Schema markup for AI search: what helps and what does not

The short answer

Google's documentation on AI features is direct: there are no additional requirements to appear in AI Overviews or AI Mode, no special optimizations, and no special schema.org structured data you need to add. To be shown as a supporting link, a page must be indexed and eligible to appear in Search with a snippet. Google also lists "making sure your structured data matches the visible text on the page" among ordinary good practices.

Google's guide to optimizing for generative AI features, added in May 2026, goes further. It lists "overfocusing on structured data" under things you don't need to do: structured data isn't required for generative AI search, and there's no special schema.org markup to add. The same guide says it's still a good idea to keep using it as part of your overall SEO.

So structured data is not an AI ranking lever. It is a way to state facts precisely. If those facts are wrong, missing from the page or invented, the markup works against you.

What structured data still does

Google says it uses structured data it finds on the web to understand the content of the page. In practice that means:

• Entity facts: your organization's name, URL, logo and official profiles stated once, unambiguously.
• Page facts: whether a page is a product, an article or a local business, with its price, author or opening hours.
• Freshness: datePublished and dateModified that agree with the visible date.
• Search features: eligibility for the rich results Google still supports, when all required properties are present.

Other engines do not document how they use markup. Treat any effect there as unknown, and make sure the same facts are readable in the visible text.

What the evidence says

The best public test we know of points the same way as Google's guidance. In a study published on 2026-05-11, Ahrefs followed 1,885 pages that added JSON-LD between August 2025 and March 2026 and compared their AI citations in the 30 days before and after with matched control pages:

Ahrefs concluded that adding schema produced no major uplift in citations on any platform. The study has limits that its authors state: every page was already heavily cited before the change, schema often arrived together with other site changes, and the window was only 30 days. It says little about a page that AI systems have not picked up yet.

The same article reports a separate searchVIU test in which five AI systems fetched pages live and extracted only the visible HTML, ignoring JSON-LD and hidden Microdata or RDFa. Ahrefs gives no methodology for that test, so treat it as a hint rather than proof. Both results support the practical rule on this page: any fact you want an engine to use belongs in the visible text first.

Types worth adding

Use a type only when it describes what a person can see on the page.

Sector-specific types such as LegalService, Dentist or Hotel are covered in our industry playbooks.

What to skip or stop expecting

• FAQ rich results. Google stopped showing them in Search on May 7, 2026, and removed the documentation in June 2026. Existing FAQPage markup that matches visible questions does no harm, but new FAQ markup will not earn a rich result.
• "AI schema" or special AI tags. Google says none is required. Be wary of plugins that promise one.
• Markup for invisible content. Google's guidelines say not to mark up content that is not visible to readers. Reviews, ratings and prices in JSON-LD must appear on the page.
• Self-serving claims. Awards, rankings or "best" labels in markup that the page cannot support do not become true by being structured.

A minimal starting point

Most sites can start with an Organization and WebSite block on the home page:

<script type="application/ld+json">
[
  {
    "@context": "https://schema.org",
    "@type": "Organization",
    "name": "Your brand",
    "url": "https://example.com/",
    "logo": "https://example.com/logo.png",
    "sameAs": ["https://www.linkedin.com/company/your-brand/"]
  },
  {
    "@context": "https://schema.org",
    "@type": "WebSite",
    "name": "Your brand",
    "url": "https://example.com/"
  }
]
</script>

Replace every value with a verified fact, and only list profiles you control.

How to check your markup

1. Validate the syntax with the Schema Markup Validator, and check rich-result eligibility with Google's Rich Results Test.
2. Compare against the page. Every name, price, date and rating in the markup should be visible in the text.
3. Put it in the server HTML. Google can read JSON-LD injected by JavaScript, but a crawler that does not run JavaScript cannot. Server-rendered markup works for both.
4. Recheck after template changes. Theme and plugin updates often duplicate or drop structured data.

The free AI visibility checker reports whether a page's JSON-LD parses and which types it declares, as one of its 26 GEO audit checks.

Sources

• Google Search Central: AI features and your website
• Google Search Central: Guide to optimizing for generative AI features
• Ahrefs: We tracked 1,885 pages adding schema. AI citations barely moved.
• Google Search Central: Introduction to structured data markup
• Google Search Central: General structured data guidelines
• Google Search Central: Latest documentation updates (FAQ rich result removal)
• Schema.org: Schema Markup Validator
• Google: Rich Results Test

Reviewed on 2026-10-23. Search features change; check Google's documentation updates before relying on a rich result type.