FAQ rich results are gone. What to do with the schema you already shipped.
Google has dropped FAQ rich results for almost all sites. Here's what to do with the FAQPage schema you've already deployed — and what still earns visibility.
For two years, FAQ rich results were the easiest visible win in SEO. Add FAQPage schema to a page, get expandable answers under your listing in Google, take up more space than your competitors. That's now over.
Google has confirmed it no longer surfaces FAQ rich results for the vast majority of sites — only a narrow list of government and authoritative health domains will continue to see them. If your team rolled out FAQ schema at scale, you now have a decision to make about thousands of pages of structured data that no longer earns the result it was built for.
What changed, in plain language
Google's FAQ rich results were the expandable question-and-answer snippets shown directly in search results for pages with FAQPage schema. Two years ago they were ubiquitous; over the last 18 months Google narrowed them progressively. The May 2026 update finishes the job. FAQPage markup will still validate. Pages still get indexed. The visible search feature is gone for commercial sites.
What you might be tempted to do — and why both options are wrong
The first instinct is to rip out the markup wholesale. Don't. The second instinct is to leave it untouched on the basis that "it can't hurt." That's also wrong — leaving thousands of unused FAQ blocks in your templates creates maintenance debt, slows page load if the underlying components render hidden content, and clutters your data layer. The right move is a triage, not a binary.
The three things FAQPage schema still earns
1. AI Overviews citations. Google's generative search system continues to pull structured answers as candidate citations. Pages with cleanly scoped Q&A blocks — schema or not — are over-represented in the source lists. The schema isn't a ranking factor here, but the underlying content pattern (a question, a short answer, named in the same place on the page) is what AI Overviews extracts.
2. Voice and assistant surfaces. Google Assistant and third-party voice surfaces still consume FAQPage schema in some markets. The traffic is small but the cost of keeping the schema valid is also small.
3. LLM crawler indexing for AEO. GPTBot, ClaudeBot and PerplexityBot all parse structured data when present. Question-and-answer pairs in machine-readable form are easier for them to lift cleanly into responses with attribution.
How to triage your existing FAQPage markup
Run through pages in this order.
Tier 1 — Keep and improve
Pages where the FAQ content directly addresses the search intent of the target keyword, the questions are real ones people ask (check Search Console queries, not your imagination), and the answers are two to four sentences of genuine information. For these, keep the schema and treat the FAQ block as primary content, not decoration. Move it above the fold where appropriate. Rewrite weak answers. These pages will continue to earn AI Overview citations and LLM mentions.
Tier 2 — Demote and reword
Pages where the FAQ was inserted to chase the rich result rather than to answer a real question. Symptoms: generic questions ("What is X?"), one-line answers, identical FAQ blocks duplicated across many pages. Remove the schema. Keep or rewrite the visible content into actual page copy. Free yourself from the schema obligation.
Tier 3 — Remove entirely
Pages where the FAQ was scraped from competitors, AI-generated without review, or thin in a way that suggests low quality. Take both the visible content and the schema off the page. This is also a good moment to audit those pages against the Helpful Content guidelines — pages that needed FAQ schema to look fuller are often the same pages flagged by quality systems.
What to use the freed-up budget on
Many SEO teams have ongoing maintenance contracts for FAQ schema generation and validation. That capacity is now better spent on building out actually-useful Q&A content as page copy, auditing AI Overview presence for high-intent commercial queries, and creating dedicated comparison and "alternatives to" pages — which AI Overviews favours and which traditional results don't reward as well.
There's an irony here: the agencies that pushed FAQ schema hardest for years are now selling AEO services. The work being done is broadly the same — structured, factual, machine-readable content. The output surface has changed.
A note on HowTo schema
HowTo rich results were removed in 2023 with similar drama. We have three years of data on what happened to pages that kept the markup versus pages that removed it. The short version: no measurable difference in organic traffic from removal. We expect the same pattern with FAQPage
FAQ
Should I remove all FAQPage schema from my site?
No. Keep it on pages where the FAQ content is genuinely answering the query intent. Remove it where it was inserted to chase a rich result.
Will keeping invalid FAQPage schema hurt rankings?
No direct ranking penalty. But Google's quality systems do look at the gap between marked-up content and useful content; if your FAQ blocks are weak, the page may be flagged for thin or unhelpful content regardless of schema.
What about HowTo and other schema types?
HowTo went two years ago. Product, Article, Review, Recipe, Event, and FAQ-on-government-sites are still supported. Check the Search Central rich results documentation for the current list.
23.png)
