Keep FAQ schema only where it accurately marks up useful questions and answers that a customer can read on the page. The rich result it used to earn is gone. Google's own changelog entry from May 8, 2026 states it in one line:
This feature will no longer appear in Google Search starting May 7, 2026.
Last verified: September 2, 2026.
The questions themselves can still help customers. The markup should not be sold as a ranking shortcut for Google or AI search.
What changed
This was a two-step retreat, not a surprise. On August 8, 2023, Google narrowed the feature:
Going forward, FAQ (from FAQPage structured data) rich results will only be shown for well-known, authoritative government and health websites.
For a plumbing company or a med spa, the rich result had already been gone for nearly three years at that point. May 7, 2026 removed it for everyone, and Google deleted the feature documentation on June 15, 2026.
That change affects the Search presentation. It does not make a useful FAQ section bad content. Buyer questions are still worth answering well: in the Cheers panel for the 28 days ending September 2, 2026, prompts shaped as how, what, why, or should-I questions returned the highest pooled citation rate of any prompt family, 27.3% against 19.6% for "best near me" prompts, across 228 tracked prompts and 62 organizations. Answering the question is what earns the citation. The markup is not the lever. The prompt-shape cuts are in AI visibility statistics, with the sampling rules in the methodology.
When to keep the markup
Keep it when the same questions and complete answers are visible to customers, the information is current, and another system or internal consumer uses schema.org FAQPage data.
Google's AI optimization guide is blunt about the AI angle:
Structured data isn't required for generative AI search, and there's no special schema.org markup you need to add.
That is the standard I use. Accurate markup can stay, and it does not get an AI visibility claim attached to it.
When to remove or rewrite it
Remove markup that describes hidden content, repeats stale answers, conflicts with branch policy, or marks up promotional copy as customer guidance. Rewrite the visible answer when hours, services, warranties, pricing, eligibility, or location rules have changed.
Do not swap FAQPage for QAPage. Google says QAPage is for one question where users can submit answers. It is not valid for a company-written FAQ section or a blog post with one fixed answer.
The location-page checklist covers visible branch information. The JSON-LD guide covers how structured data should reflect source truth.
Next step
Is AI recommending your business?
Find out how visible you are across ChatGPT, Gemini, Perplexity, and AI Overviews.
The practical decision
I would not fund an FAQ-schema project for Google rich results. I would fund accurate answers to questions buyers repeatedly ask, then keep the markup only if it remains truthful and useful to a known consumer.
Use the AI Visibility Grader to inspect current answer visibility. Do not credit an FAQPage block for a result unless a controlled test can support that claim.
Sources
Every link below was opened and checked on September 2, 2026.
- Google Search documentation updates. Google Search Central changelog. The May 8, 2026 entry announcing the FAQ rich result would stop appearing on May 7, 2026, and the June 15, 2026 entry removing the documentation.
- Changes to HowTo and FAQ rich results. Google Search Central Blog, August 8, 2023, updated September 14, 2023. The original narrowing to well-known, authoritative government and health websites, and the note that unused structured data does not cause problems for Search.
- Optimizing your website for generative AI features on Google Search. Google Search Central, last updated July 2026. No special schema required for generative AI features.
- QAPage (Question and Answer) structured data. Google Search Central, last updated June 15, 2026. Why a site-written FAQ page is not a valid QAPage use case.
- Cheers panel, 28 days ending September 2, 2026. 119 home-services organizations, pooled rates by prompt shape, aggregate only. Methodology.
Dylan Allen-Arnegard is the CEO and Co-Founder of Cheers. I help multi-location service brands get recommended by AI search and Google, with results tracked by location, employee, and competitor.