Google Removed FAQ Rich Results. Should You Keep FAQPage Schema?

From its 2019 launch to 2026, FAQPage schema was the cheapest rich result on the web: add a few questions in JSON-LD and your listing grew. Google narrowed it in 2023 and ended it in 2026. A lot of sites still carry the markup, and a lot of advice still recommends adding it. Here is what Google actually recorded, and what follows.

What Google recorded, in its own words

The three entries below are from Google’s structured data changelog. Together they are the whole story.

Read in order: the feature was restricted to a narrow set of government and health sites in September 2023, so for almost every business it had already stopped producing anything. The May 2026 entry ended it for the remaining sites, and the June 2026 entry, titled “Removing documentation for the FAQ rich result feature,” took the page down. There is no longer a Google document describing FAQ rich result eligibility, because there is no longer a FAQ rich result.

What that changes, and what it does not

Two things were always bundled under “FAQ schema”: the markup (a FAQPage node in JSON-LD) and the content (a question a reader can see, with an answer under it). The rich result rewarded the markup. Everything else rewards the content.

Google’s AI features documentation says there is “no special schema.org structured data that you need to add” to appear in AI Overviews or AI Mode, and that a page needs to be indexed and eligible for a snippet. Its May 2025 guidance on AI search adds the rule that decides what to keep: if you use structured data, make sure that “all the content in your markup is also visible on your web page.”

On the assistant side, the mechanics point the same way. Anthropic’s web search documentation says Claude’s citations carry “up to 150 characters of the cited content.” That is one plain sentence. A question-shaped heading with a direct one-sentence answer beneath it is exactly the unit that fits, which is why visible FAQ content keeps earning citations after the rich result that once decorated it has gone. How Claude picks and quotes sources is in a separate post; the Google AI Mode version, where one question becomes many sub-queries, is in another.

What to do with the FAQ blocks you already have

1. Keep every visible question and answer. If a reader can see it and it answers something customers ask, it is doing its job for snippets, People Also Ask and assistant citations. Tighten the first sentence of each answer so it stands on its own.
2. Keep FAQPage markup only where it mirrors that visible text. Same questions, same answers, same page. It no longer earns a rich result, but it is harmless and gives a parser a clean structured copy of what the page says.
3. Remove markup-only FAQs. A FAQPage node whose questions do not appear on the page was outside Google’s guidelines when the rich result existed and is pure liability now: it is structured data describing content that is not there.
4. Do not stuff. Twelve near-duplicate questions written to catch phrasings add nothing an assistant will quote. Three real questions with one-sentence answers beat them.
5. Validate what remains. Our schema validator fetches a live URL and reports parse errors, duplicate ids and dangling references, and the schema generator writes a clean FAQPage block when you do want one.

Where FAQ content fits now

The honest summary is that the acronym outlived the feature. The practice people called FAQ schema was always two practices, and the one that survives is the content one: answer engine optimization as the shaping of a page so a direct answer can be lifted out of it. Our AEO service does that work on the pages that matter, and the AI Search Playbook sets out where quotable content sits among the four things every assistant needs from a site.

Frequently asked questions

About the work

Webvello is an AI search optimization and index engineering company founded in 2024, based in Roseville, California. We’ve delivered 50+ projects across 37+ cities, with average client ROI in the 3–5x range on properties where the technical foundation was the binding constraint. Founder George Shvaya came to search from cybersecurity and systems engineering, which is why this post quotes the changelog rather than the folklore that grew up around FAQ schema.

Sources: the three dated entries are quoted from Google’s structured data changelog; the visibility rule is from Google’s AI features documentation and its May 21, 2025 Search Central post; the citation length is from Anthropic’s web search tool documentation. All re-read on October 16, 2026. The FAQ page on this site mirrors its visible questions, which is the standard described above.