FAQ Schema: Markup Example and What Google Changed

FAQ schema is structured data, written with the schema.org FAQPage type, that marks up a list of questions and their answers on a page so machines can read them as pairs. For years it earned expandable FAQ rich results in Google. That is over: Google limited FAQ rich results to government and health sites in 2023, and its own changelog says the feature stopped appearing in Google Search on May 7, 2026. Below is a working FAQ page schema example, what the changes mean, and when the markup is still worth adding.

What is FAQ schema?

FAQ schema is the common name for FAQPage markup from schema.org. schema.org defines a FAQPage as a web page presenting one or more frequently asked questions. Each question is a Question item with a name (the question text) and an acceptedAnswer, which is an Answer with a text property.

You usually write it as JSON-LD: a block of JSON inside a <script type="application/ld+json"> tag. JSON-LD is separate from the visible HTML, so you can add it without touching the page layout.

Two related types get confused with it:

• QAPage is for a page built around one question with answers from users, like a forum thread. FAQPage is for questions and answers written by the site owner.
• HowTo marks up step-by-step instructions. Google deprecated HowTo rich results in September 2023.

FAQ page schema example (JSON-LD)

Here is a complete FAQPage block for a page with three questions. Copy it, replace the text, and keep each answer identical to what the page shows.

{
  "@context": "https://schema.org",
  "@type": "FAQPage",
  "mainEntity": [
    {
      "@type": "Question",
      "name": "How long does shipping take?",
      "acceptedAnswer": {
        "@type": "Answer",
        "text": "Orders ship within 1 business day. Delivery in the US takes 3 to 5 business days."
      }
    },
    {
      "@type": "Question",
      "name": "Can I return an item?",
      "acceptedAnswer": {
        "@type": "Answer",
        "text": "Yes. Unused items can be returned within 30 days for a full refund. <a href=\"https://example.com/returns\">Start a return</a>."
      }
    },
    {
      "@type": "Question",
      "name": "Do you ship outside the US?",
      "acceptedAnswer": {
        "@type": "Answer",
        "text": "We ship to Canada, the UK and the EU. Duties are calculated at checkout."
      }
    }
  ]
}

Wrap it in a script tag in the page HTML:

<script type="application/ld+json">
  { "@context": "https://schema.org", "@type": "FAQPage", "mainEntity": [ ... ] }
</script>

Notes on the example:

• mainEntity is an array, one Question per FAQ item.
• name holds the full question text. acceptedAnswer.text holds the full answer.
• Answer text may include basic HTML such as links, as in the second item. Keep it short and simple.
• One FAQPage block per page. If the page has 12 questions, put all 12 in one mainEntity array.

If you don't want to write JSON by hand, our FAQ schema generator turns a list of questions and answers into this block. Then run the live page through our schema validator or the Schema Markup Validator to catch syntax errors.

What did Google change about FAQ schema?

There were two changes, three years apart.

August 2023: FAQ rich results limited to government and health sites

On August 8, 2023, Google published changes to HowTo and FAQ rich results. FAQ rich results from FAQPage markup would from then on only be shown for well-known, authoritative government and health websites. For every other site, the rich result would no longer be shown regularly. HowTo results were first limited to desktop and then deprecated on September 13, 2023.

The same post answered the question every site owner asked: you could drop the markup, but there was no need to proactively remove it. Structured data that isn't used doesn't cause problems for Search, but also has no visible effect.

May–June 2026: FAQ rich results removed entirely

Google's Search Central documentation changelog has two entries that finish the story:

• May 8, 2026: Google added a deprecation notice to the FAQ rich result documentation, noting the feature would no longer appear in Google Search starting May 7, 2026.
• June 2026: Google removed the FAQ rich result documentation, because the feature is no longer shown in Google Search results.

So as of September 2026, "google faq schema" has no Google-specific meaning left. There is no FAQ rich result to be eligible for, even for government and health sites, and the old FAQ documentation URL now redirects to the changelog.

What Google says about structured data and AI features

Google's AI optimization guide addresses this head-on in its list of things you can ignore. It says structured data isn't required for generative AI search and there is no special schema.org markup to add, while still recommending structured data as part of normal SEO because it makes pages eligible for rich results.

Put plainly: FAQPage markup gets you nothing visible in Google today, and Google says it isn't a lever for AI Overviews or AI Mode either.

Is FAQ schema still worth adding?

It depends on which system you care about and how much the markup costs you to maintain.

Goal	Does FAQPage markup help? (as of September 2026)
FAQ rich results in Google	No. The feature was removed in May 2026.
Google AI Overviews / AI Mode	Google says no special markup is needed.
Other AI engines (ChatGPT, Perplexity, Claude)	Unproven. No engine has published that it uses it.
Other consumers of schema.org (some search engines, site search, internal tools)	Possibly, depending on the system.
Keeping Q&A content well organized	Indirectly: writing the markup forces you to write clean pairs.

The case for keeping it (reasoning, not proof)

We want to be careful here, because much of what is written about FAQ schema and AI is guesswork presented as fact. Here is the reasoning, labeled as reasoning:

• It is an unambiguous format. A Question with an acceptedAnswer states exactly which text answers which question. An AI retrieval system that parses JSON-LD gets clean pairs with no layout noise. Whether any given engine parses it is not public.
• It costs little if generated. If your CMS or template produces FAQPage markup from the same source as the visible FAQ, the maintenance cost is close to zero.
• It does no harm in Google. Google said in 2023 that unused structured data causes no problems for Search.

What we can say with more confidence is that the visible question-and-answer content matters. AI engines retrieve and quote page text. Research on generative engines points the same way: the GEO paper (Aggarwal et al., KDD 2024) found that content changes such as adding quotations, statistics and cited sources raised a source's visibility in generated answers. That work was about page content, not markup.

The case against

• Duplicate maintenance. Hand-written JSON-LD drifts out of sync with the page. Markup that says one thing while the page says another is worse than no markup.
• Opportunity cost. An hour spent on FAQ markup is an hour not spent writing a better answer, which is what both Google and AI engines read.
• Bloat. Some plugins add FAQPage markup to every page, including pages with no FAQ. Clean that up.

Our recommendation: keep FAQPage markup where it is generated automatically from a real, visible FAQ. Don't hand-build it for pages that have no FAQ, and don't expect a ranking or citation lift from it.

How to write FAQ content that AI engines can use

Since the visible text is what counts, these rules matter more than the markup:

1. Use real questions. Pull them from Search Console queries, support tickets and sales calls. Our AI prompt generator suggests the questions buyers ask AI engines.
2. Answer in the first sentence. "Orders ship within 1 business day." beats "Great question! Shipping depends on many factors…". Retrieval systems often lift the first sentence under a question.
3. Say the number. Prices, time frames, limits, compatibility. Specific answers are easier to quote.
4. Keep each answer self-contained. An answer that says "see above" is useless when quoted out of context.
5. Make questions visible headings or clear labels. An AI crawler reading raw HTML should see which text is the question. Check with our AI crawler view, which shows the page without JavaScript. FAQ accordions that load their content by script can be invisible to some bots.
6. Don't hide answers behind clicks that require JavaScript to fetch. Collapsed-but-present HTML is fine; content fetched only on click may never be seen.

For the wider picture of structuring pages for AI answers, see our guide on how to optimize content for AI search.

FAQ schema markup rules that still apply

Google's FAQ rules are gone with the feature, but schema.org and common sense still apply:

• Mark up only questions and answers that are visible on the page.
• Use FAQPage for owner-written Q&A. Use QAPage for user-generated answers.
• One answer per question in acceptedAnswer.
• No advertising copy dressed up as an FAQ.
• Keep the JSON valid. One missing comma breaks the whole block, and most parsers then ignore it entirely.
• Serve the JSON-LD in the initial HTML if you want non-rendering crawlers to see it.

Common errors to check

• Invalid JSON: trailing commas, curly quotes pasted from a word processor, unescaped double quotes inside answer text.
• Wrong nesting: acceptedAnswer placed outside the Question, or mainEntity as an object when there are several questions.
• Mismatch: the markup answer differs from the visible answer after a copy edit.
• Duplicates: two FAQPage blocks on one page, often one from a theme and one from a plugin.

Our schema validator lists every JSON-LD block on a URL, which makes duplicate blocks easy to spot.

FAQ schema checklist

• Decide whether the page has a real, visible FAQ. If not, skip the markup.
• Write answers that lead with the fact, in plain text.
• Generate the JSON-LD with the FAQ schema generator or from your CMS.
• Validate the live URL with the schema validator.
• Confirm the questions and answers appear in the raw HTML with the AI crawler view.
• Stop reporting "FAQ rich result" as a goal. Track whether AI engines cite the page instead.

If you want to know whether AI engines quote your FAQ answers, run a free check with the AI visibility checker.