SEO & AI
FAQPage schema: the markup that helps AI cite you
AI assistants like short, structured answers. FAQPage markup in JSON-LD presents your questions and answers in a format machines read without ambiguity — and it's one of the simplest to add to a page.
When an AI assistant builds an answer, it looks for short, self-contained passages that are easy to extract. A page that explicitly states "here's a question, here's its answer" does the work for it. schema.org's FAQPage markup does exactly that — in JSON-LD, in a format machines read without interpreting your layout.
What FAQPage markup is
It's a block of structured data describing a list of questions and their answers. You add it to your page as a JSON-LD script; it mirrors, in machine language, what the user already sees on screen.
What it really does (and doesn't)
Let's be honest about expectations. What FAQPage no longer does: since 2023, Google has heavily restricted FAQ rich results in its results — reserved for a few categories (governments, health). So don't expect a systematic dropdown under your link.
What FAQPage still does: the markup is still read by search engines and, above all, by systems that parse JSON-LD to understand a page — including AI. It turns a block of prose into unambiguous question/answer pairs, exactly the format an assistant likes to extract and cite.
How to write it
<script type="application/ld+json">
{
"@context": "https://schema.org",
"@type": "FAQPage",
"mainEntity": [
{
"@type": "Question",
"name": "Your question, as it is asked?",
"acceptedAnswer": {
"@type": "Answer",
"text": "The answer — identical to the text visible on the page."
}
}
]
}
</script>
Question object for each entry. The answer's text must match the content actually displayed.The rules to follow
- Visibility — every marked-up Q&A must exist, visible, on the page. No hidden content reserved for the JSON-LD.
- One FAQPage per page, reserved for genuine frequent questions (not a marketing catch-all).
- Self-contained answers — an answer that makes sense out of context is easier to cite.
- Consistency — the markup text must match, word for word or nearly, the visible text.
Beyond FAQPage
Markup is just one brick. The signals AI seems to weigh most come from the writing: a synthetic answer in the first paragraph, short paragraphs, H2s phrased as questions, visible publication and update dates. This very page, for instance, carries its own FAQPage markup — proof by example.
FAQPage doesn't magically "boost" your positions. It makes your answers machine-readable, without ambiguity. Add it to your high-intent pages, keep it honest (real, visible Q&A), and combine it with clean HTML and allowed AI crawlers.
Verify
After adding it, validate your markup with Google's Rich Results Test, and check your overall setup with our free AI-visibility checker. For the big picture, start by knowing whether ChatGPT already cites you.
Does FAQPage schema improve SEO?
Not directly: since 2023, Google reserves FAQ rich results for a small number of sites (governments, health). But the markup is still read by search engines and by AI that parses the JSON-LD; it structures your answers without ambiguity.
Where do I place FAQPage markup?
In a <script type="application/ld+json"> block on the page, ideally in the head or at the end of the body. The marked-up questions and answers must match content actually visible on the page.
Can I mark up Q&A not visible on screen?
No. schema.org and Google rules require the marked-up content to be visible to the user. Content hidden only in the JSON-LD is a bad practice that can be penalised.
Is FAQPage enough to be cited by AI?
No. It's a useful structure signal, not a magic lever. It adds to fundamentals: allowed AI crawlers, clean HTML, a synthetic answer in the first paragraph and content worth citing.
Published June 2026. AI bot names and behaviours change fast; check the publishers' documentation before freezing a configuration. General information, not individualised advice.