity consistency
What is tested. Compare organization name, description, address, contact information, and identifiers across major pages and structured data.
What is recorded as evidence. Source locations and conflicts.
What fails it. Multiple unexplained names; obsolete descriptions; inconsistent locations; unclear legal and product identities.
How to fix it.
- Decide which name is the operating entity and which, if any, are brands or products under it.
- Use that name identically in the <title> suffix, the og:site_name, the footer and the Organization structured data.
- Publish one page that states who operates the site, where, and how to reach a human, and link it from the footer.
- Add Organization structured data with name, url, logo, and sameAs pointing at the profiles you actually control.
How to confirm it is fixed. Search your own brand name and confirm the entity a result describes is the one you meant.
Reference.
- Organization structured data — Google Search Central
- Organization — schema.org
- Schema Markup Validator — schema.org
- Structured Data for beginners — Google Search Central
EN-02 — Product and service identity
What is tested. Determine whether each primary product or service has a stable name, URL, concise definition, audience, and relationship to the organization.
What is recorded as evidence. Product pages, navigation, structured data, internal-link relationships.
What fails it. Interchangeable product names; orphaned offerings; marketing language that never explains what is sold.
How to fix it.
- List what you actually sell, in the words a buyer would use rather than an internal name.
- Give each one a page that says what it is, who it is for, what it costs or how pricing works, and what it excludes.
- Add Service or Product structured data on those pages, with the same names the page uses.
- Remove or rename anything the site still advertises that you no longer sell.
How to confirm it is fixed. Ask an answer engine what your company sells and compare the answer with your own list.
Reference.
- Product structured data — Google Search Central
- Service — schema.org
- Schema Markup Validator — schema.org
- How do I get started with structured data? — Google Search Central
EN-03 — People and authorship identity
What is tested. Inspect author bylines, biographies, leadership pages, credentials, and profile relationships.
What is recorded as evidence. Author URLs, biography content, structured-data relationships when present.
What fails it. Unattributed expertise content; placeholder authors; contradictory roles; credentials that cannot be tied to a person.
How to fix it.
- Give every piece of authored content a named author, not a company byline or 'admin'.
- Give each author a page with their role, their relevant background, and how to verify it.
- Link author to organisation both ways, in the prose and in Person or Article structured data.
- Keep the same spelling of a name everywhere it appears.
How to confirm it is fixed. Open an article and confirm you can get from the byline to a page that says who wrote it.
Reference.
- Article structured data — Google Search Central
- Person — schema.org
- Structured data for developers — Chrome for Developers
EN-04 — External identity reconciliation
What is tested. Compare important owned profiles and repositories named by the customer with the website's identity claims.
What is recorded as evidence. Explicitly linked profiles, repository ownership, material inconsistencies.
What fails it. Old brand profiles; conflicting domains; unlinked authoritative properties.
How to fix it.
- List every external profile that names your organisation: directories, registries, social, review sites, industry bodies.
- Correct the name, address, URL and description on each so they match the site.
- Claim or close the ones you do not control but that carry stale details.
- Point sameAs at the corrected profiles once they agree.
How to confirm it is fixed. Read the top ten external results for your brand name and check each against the site.
Reference.
- Organization structured data — Google Search Central
- Organization — schema.org