Does FAQ schema still help SEO and AI search in 2026?

FAQ schema (FAQPage markup) no longer gets you anything visible in Google Search. Google limited FAQ rich results to government and health sites in August 2023, and on 7 May 2026 it stopped showing them for every site. The markup is still valid, it does no harm and it costs almost nothing to keep. But no study so far shows it gets you cited more often by Google AI Overviews, ChatGPT or Perplexity.

My short answer: keep FAQ schema if you already have it, add it where you already have a real FAQ, and don't pay anyone to "optimise your FAQ schema for AI". AI answers quote the questions and answers people can actually see on your page. The markup is a small extra on top of that.

What happened to FAQ rich results in Google?

Google took FAQ rich results away in two steps. First it restricted them to government and health sites in August 2023, then it removed them for everyone on 7 May 2026. Back in 2023, Google said FAQ rich results would only appear for "well-known, authoritative government and health websites", as Search Engine Journal reported. For almost every business site, those expandable question dropdowns under a search listing vanished that month.

The second step landed on 7 May 2026. FAQ rich results stopped appearing in Google Search altogether, government and health sites included. Google's FAQ structured data documentation now carries a deprecation notice. The same Search Engine Journal report says the FAQ search appearance filter, the rich result report and Rich Results Test support were due to go in June 2026, with Search Console API support following in August 2026.

What this means for you:

• If you were counting on FAQ dropdowns to take up extra room on the results page, most sites lost that in 2023. Now nobody has it.
• If a reporting dashboard pulls FAQ appearance data from the Search Console API, that data stops. Take it out of the report, or someone will read the drop to zero as a traffic problem.
• Google's Rich Results Test won't validate FAQPage any more. Check syntax with the Schema.org validator instead.

Does FAQ schema hurt your rankings now?

No. FAQ schema doesn't hurt your Google rankings. Google's FAQ structured data documentation says the markup can stay where it is.

The only real risk is the same one as always: markup that doesn't match what's on the page.

Does FAQPage markup help you get cited in AI Overviews?

Not specifically, according to Google, and the best independent data says the same. Google's guidance on AI features in Search says there's no special schema.org structured data you need to add to appear in AI Overviews or AI Mode. To show up in AI Overviews, a page has to be indexed and eligible for a snippet. Those are the same requirements as ordinary Google Search.

Ahrefs tracked 1,885 pages that added JSON-LD between August 2025 and March 2026 and compared them with matched control pages. As Search Engine Roundtable summarised, adding schema made no major difference to citations on any platform. AI Overviews actually showed a small 4.6% decline against the controls. AI Mode and ChatGPT showed slight gains that could easily be noise.

The study covered every JSON-LD type, not just FAQPage. I don't see any reason FAQ markup on its own would beat that overall result, though.

Why care? Because AI answers now sit on top of most results. Search Engine Roundtable reported in October 2026 that over 80% of Google searches show AI Overviews. What gets you quoted there is content that answers the question clearly. I go through that in how to get cited in AI Overviews and ChatGPT.

What about ChatGPT, Perplexity and Copilot?

Microsoft is the only one that has said schema helps. Bing's Fabrice Canel said in March 2025 that schema helps Microsoft's LLMs understand content. I couldn't find any public statement from OpenAI about ChatGPT search, or from Perplexity.

AI search tools fetch your page and read the text. A clean question-and-answer block is easy for them to lift, markup or no markup. The bigger risk is a page they can't read at all. Vercel's analysis of AI crawler traffic found that major AI crawlers, including OpenAI's and Anthropic's, don't run JavaScript, so if your FAQ and your JSON-LD are both injected client-side, the crawler may see an empty page. Check that your robots.txt lets these crawlers in, too. My free AI Crawler Checker tells you in a few seconds.

So what does FAQ schema still do in 2026?

It labels question-and-answer pairs for systems that read JSON-LD, Bing being one, and it pushes you to write short answers that stand on their own. That's about it. It won't bring traffic or rankings by itself.

• It labels the pairs clearly. Anything that reads JSON-LD, including Bing by its own account, can see exactly which text is the question and which is the answer. Your own site search or chatbot can reuse the same data.
• It forces better editing. The markup only works when each answer makes sense alone and matches the page. Short, self-contained answers are what AI tools like to quote anyway.
• It's cheap. In WordPress it's a block or a setting. In Next.js it's a few lines of code built from the same data as the visible FAQ.

Anyone promising traffic, rankings, rich results or guaranteed citations from FAQ schema is selling 2022 advice.

When is FAQ schema worth adding, and when should you skip it?

Add FAQPage markup where a genuine FAQ is already visible, like a service page or a blog post with real reader questions. Skip it for filler, and never mark up questions that aren't on the page.

How do you add FAQ schema in WordPress?

The easiest route is the FAQ block in Yoast SEO or Rank Math. You write the questions in the editor, they show on the page, and the plugin outputs FAQPage JSON-LD from that same content. Some page builders have an FAQ schema setting on their accordion widget too. Check yours, because a lot of accordions output plain HTML and nothing else.

1. 1

   Collect real questions

   Pull them from sales emails, enquiry forms, calls and the queries in Google Search Console, not from guesswork.

2. 2

   Put them on the page

   Write a short, direct answer to each question and publish it where visitors can see it.

3. 3

   Generate the markup

   Use your SEO plugin's FAQ block, your builder's schema setting or a generator, so the JSON-LD matches the visible text.

4. 4

   Check for duplicates

   If your SEO plugin and your builder both output FAQPage, switch one off so the page has a single FAQ block.

5. 5

   Validate and view source

   Run the URL through the Schema.org validator and confirm the JSON-LD is in the page source, not added later by JavaScript.

If your theme or builder has no FAQ option, try my free FAQ Generator. It reads your page or service details, writes the questions customers actually ask with answers taken only from your content, and gives you FAQPage JSON-LD to paste into a code block. It also flags questions your page can't answer yet, which is a handy list of what's missing. For other schema types, my free Schema Markup Generator does the same job.

How do you add FAQ schema in Next.js?

Keep the FAQ content in one array and render both the visible list and the FAQPage JSON-LD from it in a server component. Then the markup can't drift away from the page.

const faqs = [
  { q: "Do you build on Bricks Builder?", a: "Yes. I build WordPress sites on Bricks from your Figma design." },
  { q: "Can you move my site without downtime?", a: "Yes. I plan DNS and email so the site stays online." },
];

export default function FaqSection() {
  const jsonLd = {
    "@context": "https://schema.org",
    "@type": "FAQPage",
    mainEntity: faqs.map((f) => ({
      "@type": "Question",
      name: f.q,
      acceptedAnswer: { "@type": "Answer", text: f.a },
    })),
  };

  return (
    <section>
      <h2>Frequently asked questions</h2>
      {faqs.map((f) => (
        <details key={f.q}>
          <summary>{f.q}</summary>
          <p>{f.a}</p>
        </details>
      ))}
      <script
        type="application/ld+json"
        dangerouslySetInnerHTML={{ __html: JSON.stringify(jsonLd).replace(/</g, "\\u003c") }}
      />
    </section>
  );
}

A few things in there matter:

• Render it on the server. In the Next.js App Router, a server component like this puts both the answers and the JSON-LD in the first HTML response. That's what crawlers that skip JavaScript need. Check it with view-source or curl.
• Use details and summary. The answers stay in the HTML even while they're collapsed. Some accordion libraries only load the answer text on click, which isn't what you want here.
• Escape < in the JSON. The replace call stops a stray </script> in your content from breaking out of the tag.

On a headless CMS, map the CMS FAQ field into the same array. My JSON-LD structured data guide covers the rest of the markup a marketing site needs.

What should you ask a developer before paying for schema work?

Ask what it will measurably change, whether the markup is generated from the visible content and server-rendered, and whether something more urgent (indexing, crawler access) should come first. Adding FAQ schema is a small job. If it's quoted as a big "AI visibility" project, be wary.

• What will this measurably change? If the answer is rich results or guaranteed AI citations, the advice is out of date.
• Is the markup generated from the visible content? It should come from the same source, not get typed into a separate field that drifts over time.
• Is it in the server-rendered HTML? If it's injected client-side, AI crawlers may never see it.
• Will it duplicate what my SEO plugin already outputs? Two FAQPage blocks on one page is a common mess.
• What's more urgent? Pages that aren't indexed, failing Core Web Vitals or robots.txt rules that block AI crawlers usually cost far more visibility than missing FAQ markup.

When I do technical SEO and Core Web Vitals work, schema is one line on a longer list. I sort out indexing and crawl access first, because that's usually where the lost visibility is.

What should you do next?

Pick one of your key service pages and run it through the FAQ Generator. Publish the questions that fit the page, then paste in the JSON-LD. It takes about ten minutes and it's free. If you want someone to check structured data, AI crawler access and indexing across a whole WordPress or Next.js site, have a look at my technical SEO service or book a 20-minute call.