FAQ Schema Won't Get You Rich Results Anymore. Add It Anyway.

Missing FAQ schema means you’ve got a block of question-and-answer content on a page, but no FAQPage structured data telling machines that’s what it is. It’s worth flagging because the page is doing the work of an FAQ without labeling itself as one.

Before you add it, know one thing: the old reason to add FAQ schema is dead. Google stopped showing FAQ rich results in Search on May 7, 2026, finishing a pullback that started in 2023. So if you’re doing this hoping for those expandable Q&A snippets under your listing, don’t. They’re gone. The reason to do it now is different, and it’s worth understanding first.

Why this still matters for AI traffic

Getting cited by ChatGPT, Perplexity, or Google’s AI Mode is a retrieval game. These systems break a question into parts, pull the chunks that answer each part, and stitch them together. A clean question-and-answer pair is the most extractable unit you can hand them: one query, one answer, exactly the shape an AI engine wants.

The SEO world oversells the schema’s role, so let me be straight about it. Some tools, Frase among them, claim FAQ schema directly drives AI citations (they cite a 3.2x lift in AI Overviews). I don’t buy that the markup itself is the driver. Testing by Andrew Williams-Cook (via ZipTie) found that large language models tokenize your JSON-LD as raw text rather than parsing it as structured data. That’s one experiment from limited public testing, not a Google-confirmed law of physics, but it lines up with how these models actually work. What gets you cited is the visible, well-formatted Q&A content on the page, the stuff a human reads.

So the markup’s value is indirect. It feeds Google’s traditional index and Knowledge Graph cleanly, and it forces your page into a tight question-answer structure in the first place. That structure is the structure AI engines reward. That’s the case for adding it: a supporting signal, not a magic switch.

What the markup has to do

Valid FAQPage JSON-LD wraps the FAQ content already on the page. Each question becomes a Question with an acceptedAnswer, using the text a reader can see. Don’t invent questions or answers for the markup. Mark up what’s there.

If the Q&A is buried in loose prose, tighten it into clean pairs first, so the markup is honest and the content stays extractable. The markup and the visible text have to match. That’s a Google requirement, not a nice-to-have, and it matters: structured data that describes content not visible on the page can be flagged as spam.

That restructuring is worth more than the markup, by the way. The answer-first, one-question-one-answer format is what both Google and AI engines pull from.

When it’s worth adding, and when to skip

Add it when:

• The page has genuine FAQ content you want AI engines to cite: real questions buyers ask, with real answers.
• The page matters for AI visibility: a product page, a service explainer, a doc that answers common questions.

Skip or deprioritize this when:

• The “FAQ” is marketing fluff: invented questions nobody asks, padded to look thorough. Marking up filler doesn’t help, and the structure just makes the filler more obvious.
• The page barely has Q&A content. Schema can’t fix thin content. If there’s nothing substantive to mark up, fix the content first, then add the schema.

The effort here is low and the downside is near zero, but the upside is modest too. Google has said that leaving in FAQ markup it no longer uses for rich results won’t cause problems: it won’t produce visible results in Search, but it won’t hurt you either. So adding valid FAQ schema to real FAQ content is a safe, cheap bet. Just don’t expect it to move traffic on its own. The visible content does the heavy lifting.

Tightening the Q&A is where the value is. Wrapping it, once it’s tight, is the cheap part: Get AI Traffic applies the markup with upsert_schema, which stores one block per schema type, so a re-run replaces your FAQPage markup instead of stacking a second copy on the page.

Frequently asked questions

Does FAQ schema still get rich results in Google? No. Google stopped showing FAQ rich results on May 7, 2026, and is removing the related reporting tools through 2026. As of that date the rich result is gone for everyone. (Government and health sites kept it during the 2023 to early-2026 pullback, but the May 2026 change eliminated even that carve-out.)

If the rich result is gone, why add FAQ schema at all? Because it feeds Google’s index and Knowledge Graph cleanly, and it forces a tight question-answer structure on the page. That structure is what AI engines extract and cite. The schema is a supporting signal; the visible Q&A content is the real driver.

Does FAQ schema directly make ChatGPT or Perplexity cite me? Not directly. Public testing suggests LLMs read your JSON-LD as plain text, not as structured data. They cite the visible Q&A content on the page. The schema helps indirectly, through Google’s pipeline.

Should I remove FAQ schema I already have? No. Google’s documentation says existing FAQ structured data can stay in place: it won’t cause problems, and it won’t produce visible results in Search. Removing it gains you nothing, as long as the markup still matches your visible content.

Sources

• Google Search Central: Mark up FAQs with structured data
• Search Engine Journal: Google Drops FAQ Rich Results From Search
• ZipTie: FAQ Schema for AI Answers, does it actually get you cited?
• Frase: Are FAQ Schemas Important for AI Search, GEO and AEO? (the opposing 3.2x-lift view this article disputes)