Quick answer: Schema markup is not a switch that makes ChatGPT or Google AI cite you. A 2026 Ahrefs study of 1,885 pages that added JSON-LD found no measurable citation lift on any major AI platform. Schema still matters for rich results, entity clarity and product and local data. Add the six types below for those reasons, then win citations with content and authority.
Scan almost any GEO checklist and “add schema” sits near the top, often with a promise that it helps AI “understand” your content. Some of that is true. The part about citations, on current evidence, isn’t. This post separates the two so you spend your developer hours where they count.
What the 2026 evidence says
- No direct citation lift. Ahrefs tracked 1,885 pages that added JSON-LD and measured citations across ChatGPT, Perplexity, Claude and Gemini. Adding schema alone made no measurable difference.
- Google needs no special AI markup. Google says there is nothing extra to add to appear in AI Overviews or AI Mode; pages need to be indexed and eligible for a snippet. Standard SEO best practice applies.
- AI reads your text, not just your tags. Language models extract answers from visible content. Markup that describes content users can’t see is ignored at best and a policy problem at worst.
Where schema still matters
Schema isn’t useless — it just works one layer down.
| Role | What it does | Why it matters for AI search |
|---|---|---|
| Rich results | Enables stars, prices, breadcrumbs, product snippets in Google | Higher CTR from classic results, which feed AI retrieval |
| Entity clarity | Tells Google who you are (Organization, sameAs, Person) | Helps Google connect your brand to the right Knowledge Graph entity |
| Product and local data | Prices, availability, address, hours | Shopping and local answers rely on accurate structured data |
| Freshness | datePublished / dateModified | Signals recency on time-sensitive topics |
In short: schema makes you eligible and unambiguous. Content and corroboration make you citable.
Six schema types worth adding
| Type | Use it for | AI-search relevance | Priority |
|---|---|---|---|
| Organization (+ sameAs, logo) | Your brand, sitewide | Entity disambiguation; links your site to your profiles | High |
| Person (author) | Named authors and reviewers | Expertise signals, especially in YMYL topics | High |
| Article / BlogPosting | Every post | Dates, author and publisher in one place | High |
| Product / Offer | Ecommerce and D2C | Prices and availability for shopping answers | High (if you sell) |
| LocalBusiness | Branches, clinics, stores | Address, hours and geo for local answers | High (if local) |
| FAQPage | Genuine Q&A sections | Rich results are now limited to a few site types; keep it for structure, not for snippets | Low |
Skip HowTo markup for rich results — Google no longer shows them in Search.
JSON-LD examples for an Indian brand
Organization (homepage or sitewide, using a fictional brand):
{
"@context": "https://schema.org",
"@type": "Organization",
"name": "Example Finserv Pvt Ltd",
"url": "https://www.example.in/",
"logo": "https://www.example.in/logo.png",
"address": {
"@type": "PostalAddress",
"addressLocality": "Noida",
"addressRegion": "Uttar Pradesh",
"postalCode": "201301",
"addressCountry": "IN"
},
"sameAs": [
"https://www.linkedin.com/company/example-finserv",
"https://x.com/examplefinserv"
]
}Article with a named author:
{
"@context": "https://schema.org",
"@type": "Article",
"headline": "Home Loan Eligibility for Salaried Employees",
"datePublished": "2026-10-01",
"dateModified": "2026-10-09",
"author": {
"@type": "Person",
"name": "Expert Marketer",
"url": "https://www.example.in/author/expert-marketer/"
},
"publisher": { "@type": "Organization", "name": "Example Finserv Pvt Ltd" }
}Schema in WordPress with Yoast
Yoast outputs a connected schema graph by default: WebSite, WebPage, Article, BreadcrumbList, and Organization or Person depending on your settings. To get the most from it:
- In Yoast → Settings → Site representation, choose Organization and add your logo and social profiles (these become sameAs).
- Complete each author’s WordPress profile, including bio and social links, so Person markup is filled.
- Use Yoast’s FAQ and How-to blocks only for real Q&A or steps on the page.
- For Product or LocalBusiness, use WooCommerce SEO or Local SEO add-ons, or a dedicated schema plugin.
Validate before you publish
- Rich Results Test — checks Google rich result eligibility.
- Schema Markup Validator (validator.schema.org) — checks general schema.org validity.
- Search Console → Enhancements — watch for errors across the site after rollout.
FAQ
Does Google AI Overviews use schema? AI Overviews draw on Google’s index, where structured data helps Google understand pages. But Google says no special markup is needed to appear, and schema alone won’t get you cited.
Is FAQ schema still worth adding? Only where the page has a genuine FAQ. Google shows FAQ rich results for a limited set of sites, so the benefit today is structure and clarity, not snippets.
What schema does ChatGPT read? OpenAI hasn’t documented using schema to choose citations. ChatGPT retrieves pages and reads the visible content, so clear text matters most.
Does Yoast add schema automatically? Yes. Yoast outputs Article, WebPage, BreadcrumbList and Organization or Person schema by default. Fill in your site representation and author profiles to make it complete.
Check your site’s schema and AI readiness with the 10-check Universal Search audit.