Schema markup: definition and what the evidence shows

99Visibility Research · Updated October 21, 2026 · 3 min read

Short answer

Schema markup is code that labels the things on a web page using the schema.org vocabulary, so machines can read them as typed data. Most sites add it as a JSON-LD block. Google says structured data isn't required for generative AI search. Ahrefs' matched test found adding schema produced no major uplift in AI citations, and AI Overview citations fell 4.6%.

Key takeaways

Schema markup is schema.org vocabulary added to a page, usually as JSON-LD, describing entities like an organization, product or article.
Google states structured data isn't required to appear in AI Overviews or AI Mode.
Ahrefs tracked 1,885 pages adding schema and found no major citation uplift on any platform, with a statistically significant 4.6% drop in AI Overview citations.
Every page in that test already had 100+ AI Overview citations, so whether schema helps an uncited page enter AI answers is untested, not refuted.
Existing markup does no harm, so the honest position is: keep it if it serves other purposes, but don't treat it as a lever for AI answers.

What schema markup is

Schema markup is structured data written in the schema.org vocabulary. It tells a machine what a piece of content is: an Organization, a Product, an Article, an Event. Each type has defined properties, such as name, price or datePublished.

The markup describes content that is already on the page. It doesn't add new content for readers. For a deeper look at the AI question specifically, read our guide on whether schema markup helps AI citations.

How it works

Most sites add schema as a JSON-LD script in the page's HTML. Microdata and RDFa embed the same vocabulary in HTML attributes instead. A minimal JSON-LD block looks like this:

<script type="application/ld+json">
{"@context": "https://schema.org", "@type": "Organization", "name": "Example Co", "url": "https://example.com"}
</script>

Search engines read the block when they crawl the page. Google uses some types to power rich results in classic search. The markup should match what the visible page says. A JSON-LD checker can show where the two disagree.

Why it matters for AI visibility

The evidence says schema is not a lever for AI answers. Google's AI features documentation, as of October 2026, says a page needs to be indexed and eligible for a snippet to appear in AI Overviews or AI Mode, with no additional technical requirements. Its optimization guide adds that you don't need new markup to appear in Google's generative features.

Ahrefs ran the best test available. It tracked 1,885 pages that added schema, each matched against three control pages. Ahrefs found that adding schema "produced no major uplift in citations on any platform." AI Overview citations fell 4.6%, a statistically significant drop. AI Mode rose 2.4% and ChatGPT 2.2%.

The limit matters. Every treated page already had 100+ AI Overview citations before schema was added. So the test says nothing about pages that aren't cited yet. That question is untested, not settled either way.

What does hold up: stating a specific fact (a price, a number, a one-line definition) in fetchable page text gives an engine a sentence it can quote. See structured data vs plain-text facts for the comparison.

Common misconceptions

"Add FAQ schema to get cited." Google deprecated FAQ rich results in May 2026 and removed the documentation in June 2026. No AI-specific schema type exists.
"Schema is a ranking factor for AI." No engine documents this, and the matched test found no lift. See what holds up for AI Overviews ranking signals.
"Schema is useless, so remove it." Also unsupported. Google says existing markup does no harm. Read whether removing schema markup hurts AI visibility.
"GEO studies prove markup works." The critical survey of 45 GEO studies found no technique with a stable cross-platform effect on discoverability.

Frequently asked questions

Does schema markup help ChatGPT or Perplexity cite my site?

No engine documents using schema as a citation signal. In Ahrefs' matched test, ChatGPT citations rose 2.2% after pages added schema, which Ahrefs did not treat as a major uplift. The test only covered pages already heavily cited. For ChatGPT, the documented access levers are allowing OAI-SearchBot in robots.txt and making sure your firewall lets it through.

Should I still add schema markup to new pages?

Add it if it serves a purpose you can name, such as eligibility for a rich result type Google still supports. Don't add it expecting AI citations. Spend that effort first on access: content in the HTML before JavaScript runs, crawlers allowed in robots.txt, and pages indexed and snippet-eligible.

What is the difference between schema.org and JSON-LD?

Schema.org is the vocabulary: the list of types and properties, like Organization and name. JSON-LD is one format for writing that vocabulary into a page, as a script block. You can express the same schema.org data in Microdata or RDFa instead. Google recommends JSON-LD because it sits apart from the visible HTML and is easier to maintain.

Why did AI Overview citations drop after adding schema?

Ahrefs reported the 4.6% drop as statistically significant but did not establish a cause. It's a measured result on one population: pages that already had 100+ AI Overview citations. Treat it as evidence that schema didn't help those pages, not as proof that schema harms visibility in general.

Sources

This page is AI-assisted. It was written from a verified evidence base and checked against our editorial policy. Engine behaviour changes often; facts are as of October 21, 2026.