# FAQ schema
JSON-LD that exposes a page's Q&A pairs as extractable structured data. Now AI-side load-bearing — Google deprecated the Search rich result in May 2026.
By AgentSite · 3 min read · Updated 2026-05-23
FAQ schema is JSON-LD that marks a page's question-and-answer pairs as machine-readable `FAQPage` structured data. Answer engines extract those Q&A pairs verbatim into "what is X" responses. Google deprecated the FAQ rich result in Search as of May 7, 2026; the schema's remaining value is on the AI side, where the extraction surface is largely the same.
## What it looks like
```json
{
"@context": "https://schema.org",
"@type": "FAQPage",
"mainEntity": [{
"@type": "Question",
"name": "What is FAQ schema?",
"acceptedAnswer": {
"@type": "Answer",
"text": "FAQ schema is JSON-LD that marks a page's question-and-answer pairs as machine-readable FAQPage structured data."
}
}]
}
```
The page itself carries the same question and answer in human-readable HTML; the schema is a parallel surface for the same content.
## What the spec says
[Schema.org defines FAQPage](https://schema.org/FAQPage) as "A WebPage presenting one or more 'Frequently asked questions'." The minimum useful shape is `FAQPage.mainEntity` as an array of `Question` objects, each with a `name` (the question) and an `acceptedAnswer.text` (the answer). Other properties exist on the parent `WebPage` and `Thing` types; most agents only read the question-and-answer pair.
## Where it's still load-bearing
Google's documentation states bluntly that "as of May 7, 2026, FAQ rich results are no longer appearing in Google Search" ([Google FAQPage structured data docs](https://developers.google.com/search/docs/appearance/structured-data/faqpage)). The schema is now restricted to government and health sites for any Search rich-result purpose at all.
The AI side has not deprecated anything. Generative engines that read structured data still lift FAQ pairs as a preferred quote-and-attribute surface, and the Princeton GEO paper measured up to 40% visibility lift from optimization tactics rooted in extractable-structured content ([Aggarwal et al., KDD 2024](https://arxiv.org/abs/2311.09735)). For an AEO-focused site, FAQ schema is now an AI-side investment with effectively no Search-side return.
## Anti-patterns
Google's content guidelines for the schema remain in force even with the rich-result deprecated, and they happen to be AEO-aligned:
1. **One answer per question.** Multiple-choice FAQ blocks are rejected. Agents extract a single quote per question.
2. **All Q&A visible on the page.** Hidden or accordion-collapsed FAQ content that requires interaction to reveal is excluded. Agents do not interact.
3. **No repeating FAQ content across multiple pages.** Each FAQ should be unique to its page. Agents that deduplicate across the site drop boilerplate FAQ schema entirely.
## Where this fits
FAQ schema is a Layer 3 dimension — the page's "what kind of thing is this" declaration. It pairs naturally with a [direct answer](/direct-answer) lede that often serves as the implicit Question of the page itself. The longer thesis on why the structured-data layer matters at all is in [agent readability](/agent-readability), and the dimension's place in the layered model is in [the five layers of AEO](/five-layer-aeo).