FAQPage Schema: The Highest-ROI Signal for AI Overviews

TL;DR: Google stopped showing FAQ rich results on May 7, 2026, but FAQPage schema is still worth keeping because it hands AI Overviews and chatbots clean, pre-parsed question-answer pairs to cite.

If you read the May 2026 headlines and reached for the delete key, stop. Google retired the FAQ rich result - the visible accordion in the blue links - not the markup itself. The schema that lost its SERP slot is the same schema answer engines now lean on to pull citable answers, and ripping it out trades a dead feature for a live one.

What is FAQPage schema, and did Google really kill it?

FAQPage is a schema.org type that labels a block of question-and-answer content so machines can read it without parsing your layout. At minimum it needs a mainEntity array holding one or more Question items, each with a name (the full question) and an acceptedAnswer whose text carries the full answer.

What Google retired is the rich result, on a published timeline. Per Google Search Central: as of May 7, 2026 FAQ rich results no longer appear in Search; the FAQ search appearance, rich-result report, and Rich Results Test support are dropped in June 2026; and Search Console API support for the FAQ rich result is removed in August 2026. The doc is explicit that the FAQPage type itself stays valid - the markup can live on your pages without causing problems.

Why did Google deprecate the FAQ rich result?

The accordion had been fading for years. Since 2023, Google restricted the FAQ rich result to "well-known, authoritative websites that are government-focused or health-focused." If you run a SaaS blog, a docs site, or a marketing page, you almost certainly had no visible FAQ rich result before May 2026 either - so the 2026 change took away something you had already lost. That is the part the panic posts skip: for most sites, nothing on the page changed.

What did change is the framing. Google moved FAQ markup out of the "earn a SERP feature" column and into the "structured signal we read" column. That is a downgrade for snippet-chasers and an upgrade for anyone optimizing for answer engines.

Why is FAQPage schema the highest-ROI signal for AI Overviews?

Answer engines - Google's AI Overviews, ChatGPT, Perplexity, Claude - work by retrieving a page and extracting the span that answers a query. A wall of prose makes them guess where the answer starts and ends. A FAQPage block hands them the boundary already drawn: this string is the question, this string is the accepted answer. There is no ambiguity to resolve and nothing to summarize away.

That matters because extraction confidence is one of the signals AI assistants weigh when deciding what to cite. A clearly delimited answer is cheaper to lift and safer to attribute than a paragraph the model has to compress. The same markup that lost its blue-link visibility became a stronger citation hook - the value moved, it did not disappear. For the broader playbook, our GEO and AEO checklist covers where FAQPage fits among the other on-page levers.

When should you keep FAQPage schema, and when not?

• Keep it on pages with genuine question-and-answer content that mirrors how people phrase searches.
• Keep it on documentation, pricing, and troubleshooting pages where the answer is short and self-contained.
• Drop it on pages where you invented questions just to hold the markup - thin or padded Q&A reads as manipulation to both Google and answer engines.
• Never wrap promotional copy in a Question to smuggle keywords; the markup is for answers, not slogans.

There is no penalty for leaving valid FAQPage markup in place, so the only reason to remove it is if the Q&A content was fake to begin with - in which case the content was the problem, not the schema.

How do you add FAQPage schema correctly?

The cleanest source is a real FAQ section on the page, with each question as a heading and each answer as the text directly beneath it. Generate the JSON-LD from that block so the markup and the visible content always match - mismatched schema is the fastest way to lose trust with a crawler. Most modern CMS platforms emit FAQPage automatically from an FAQ block; Ghost, for example, builds it from the post's question headings, so the markup stays in sync without hand-editing. If you want to confirm a page is actually exposing answer-ready content, a free AI SEO checker will show you what an answer engine sees.

FAQ

Is FAQPage schema deprecated in 2026?

No. Google deprecated the FAQ rich result, not the FAQPage schema type. The markup is still a valid schema.org type and Google still crawls and reads it.

Should I remove FAQ schema from my site?

No. There is no penalty for keeping valid FAQPage markup, and it remains a useful extraction signal for AI Overviews and chatbots. Only remove it if the Q&A content was padding.

Does FAQPage schema help with ChatGPT and Perplexity citations?

It helps because it delivers pre-structured question-answer pairs that are easy to extract and attribute. It is not a guaranteed citation, but it lowers the cost of citing your page versus a competitor's prose.

What is the difference between FAQPage and HowTo schema now?

FAQPage marks up question-and-answer pairs; HowTo marks up ordered steps to complete a task. Google also deprecated the HowTo rich result, but like FAQPage the type stays valid and useful for answer extraction.

Will Google penalize me for keeping FAQ structured data?

No. Google's documentation states the type stays valid and the markup can remain on your pages. Penalties only come from spammy or mismatched markup, not from valid FAQPage data.