How to Create FAQPage Schema in 2026 (Post-Deprecation)

Chapter 1. Before you start

FAQPage schema is in a different place in 2026 than it was in 2024. Google deprecated the visible rich result for most sites in May 2025, but the schema itself remains a recognized Schema.org type and continues to feed AI engines that extract Q&A pairs from page markup. The decision to ship it now is about AI visibility and structured-data hygiene, not Google SERP real estate.

• Confirm the FAQs are genuine. Real user-relevant questions and answers, not keyword-stuffed dressing. Google still polices fake FAQ markup.
• Each Q&A must be visible on the page. Schema content has to match what users see; hidden FAQ schema is a manual-action risk.
• Decide the FAQ section's role. Editorial value for users, AI extraction signal for engines, or both. All three are valid; "to get the rich result" is no longer in the list.
• Pick the right schema type. FAQPage for a single-author multi-question page; QAPage for a single user-submitted question with multiple community answers.

Chapter 2. What happened to FAQ rich results in 2025?

On May 7, 2025, Google officially removed FAQ rich results from Google Search for the vast majority of sites. The earlier 2023 narrowing already restricted eligibility to "well-known, authoritative government or health-focused websites"; the 2025 announcement formalized the full retirement for everyone else.

What this means concretely:

• The accordion snippet no longer renders in Google SERPs for ecommerce, B2B, blog, or non-authoritative sites.
• Search Console's FAQ rich result report is being phased out. The Rich Results Test still parses the markup but no longer indicates Google SERP rendering.
• The Schema.org vocabulary itself is unchanged. FAQPage, Question, and Answer remain valid types Google parses.
• AI engines did not deprecate anything. ChatGPT, Perplexity, Gemini, and Google's own AI Overviews continue to extract Q&A pairs from valid FAQPage markup.

Chapter 3. Should you still ship FAQPage schema?

Yes, for three reasons. The Google SERP rich result is gone for most sites, but the schema still produces value:

1. AI engine extraction. ChatGPT, Perplexity, and Gemini ingest FAQPage markup to extract clean Q&A pairs. Pages with FAQPage schema are easier for these engines to cite as authoritative answers.
2. Semantic clarity. The markup explicitly tells every parser "this is a question, this is the answer" - which improves NLP extraction even for engines that don't have a formal FAQPage handler.
3. Future-proofing. Google has deprecated and restored features in the past. The cost of leaving valid FAQPage schema in place is zero; the cost of removing it and re-adding it later is non-zero.

The case for removal is weak. The schema doesn't hurt rankings, doesn't trigger manual actions when properly used, and continues to serve AI-search purposes. Sites that ripped out FAQPage schema after the deprecation saw no measurable upside; sites that kept it retain the AI-citation benefit.

Chapter 4. Required and recommended properties

FAQPage requires mainEntity containing an array of Question entries. Each Question needs name (the question text) and acceptedAnswer (an Answer with text).

{
  "@context": "https://schema.org",
  "@type": "FAQPage",
  "mainEntity": [
    {
      "@type": "Question",
      "name": "How long does GA4 setup take?",
      "acceptedAnswer": {
        "@type": "Answer",
        "text": "About 90 minutes if you have admin access on Google Analytics, your website, Google Ads, and Search Console."
      }
    },
    {
      "@type": "Question",
      "name": "Do I still need Universal Analytics installed?",
      "acceptedAnswer": {
        "@type": "Answer",
        "text": "No. Google sunset Universal Analytics on July 1, 2023 and deleted all historical UA data on July 1, 2024."
      }
    }
  ]
}

Answers can contain plain text or HTML. Plain text is cleaner for AI engines to extract; HTML supports formatted answers with paragraphs, lists, and inline links. Both validate. Whatever you choose, the answer text in the schema must match the answer text visible on the page exactly.

Chapter 5. Content match and quality rules

Google's structured data policies still apply to FAQPage schema even after the rich-result deprecation. Violating them can trigger a manual action that affects rich-result eligibility for other schema types on the site.

• Every Q&A in the schema must appear verbatim on the visible page. Hidden FAQ schema is a violation.
• The full question and the full answer must be marked up, not just snippets. Truncating answers in the schema is a violation.
• No advertising. FAQPage cannot be used as an ad delivery surface.
• No obscene, violent, dangerous, or illegal content. Standard Google content policy.
• Real Q&A, not keyword-stuffed dressing. "What is the best X for Y in 2026?" repeated 20 times across the page is a violation even if technically valid markup.

Chapter 6. Where do you place FAQPage schema?

On any page that has a genuine FAQ section, embedded inside a <script type="application/ld+json"> tag in <head> or before </body>. One FAQPage schema per page maximum.

Common placements:

Chapter 7. The breakages we see most often

Ranked by frequency across 47 ecommerce schema audits over the past 24 months:

• Content mismatch between schema and visible page: answer text in JSON-LD differs from rendered HTML. Policy violation. 12 of 47 sites still shipping FAQPage.
• Truncated answers in the schema: full visible answer not represented in the markup. 9 of 47.
• FAQPage shipped without visible FAQ on the page: hidden schema, manual-action risk. 7 of 47.
• Keyword-stuffed FAQ sections: artificial questions inserted for SEO. 5 of 47.
• FAQPage used for QAPage scenarios: single question with multiple user answers should use QAPage instead. 3 of 47.
• Multiple FAQPage schemas on the same page: only one allowed. 2 of 47.

We track FAQPage schema integrity through our Sentry product; the structured-data rule set catches every breakage above on running sites.