How to Add FAQPage JSON-LD to Make FAQ Content Easier for AI to Understand

What is FAQPage JSON-LD?

FAQPage is a schema.org type for a page containing a list of frequently asked questions, each with an accepted answer written by the site. In JSON-LD it looks like this:

<script type="application/ld+json">
{
  "@context": "https://schema.org",
  "@type": "FAQPage",
  "mainEntity": [
    {
      "@type": "Question",
      "name": "Do you deliver to Northern Ireland?",
      "acceptedAnswer": {
        "@type": "Answer",
        "text": "Yes. Standard delivery to Northern Ireland takes 3 to 5 working days and costs £6.95, free on orders over £60."
      }
    },
    {
      "@type": "Question",
      "name": "Can I return an opened item?",
      "acceptedAnswer": {
        "@type": "Answer",
        "text": "Yes, within 30 days, as long as it is unused and in its original packaging. Return postage is free."
      }
    }
  ]
}
</script>

The questions and answers must match what visitors see on the page. The example values are illustrative.

Does FAQ schema improve visibility in AI search?

There is no good public evidence that FAQPage markup, on its own, increases citations in AI answers. The reasons to be cautious:

• Google says no special structured data is needed to appear in AI Overviews or AI Mode, and that structured data should match visible text.
• AI assistants usually work from page text. A clear question heading with a direct answer on the page already gives them what the markup describes.
• Reported gains are usually confounded. When teams add FAQ markup, they often also add new FAQ content, rewrite answers and restructure pages. The content changes are the more likely cause of any improvement.

Markup can make the question-and-answer structure unambiguous for systems that read it, which is a modest benefit. It is not a lever that moves AI visibility by itself; see whether AI systems cite visible page text instead of JSON-LD.

What did Google change about FAQ rich results, and does it matter?

Google reduced FAQ rich results in August 2023, limiting them to well-known, authoritative government and health websites, and restricted How-to rich results, which it later deprecated. In May 2026 it went further: FAQ rich results stopped appearing in Google Search from 7 May 2026, for every site. Google also announced it would drop the FAQ search appearance filter, the rich result report and Rich Results Test support in June 2026, and FAQ support in the Search Console API in August 2026.

That removed the last visual reason to add FAQ markup for Google: the expandable questions under a search result. It does not make the markup invalid or harmful. FAQPage is still a valid schema.org type, and Google has said unused structured data does not cause problems. It does not affect the value of the FAQ content itself either. For AI search, the change matters little, because the value there was always in the content.

Should you keep adding FAQPage schema?

A practical rule:

• Keep or add it where the page genuinely contains an FAQ, the answers are accurate and maintained, and your platform adds it at little cost.
• Do not add it to pages where questions are hidden, marketing-driven or duplicated across many pages.
• Remove or fix it where the marked-up answers no longer match the page.
• Do not prioritise it over writing better questions and answers, which is where the real benefit lies.

How do you add FAQPage JSON-LD correctly?

1. Use it only on pages with a visible FAQ written by you, not user-submitted answers.
2. Include every marked-up question and answer on the page, word for word or very close.
3. Use one FAQPage block per page, containing all the questions.
4. Avoid advertising in answers; keep them informational.
5. Do not repeat the same FAQ markup across many pages; mark it up where the FAQ primarily lives.
6. Validate with the Schema Markup Validator, and keep the markup updated when answers change.

Many content management systems and SEO plugins can generate the markup from FAQ blocks automatically, which keeps it in sync with the visible text.

How do FAQPage, QAPage and Speakable differ?

For quotable answers, none of these replaces clear writing. The quotability comes from a self-contained answer sentence, not the markup.

What makes FAQ content itself useful to AI?

• Real questions, taken from customers, sales and support, in their words.
• Direct first sentences that answer the question fully, with numbers or conditions.
• Specifics: prices, timings, eligibility, limits.
• No duplication: each question answered in one place, with other pages linking to it.
• Links to fuller pages for questions that need more than a paragraph.
• Maintenance: answers updated when policies change.

Where FAQs should live, on existing pages or a separate FAQ page, is covered in whether FAQs should go on existing pages or separate FAQ pages.

What does a weak FAQ answer look like next to a strong one?

Weak:

What are your delivery options? We offer a range of flexible delivery options to suit your needs. Please see our delivery page for more details.

Strong:

What are your delivery options? We offer standard delivery in 3 to 5 working days for £4.95, free on orders over £50, and next-day delivery for £8.95 if you order by 2pm Monday to Thursday. We deliver to the UK mainland; Highlands and islands take one or two extra days.

The weak answer gives an AI engine nothing to quote and sends readers elsewhere. The strong one answers the question completely, with the conditions that matter, and would be just as useful with or without markup.

How do you test whether FAQ changes help?

Change one thing at a time. For example, add markup to a group of existing FAQ pages without changing their content, and leave a similar group unmarked. Track whether the pages are cited for their questions in AI answers over the following weeks, with several runs per question. If the marked-up group shows no difference, the markup alone is not moving AI visibility for you, and effort is better spent on the content.

What do people ask about FAQ schema?

Is FAQ schema a spam risk?

Not when it accurately describes a visible FAQ. Marking up content that is not on the page, or adding promotional FAQs everywhere, goes against structured data guidelines.

Should every page have an FAQ section?

No. Add FAQs where real questions exist that the page does not otherwise answer. Forced FAQs add length without value.

Does Bing use FAQ markup?

Bing supports schema.org markup generally. As with Google, the visible content matters most for AI answers.

Can FAQ markup hurt my site?

Only if it is inaccurate, hidden or misleading. Accurate markup on a genuine FAQ is harmless.