Short answer: a publisher can have author profiles, Organization markup and consistent name without automatically seeing multiple AI citations. Entity resolution resolves identification, does not guarantee source selection. Google documents Organization, ProfilePage, and Article structured data, but does not publish an entity strength score that produces visibility. If the external output does not change, check the following causes before concluding that the "schema does not work".

Cause 1: the entity is clear, but the page does not respond to the query

A publisher may be perfectly identified, and the article may not be the most appropriate source for the question. Identity and relevance are different issues.

Cause 2: author identity is good, provenance is weak

An author profile does not replace sources. If technical claims are not supported, the transparency of the author does not solve the evidence.

Cause 3: Organization and publication are mixed up

The company may own several media brands. Organization markup, site name and editorial branding should reflect the actual relationship, not compress all entities into a single name.

Cause 4: the rebrand is incomplete

The site uses the new name, author profiles or external directories on the old one. Keep aliases and redirects, but clarify current identity.

Cause 5: author profiles are duplicated

CMS migration can create multiple profiles for the same person. Historical bylines can point to different URLs, which creates internal ambiguity.

Cause 6: syndicated copies complicate ownership

An article can appear on more than one domain. Publisher entity resolution does not automatically resolve canonicalization or syndication agreements.

Cause 7: site name and canonical domain are not stable

Subdomains, mirror sites or migrations can create multiple surfaces that look like distinct brands. Check canonical domain and redirects.

Cause 8: the byline does not correspond to the Article markup

The schema can declare one person and the page another. This is a real finding, but its repair does not guarantee external visibility.

Cause 9: External profiles are not the same entities

A personal social profile, publication page and parent company account have different roles. `sameAs' should not be used to bind different entities just for coverage.

Cause 10: the corpus has intent duplicates

Even if the entities are clear, multiple items can compete for the same query. Entity resolution does not automatically decide the editorial owner.

Cause 11: Measurement is anecdotal

A single unbranded AI response does not prove a lack of visibility. Use query set and repeated observations.

Cause 12: The KPI is wrong

If success is defined as "owned citation in any answer", you will mark failure even when the primary source is more appropriate. It measures factual accuracy and source appropriateness separately.

Reproducible decision tree

  1. Are the publishing brand and organization differentiated correctly?
  2. Are the site name and canonical domain stable?
  3. Are author profiles unique and accessible?
  4. Does the byline match the markup?
  5. Are aliases/rebrands documented?
  6. Does `sameAs' bind the same entity?
  7. Is syndication/canonicalization clear?
  8. Is the owner of the query clear?
  9. Do claims have provenance?
  10. Is the query set versioned?
  11. Is factual accuracy measured?
  12. Doesn't the KPI include citation as a mandatory result?

If 1-7 are clean, entity resolution can be considered healthy even if citations do not increase.

Baselines

Keep organization name, site name, canonical domain, author registry, duplicate profiles, alias map, syndication rules and query observations. Version the snapshot.

Severity

P0: wrong entity, wrong canonical domain or major author mismatch. P1: duplicate profiles, incomplete rebrand, syndication ambiguity. P2: secondary external profiles. P3: cosmetic differences.

How do you measure entity health

Identity mismatch rate, duplicate-author count, canonical-domain consistency, author-profile integrity and time-to-resolution are direct metrics.

How do you measure external visibility

Brand mentions, owned citations, third-party citations and factual accuracy are separate series. Do not combine them with identity health into a single score.

False attribution

If citations increase after identity cleanup, there may be other causes: new article, backlinks, breaking news, Search/model updates or changes in query mix.

It reports association, not automatic causation.

Example

A publisher rebrands, consolidates author profiles and fixes Organization markup. The citations remain stable, but the systems no longer use the old name. This is a real identity resolution result even without lift de citations.

Stop criterion

Move the project to monitoring when the identity of the organization and authors are consistent, duplicates are dealt with and rebrand aliases are stable. Don't keep adding external profiles just to increase a digital footprint.

How to audit the rebrand without losing editorial history

For publishers, the rebrand may affect the domain, site name, author profiles and social accounts at different times. Don't just delete the old name from all materials. Keep aliases where they help continuity and use redirects for moved URLs, but make the current identity explicit on active surfaces.

A good audit separates active canonical name',historical alias' and `legacy URL'. Thus, an old mention is not automatically treated as an error, and a current page using the old brand can be properly prioritized.

How to check the author graph

Build a list of authors with canonical profiles, aliases and associated articles. It looks for two types of conflict: a person with multiple profiles and a profile that aggregates different people. Both frequently occur after migrations or imports from legacy CMSs.

For each finding, write down the source URL, target profile and consolidation rule. Do not perform mass redirects before verifying that the profiles really represent the same person.

Re-audit threshold

Resume full audit after migration, rebrand or change of authorship structure. Otherwise, incremental monitoring can track duplicate profiles, broken author URLs, and new aliases.

Claim ledger

  • FACT/EVIDENCE: Google documents Organization, ProfilePage and Article structured data.
  • PRACTITIONER GUIDANCE: entity resolution and source selection must be measured separately.
  • INFERENCE: consistent identity can reduce naming ambiguity.
  • NOT PROVEN: that entity resolution directly or guaranteed produces AI citations.

Conclusion

Entity resolution is not a distribution mechanism. It makes clear who the publisher is and who the authors are. If external visibility doesn't change, look for relevance, provenance, query ownership, and source context before adding another schema layer.

Sources reviewed

Related insights