AEO playbook

Person Schema and Authority Sites: The Quiet Signal That Decides Who AI Cites

Why Person schema and a personal authority site are the foundation of AEO. How structured identity, an entity graph, and one canonical source make AI tools cite you instead of someone else.

By Selina Eizik·April 23, 2026·11 min read

What Person schema actually is

Person schema is a block of structured data — written in JSON-LD — that lives in the <head> of your website and formally describes a human being to machines. Name. Job title. Employer. Photo. Areas of expertise. The other profiles that belong to that same person. It is invisible to visitors and extremely visible to ChatGPT, Perplexity, Claude, Google AI Overviews, and the traditional search crawlers.

Without it, every AI engine that lands on your site has to infer who you are from prose. With it, they don't guess — they read a labeled record. That's the entire game in AEO. Models that have to guess cite someone else. Models that have facts cite you.

This piece is the strategic companion to our plain-English schema markup guide. That one is the how. This one is the why — and how to use Person schema as the foundation of a personal authority site that compounds over time.

Why it matters for AI search and brand

Three shifts have made Person schema the most undervalued asset in modern marketing:

AI engines reward verified identity. ChatGPT, Perplexity, and Google's AI Overviews are designed to surface named experts, not anonymous content. A clear Person entity tells them who to attribute the expertise to.
E-E-A-T is now machine-readable. Google's Quality Rater Guidelines have always rewarded experience, expertise, authoritativeness, and trust. Person schema is the structured layer that lets algorithms — not just human raters — actually parse it.
Personal brand outperforms company brand in AI recommendations. When buyers ask "who is the best [thing] for [job]," AI tools overwhelmingly return people, not logos. If your name is the entity AI knows, you win the recommendation.

What an authority site is (and isn't)

An authority site is a website built around a single person (or small expert team) whose job is to consolidate everything that proves their expertise into one canonical source: bio, photo, credentials, employer, speaking history, writing, press, and external profiles.

It is not a portfolio, a personal blog, a Linktree, or a company website with an "About the founder" page bolted on. Those things are fine — but they do not give AI engines a single, unambiguous Person entity to reference. An authority site does.

Done right, an authority site becomes:

The canonical source AI tools cite when asked about the person or their topic.
The hub that every external mention (podcasts, articles, conference pages, LinkedIn) links back to.
The proof layer for every claim — credentials, companies, talks — backed by structured data and outbound sameAs links.

How Person schema builds your entity graph

Modern AI engines do not think in pages. They think in entities and the relationships between them. Person schema is how you put yourself on that map.

A single, well-written Person block establishes:

Identity — your name, photo, and the URL that uniquely represents you (the @id).
Affiliation — what company you work for or founded, via worksFor, with that company's URL.
Expertise — the topics you are an authority on, via knowsAbout.
Cross-platform identity — LinkedIn, Substack, Wikidata, conference profiles, your company site — all declared together via sameAs. This is how an AI engine confirms that "the Selina Eizik writing about AEO" is the same person as "the Selina Eizik who keynoted KW Family Reunion" and "the Selina Eizik who co-founded Conductor."

Once those edges exist in the graph, every new article, talk, or mention compounds. The model already knows who you are. It just adds another data point to the entity it has already resolved.

Anatomy of a great Person schema block

Most Person schema in the wild is bad. It's auto-generated by a plugin, it's missing the fields that matter, and it's duplicated across pages with conflicting values. A good Person block has eight properties working together:

@id — a stable URL fragment like https://yoursite.com/#person. Use this exact same ID everywhere you reference yourself, including as the author of every Article.
name — your full, real name, exactly the way you want to be cited.
image — a real headshot, hosted on your own domain. Avoid avatars and illustrations.
jobTitle — current role, written in human language ("Founder & Marketing Strategist," not "CEO/CMO/COO").
description — 2–4 sentences that include your biggest credentials, companies, and the specific topics you're known for. Write this for a model that has 60 tokens to summarize you.
worksFor — your current company as a nested Organization with its URL.
knowsAbout — an array of the exact topical phrases you want to be cited for. These should match the queries your buyers actually type into AI tools.
sameAs — every external profile that proves you exist: LinkedIn, your company site, speaking pages, Substack, Wikidata if you have it. This is the single highest-leverage field for AI verification.

A real, working example

Here is the exact Person schema block from selinaeizik.com. Adapt it for your own site.

<script type="application/ld+json">
{
  "@context": "https://schema.org",
  "@type": "Person",
  "@id": "https://yoursite.com/#person",
  "name": "[Your Full Name]",
  "url": "https://yoursite.com",
  "image": "https://yoursite.com/headshot.jpg",
  "jobTitle": "[Your role, in human language]",
  "description": "[2–4 sentences with credentials, companies, expertise topics — written like an AI summary of you.]",
  "worksFor": {
    "@type": "Organization",
    "name": "[Your Company]",
    "url": "https://yourcompany.com"
  },
  "alumniOf": {
    "@type": "Organization",
    "name": "[Past notable employer or association]"
  },
  "sameAs": [
    "https://www.linkedin.com/in/[you]",
    "https://[your-substack-or-newsletter]",
    "https://yourcompany.com",
    "https://[conference-or-press-profile]"
  ],
  "knowsAbout": [
    "[Topic 1 — match the AI query]",
    "[Topic 2]",
    "[Topic 3]"
  ]
}
</script>

The 7-step authority-site playbook

Here is how to build a Person-centric authority site that AI tools will actually cite — in priority order.

Pick one canonical domain. One person, one authority site. Don't split your identity across multiple domains.
Publish a real About page. Long-form, written in first person, with credentials, companies, dates, and the full story of how you got here. This page is the human-readable counterpart to your Person schema.
Add the canonical Person + Organization schema. One Person block, one Organization block, both with stable @ids, referenced everywhere.
Reuse the Person @id as the Article author. Every blog post, every speaking page, every piece of media should attribute back to the same Person entity.
Build the sameAs graph. List every legitimate external profile in sameAs — and then make sure each of those external profiles links back to your authority site. Two-way edges are how AI engines confirm the entity is real.
Publish original writing under your name. Every long-form post adds another data point AI engines can attribute to you. Quantity matters less than topical consistency — pick a lane and own it.
Make the site fast, prerendered, and crawlable. AI crawlers don't run JavaScript. If your site is a single-page app, you must prerender every route to static HTML. See how to rank higher in ChatGPT for the technical detail.

Common mistakes that kill authority signals

Duplicate Person blocks. Two Person objects on the same page with different fields = the model picks one and ignores the other. Always one canonical block, referenced via @id.
Generic stock photo or no image. A real headshot hosted on your own domain is a trust signal. Stock illustrations actively hurt.
Empty sameAs. If your Person block doesn't link out to LinkedIn, your company, and your notable profiles, you've given the model nothing to verify against.
Vague knowsAbout. "Marketing" is useless. "Answer Engine Optimization for real estate agents" is a citation magnet. Match the actual query.
Treating it like a one-time setup. Authority compounds. New companies, new talks, new credentials all need to be added to your Person block as they happen.
No prerendering. If GPTBot, PerplexityBot, and ClaudeBot fetch your site and see an empty React shell, your schema doesn't exist as far as they're concerned.

How to measure if it's working

Validate the schema. Run your homepage and About page through Google's Rich Results Test and the Schema Markup Validator. Both should show your Person and Organization entities clean with zero errors.
Check raw HTML. View source (or curl the page) and confirm your Person JSON-LD is present in the initial HTML, not injected by JavaScript.
Ask the AI engines directly. Prompt ChatGPT, Perplexity, Claude, and Gemini with "Who is [Your Name]?" and "Who is the best [your expertise] expert?" Track changes monthly. You'll see your name appear, then move up, then get cited.
Watch your branded queries. Search Console impressions for your name, company, and topical phrases should climb steadily as AI Overviews surface you.

Person schema is the cheapest, highest-leverage thing you can do for AEO. It takes an hour to write, lives forever, and quietly decides who AI tools name when your buyers ask the question.

Next read: The AEO Content Checklist — the page-by-page list every authority site should pass before publishing.

Sources

[1] Schema.org. "Person type." Schema.org. View source.
[2] Google Search Central. "E-E-A-T and the Quality Rater Guidelines." Google. View source.
[3] Google Search Central. "Introduction to structured data markup." Google. View source.
[4] Schema.org. "sameAs property." Schema.org. View source.
[5] Google. "Rich Results Test." View source.

Frequently asked questions

What is Person schema and why do I need it?

Person schema is a block of structured data (JSON-LD) that tells search engines and AI tools who you are as a named entity — your name, role, employer, expertise, photo, and the other profiles that belong to you. AI tools like ChatGPT, Perplexity, and Google AI Overviews use it to verify identity and decide who to cite as an expert on a topic.

What is an authority site?

An authority site is a website built around a single person or organization that consolidates their expertise, credentials, writing, talks, and press into one canonical source. Its job is to be the place AI tools and humans go to verify who you are and what you know. selinaeizik.com is an authority site for Selina Eizik.

How does Person schema help me get cited by ChatGPT or Perplexity?

AI engines need to resolve a query like 'best AEO expert for real estate' to a real, verifiable person. Person schema gives them a structured answer with name, job title, expertise areas (knowsAbout), and external proof (sameAs links to LinkedIn, your company, speaking pages). The more clearly your entity is defined, the more often you become the answer.

Where do I put Person schema on my site?

At minimum, on your homepage and About page. It should be a single, canonical Person object referenced everywhere via @id (e.g. https://yoursite.com/#person), and reused as the author of every Article you publish. Duplicate or conflicting Person blocks confuse the entity graph.

Do I need an authority site if I already have a company website?

Yes — if you want to be cited as an individual expert. Company sites describe what the company does. Authority sites describe who the person is. AI tools recommend people for expertise queries far more often than they recommend brands, and they need a Person-centric source to do it confidently.