FAQ Schema in 2026: Does It Still Help Google or ChatGPT?

Search "faq schema" and you'll land on a stack of near-identical how-to guides. The current top ten is almost all generic implementation walkthroughs — Neil Patel, Schema App, SEOClarity, Rank Math — sitting next to the official Schema.org and Google QAPage docs. Every one of them shows you the JSON-LD and stops there.

None of them answers the question you actually have: is this still worth my time? That's the one worth answering, because the ground moved under this topic in 2026 and most of those guides haven't caught up.

The quick answer: is FAQ schema still worth it?

Here's the whole decision in three lines, before the evidence:

• Yes, add it if your platform makes it a two-minute, no-code job — WordPress with Rank Math or Yoast, Webflow with its built-in generator. It's basically free and it doesn't hurt.
• No, don't expect it to earn you a rich result in Google or a citation in ChatGPT. Both of those doors are, for practical purposes, closed.
• Do it for the content — real questions and answers your readers can actually see — not for the markup wrapper. That's the half that still pays.

The rest of this piece is why. If you just wanted the call, you have it.

What actually happened to FAQ rich results in Google Search

Most people typing "faq schema" into Google don't know the feature they're chasing is gone. Here's the timeline, because it changes the whole math.

2023: restricted to government and health sites

The first cut came on August 8, 2023. Google announced on the Search Central blog that it was "reducing the visibility of FAQ rich results," and that going forward they would "only be shown for well-known, authoritative government and health websites." For everyone else — which is nearly every small business — the dropdown "would no longer be shown regularly."

Google did offer one reassurance in the same post: structured data that isn't being used "does not cause problems for Search," and there was no need to proactively rip it out. John Mueller framed the pullback as a "tragedy of the commons" problem. He also told site owners to "really focus on structured data to make your pages eligible for a specific treatment in search," while adding that extra structured data "can be useful to understand the content better" but that he "wouldn't assume there's a visible effect / ranking change" from it. That last line is worth keeping in your pocket.

May–June 2026: fully deprecated, docs removed

Then the last door shut. A deprecation notice landed on Google's FAQ structured data documentation on May 8, 2026, stating the feature would no longer appear in Google Search starting May 7, 2026. In June 2026, Google removed the FAQ rich result documentation entirely — because the feature no longer shows in Search at all. Independent SEO Glenn Gabe noted that after the deprecation, FAQ snippets disappeared even for the government and health sites that had held onto their narrow eligibility since 2023.

So the SERP-dropdown payoff that every implementation guide was written to chase doesn't exist anymore — not for a small business, not for a hospital, not for anyone.

Does FAQ schema help you get cited in ChatGPT or AI Overviews?

This is the real reason FAQ schema is having a second life in 2026. With the Google rich result dead, the pitch pivoted: add schema so the AI assistants quote you. It's a reasonable hunch — if you're trying to understand how AI search changes the playbook in general, we've written a full breakdown of GEO vs. SEO for small business — but when you look at the controlled evidence, the hunch doesn't hold.

The Ahrefs study: 1,885 pages, no measurable citation gain

Ahrefs ran a difference-in-differences study — the kind that isolates the effect of one change — on 1,885 pages that added JSON-LD schema between August 2025 and March 2026, matched against 4,000 control pages that didn't. The result: no statistically meaningful citation gain anywhere. Google AI Mode rose 2.4% and ChatGPT rose 2.2%, both statistically indistinguishable from zero. Google AI Overviews actually fell 4.6% — a small, real decline that Ahrefs couldn't clearly attribute to the schema itself.

There is one eye-catching number on the other side. Ahrefs found AI-cited pages are nearly three times as likely to carry JSON-LD as non-cited pages; about 53% of cited pages run schema. But the study is blunt that this is correlation, not cause. The sites that bother adding structured data, it notes, "tend to also invest in technical SEO, publish authoritative content, build links, maintain their pages, and rank well in regular search." The schema is riding the wave, not making it.

Why AI assistants don't even read your JSON-LD

There's a mechanical reason the gain sits at zero. A separate experiment from searchVIU, cited by Ahrefs, tested whether five major AI systems — ChatGPT, Claude, Perplexity, Gemini, and Google AI Mode — actually use schema markup when they fetch a page live. None of them did. In searchVIU's words: "during direct retrieval, every system extracted only visible HTML content. JSON-LD, hidden Microdata, and hidden RDFa were all ignored."

Put plainly: the assistant reads what's on the page — the same words your human visitor sees. The invisible markup wrapper gets skipped.

What Google itself says about schema and generative search

You don't have to take a third party's word for it. Google's own AI-search guidance, published May 15, 2026, has a section called "Mythbusting generative AI search" that addresses this head-on. Under "Overfocusing on structured data," it reads: "Structured data isn't required for generative AI search, and there's no special schema.org markup you need to add. However, it's a good idea to continue using it as part of your overall SEO strategy, as it helps with being eligible for rich results on Google Search."

Read that twice. Google is telling you the markup is fine to keep for ordinary SEO reasons — but there is no magic schema that buys you AI visibility.

So what should a small business blog actually do instead

None of this means FAQs are a waste. It means you've been pointed at the wrong half of the job. The markup is the wrapper; the content is the product. Spend your effort on the content — it's also how you demonstrate the E-E-A-T (experience, expertise, authoritativeness, trust) that both Google and the AI models reward. Our guide on proving real experience in AI-written content goes deeper on that.

Write real FAQ content for extractability, not for the schema

Here's the payoff that survived every deprecation. Google's own Q&A documentation (the QAPage docs) frames the markup as helping it "generate a better snippet for your page" even when the rich result itself isn't shown. Notice what's doing the work in that sentence — the answer text, not the schema wrapper. Clear, genuinely useful questions and answers on the visible page can sharpen your plain blue-link snippet and give an AI model something clean to quote. That holds whether or not you ever touch a line of JSON-LD.

So write the FAQ the way a customer would actually ask it. Short question, direct answer, one idea per block. That's the extractable format — and it happens to be exactly what a human skimming your page wants, too.

Treat the markup as free maintenance, not a growth lever

If your platform adds the JSON-LD for you, keep it. Google said back in 2023 that unused structured data "does not cause problems for Search," so there's no cleanup emergency and no penalty for leaving it in place. Just don't put it on a growth roadmap. Mueller's framing still fits: markup makes a page *eligible* for a treatment, it doesn't *guarantee* one, and you "wouldn't assume there's a visible effect / ranking change" from piling it on. Low cost, low stakes, low effort — file it there and move on.

How to add FAQ schema with no code (if you still want it)

Decided it's cheap enough to bother? Fair enough — it's a few minutes and it does no harm. Here's the no-code path on the three platforms most small business blogs run on. (If you're still deciding where to publish in the first place, our non-developer's guide to picking a CMS covers that choice.)

WordPress: Rank Math or Yoast FAQ block

The easiest of the three. Both Rank Math and Yoast SEO ship a no-code FAQ block right inside the Gutenberg editor. In Rank Math, you drop in the "FAQ by Rank Math" block, type each question-and-answer pair, and the plugin generates and inserts the JSON-LD for you on publish — no hand-editing JSON anywhere. If you're switching over from Yoast, Rank Math can auto-import and convert your existing Yoast FAQ blocks during migration, so you're not rebuilding them from scratch.

Webflow: native Webflow AI or manual JSON-LD paste

Webflow now has a built-in option. On a paid Site plan or Workspace (any role except Reviewer), you can generate schema markup with Webflow AI in a single click from Page Settings, and it's dropped into that page's <head> for you — no hand-written JSON for most standard cases.

If you're on a plan without that, the manual route still works: generate the JSON-LD with a free generator and paste it into Page Settings → Custom Code → "Inside <head> tag" for the specific page. One caveat for CMS Collection-driven FAQ sections — each item needs its own dynamic-field-populated script and a fixed number of Q&A fields, or empty fields will throw validation errors. Decide on the field count before you build the collection.

Notion: you need a publishing layer like Super or Simple.ink

This is the catch for Notion blogs. Native Notion and notion.site pages have nowhere to add custom code — as Super.so puts it, "You can not add Schema code to your Notion-based website as there is no view where you can add custom code." To get FAQ JSON-LD onto a Notion-powered blog, you need a third-party hosting layer like Super.so or Simple.ink, which expose Head/Body/CSS custom-code sections where you can paste the script, globally or per page. If you're weighing Notion for your blog at all, we've looked separately at whether Notion can run a real business blog — schema is only one of the trade-offs.

One policy rule that still matters

Two rules, actually, and both are easy to break by accident.

First: whatever you mark up in FAQPage schema has to be visible to the reader on that same page. Marking up answers a visitor can't actually see — content tucked behind a tab they never open, or living on some other page — is a guidelines violation, not a clever growth hack.

Second, set your own expectations. Even with perfectly valid markup, Google states plainly that it "does not guarantee that your structured data will show up in search results, even if your page is marked up correctly according to the Rich Results Test." Eligibility is all markup ever buys you. And in 2026, for FAQ specifically, even that eligibility leads to a feature that no longer displays.

The bottom line

Add FAQ schema if your platform makes it a two-minute, no-code job — it's free and harmless. Don't add it expecting a rich result (gone since 2026) or an AI citation (the controlled data says it barely budges). Do write real, visible, well-structured FAQ content, because that's the part Google's snippets and the AI models actually read. The markup rides along; the answers do the work.