Author byline schema: 2026 JSON-LD walk-through

TL;DR

Keep author.name to the person's name only. Push titles, employers, and credentials into dedicated properties like jobTitle, worksFor, and hasCredential.
Use sameAs to point at authoritative external profiles. Wikidata is the strongest single signal — add it before LinkedIn or social.
Use @id-based entity linking so the same Person node is referenced consistently from Article, Organization, and across pages.
Signal expertise with knowsAbout (topics) and knownFor (notable works or accomplishments).
Schema must match what's visible on the rendered page. If the byline doesn't show it, don't claim it in JSON-LD.

Author byline schema is the structured-data layer that tells search engines and LLM-based answer engines who wrote a page and why they're qualified. In 2026, this matters for two overlapping reasons: Google's E-E-A-T evaluation still leans on author entity signals, and generative engines (ChatGPT, Perplexity, Google AI Overviews, Claude) use the same machine-readable cues to decide which sources to cite. The pattern below is what we recommend to CiteFlow customers shipping production markup today.

The 2026 baseline Person object

Google's Article structured data documentation is explicit: put the author's name in author.name and use other properties for everything else. That means no "Dr. Jane Smith, PhD, Senior Editor" stuffed into a single string. Split it out.

A correct baseline Person looks like this:

{
  "@context": "https://schema.org",
  "@type": "Person",
  "@id": "https://example.com/team/jane-smith#person",
  "name": "Jane Smith",
  "givenName": "Jane",
  "familyName": "Smith",
  "url": "https://example.com/team/jane-smith",
  "image": "https://example.com/img/jane-smith.jpg",
  "jobTitle": "Senior Cardiology Editor",
  "worksFor": {
    "@type": "Organization",
    "@id": "https://example.com/#organization"
  },
  "hasCredential": {
    "@type": "EducationalOccupationalCredential",
    "credentialCategory": "degree",
    "name": "MD, Johns Hopkins School of Medicine"
  }
}

Two details worth flagging. First, the @id is a stable URI — not necessarily a real URL, but a unique identifier you can reference from every article she writes. Second, worksFor references the publisher's Organization node by @id rather than redefining it inline. This is the entity-linking pattern most 2026 implementations use, and RightBlogger's author schema guide walks through the same decision logic.

sameAs in 2026: Wikidata first, then everything else

sameAs is how you disambiguate your author from every other "Jane Smith" on the internet. The status of which links matter has shifted. Social URLs still help, but the strongest signal is a Wikidata Q-number, because Wikidata is the entity graph most LLMs and Google's Knowledge Graph reconcile against.

Order your sameAs array by authority:

"sameAs": [
  "https://www.wikidata.org/wiki/Q123456789",
  "https://en.wikipedia.org/wiki/Jane_Smith_(cardiologist)",
  "https://orcid.org/0000-0002-1825-0097",
  "https://www.linkedin.com/in/janesmithmd",
  "https://scholar.google.com/citations?user=XYZ",
  "https://twitter.com/janesmithmd"
]

If your author doesn't have a Wikidata entry yet and they publish original work, create one. Schema App's E-E-A-T guide makes the same point: an entity that exists in Wikidata is an entity that generative engines can resolve.

Expertise signals: knowsAbout and knownFor

name and sameAs answer "who." knowsAbout and knownFor answer "why you should trust them."

knowsAbout is a list of topics the author has demonstrated expertise in. Use Wikipedia or Wikidata URLs as values, not just strings, so they resolve to entities.
knownFor is for specific notable works, awards, or accomplishments. Reference them as CreativeWork, Award, or similar typed entities.

"knowsAbout": [
  "https://en.wikipedia.org/wiki/Cardiology",
  "https://en.wikipedia.org/wiki/Heart_failure",
  "https://www.wikidata.org/wiki/Q11085"
],
"knownFor": [
  {
    "@type": "CreativeWork",
    "name": "ACC/AHA Heart Failure Guidelines (2022 contributor)",
    "url": "https://www.acc.org/..."
  },
  {
    "@type": "Award",
    "name": "AHA Distinguished Scientist Award, 2021"
  }
]

Search Engine Journal's E-A-T structured-data piece and AutomateEd's 2026 author markup guide both call this out: vague claims of expertise don't help. Concrete, entity-linked claims do.

Wiring the byline into the Article

The Article references the Person by @id, not by redefining it. This keeps your graph clean and lets crawlers consolidate signals across every article that author has written.

{
  "@context": "https://schema.org",
  "@type": "Article",
  "headline": "How GLP-1s affect cardiovascular outcomes",
  "datePublished": "2026-03-14",
  "dateModified": "2026-03-18",
  "author": { "@id": "https://example.com/team/jane-smith#person" },
  "publisher": { "@id": "https://example.com/#organization" },
  "mainEntityOfPage": "https://example.com/articles/glp1-cardiovascular"
}

For multi-author pieces, author becomes an ordered array of @id references. Order matters — list the lead author first, matching the visible byline. Don't invent co-authors to pad the list, and don't collapse two people into one node.

Common mistakes that break the markup

After auditing thousands of author schemas, these are the recurring problems:

Schema-page mismatch. If JSON-LD says the author is "Dr. Jane Smith, MD" but the rendered byline just says "Editorial Team," Google treats the markup as untrustworthy. Whatever you put in schema must be visible on the page.
Name stuffing. Putting credentials, employer, and titles inside name violates Google's documented guidance and dilutes the entity match.
Dead sameAs links. A Twitter URL for an account that no longer exists is worse than no link. Audit quarterly.
Per-article duplication. Redefining the full Person object on every article creates fragmented entities. Define once, reference by @id.
No author page. If author.url 404s or returns thin content, the entity has nowhere to consolidate signals.

FAQ

Do I still need sameAs if my author has a Wikidata entry?

Yes — but the Wikidata URL goes inside sameAs. It's the highest-value entry in that array. Add ORCID, LinkedIn, and verified social profiles after it.

Can I use Person schema for a pseudonymous author?

Technically yes, but it weakens E-E-A-T signals because there's no external entity to reconcile against. If the author is pseudonymous, at minimum link sameAs to platforms where the pseudonym has an established history (Substack, GitHub, Mastodon).

Does author schema help with AI Overviews and ChatGPT citations?

Indirectly. Generative engines pull from sources that resolve cleanly as entities. A page with a well-structured Person graph, Wikidata-linked author, and matching visible byline is easier for an LLM-based system to attribute and cite than an anonymous post.

Author byline schema: a 2026 walk-through