Your existing FAQPage schema markup doesn’t need to be removed, and it hasn’t been producing visible FAQ rich results in Google Search since May 7, 2026. Google’s own documentation for the FAQ structured-data type now carries a deprecation notice confirming the feature “will no longer appear in Google Search,” closing out a pullback that started in August 2023. The markup itself is still valid Schema.org vocabulary; what changed is only what Google does with it.
What actually changed, and when
Three dates matter here, and they’re easy to blur together in secondhand summaries. In August 2023, Google first narrowed FAQ rich results to a small set of “well-known, authoritative government and health websites,” cutting everyone else out already â most sites lost this feature almost three years before the headlines caught up. On May 7, 2026, Google ended FAQ rich results for those remaining government and health sites too, and added a deprecation notice to its own documentation the next day. By June 15, 2026, Google had pulled the FAQ documentation page entirely, stating the feature “is no longer shown in Google Search results.”
None of this touches the Rich Results Test’s broader function or the other structured data types Rank Math generates for you â Article, BreadcrumbList, and the rest are unaffected. It’s specifically the FAQ snippet, the expandable question list that used to show up directly under a search result, that’s gone.
Three eras of FAQ rich-result eligibility, compared
Laid out against each other, the three phases Google’s own changelog describes show a feature narrowing twice before it was removed outright, not a single sudden change.
| Period | Who was eligible | What showed in Search |
|---|---|---|
| Before August 2023 | Any site with valid FAQPage markup | Expandable Q&A rich result for up to two questions per page, per Google’s original FAQ documentation |
| August 2023 â May 7, 2026 | “Well-known, authoritative” government and health websites only | The rich result continued, but only for that narrow set of sites; everyone else’s markup stopped producing one |
| Since May 7, 2026 | No sites | No FAQ rich result for anyone; Search Console’s FAQ report and the Rich Results Test’s FAQ support were withdrawn by June 15, 2026 |
Read this way, most sites’ FAQ rich results actually ended in 2023, not 2026 â the May 2026 change mainly closed the exception for government and health publishers who’d kept the feature for almost three years after everyone else lost it.
The two mistakes we’re seeing in the wild
Site owners reacting to this change tend to overcorrect in one of two opposite directions, and both come from treating “Google stopped showing it” as equivalent to “Google now penalizes it.”
Mistake one: ripping FAQPage schema out of every post. There’s no indication anywhere in Google’s changelog that leaving the markup in place causes a ranking or quality problem. FAQPage remains a standard, valid type in the Schema.org vocabulary, and Google’s own deprecation notice is explicit that site owners “don’t need to remove” it. Stripping it out is work that fixes nothing and, if you remove it carelessly with a find-and-replace across templates, risks breaking unrelated markup in the same block.
Mistake two: assuming it’s still buying you anything in Google Search. The opposite error is leaving FAQ sections on pages purely for the rich-result payoff that no longer exists, then wondering why click-through didn’t move after “optimizing for FAQ snippets.” If the only reason a FAQ block exists on a page is the rich result, that reason is gone. The content can still earn its place for other reasons â covered next â but not that one.
What’s genuinely still unresolved: AI citations
Whether FAQPage schema helps an answer get cited by an AI assistant is a separate question from whether Google Search shows a rich result for it, and it’s worth being precise about how little either side has actually confirmed. Google’s changelog doesn’t connect the FAQ deprecation to AI Overviews or AI Mode at all â the notice is scoped to classic Search rich results. Some SEO commentary has speculated that structured Q&A pairs make a passage easier for a language model to lift as a citation, but that’s a hypothesis about how retrieval and summarization systems parse a page, not a confirmed mechanism, and no provider has published a document saying FAQPage markup specifically improves citation odds. Treat it as unproven in either direction rather than as a replacement justification for the rich-result payoff that did disappear.
What is settled is narrower than either the “FAQ schema is dead” or “FAQ schema is secretly more important now” takes circulating since the change: Google retired one specific visual treatment in classic Search results, kept the underlying vocabulary valid, and said nothing new about AI surfaces either way. Decisions about AI citability belong on their own evidence, gathered separately, not borrowed from a changelog entry that never mentions them.
A decision checklist for Rank Math sites
Run through these four questions for each post that currently carries FAQ schema, rather than making one blanket decision for the whole site. A blog built up over a few years usually has FAQ blocks added under more than one content standard, so the right call for a post written last month and one written three years ago often isn’t the same call.
- Is the FAQ content answering questions real readers actually ask, independent of the schema? If yes, keep both the content and the markup â you’re not touching anything that was only there for the snippet. If the FAQ section exists only because a content checklist said “add three FAQs,” the content is the thing to cut, not the schema wrapper around it.
- Did you add the FAQ block through Rank Math’s Schema Generator or the FAQ block in the editor? Either way, the JSON-LD is generated automatically from the visible Q&A content on the page â there’s no separate toggle to switch off just the rich-result behavior, because Google controls that on its end, not Rank Math.
- Are you still relying on the Rich Results Test to check this schema? Google removed Rich Results Test support for the FAQ type in the same wave of changes. Use the Schema Markup Validator instead to confirm the JSON-LD is well-formed; it checks against the Schema.org spec rather than against Google-specific eligibility, which is the right tool now that there’s no Google eligibility left to check.
- Does Search Console still show a FAQ-specific report? If you’re looking for it and can’t find it, that’s expected, not a bug on your end â the dedicated FAQ filter and report were removed from Search Console in the same June 2026 wave.
What the JSON-LD looks like, unchanged
Nothing about the markup’s structure is wrong or needs rewriting. A Rank Math FAQ block still outputs the same shape it always has:
{"@type": "FAQPage", "mainEntity": [{"@type": "Question", "name": "How long does shipping take?", "acceptedAnswer": {"@type": "Answer", "text": "Orders ship within two business days and arrive in 5-7 days domestically."}}]}
That block validates against the Schema.org spec today exactly as it did in 2022. The only thing that stopped is a downstream behavior Google chose to retire on its end â not a requirement that the markup itself change shape, and not a signal that the FAQPage type was deprecated by Schema.org, which it wasn’t.
What to do with the markup itself
For the overwhelming majority of posts, the right move is to do nothing to the schema and look only at the content underneath it. If a FAQ section genuinely pre-empts a question your article’s intro, headings, and body don’t already answer, it still belongs on the page â for readers and for any system, Google’s own ranking included, that still rewards a page for thoroughly covering a topic. If it’s three generic questions bolted on to hit a content template’s checklist, that’s dead weight whether or not a rich result was ever attached to it, and removing the content â not hunting down and deleting the JSON-LD â is the actual fix.
This is also a good moment to audit FAQ content written under the older assumption that it would “win featured snippets.” That framing predates this deprecation and should be read now as a content-quality piece, not an SEO mechanism â the snippets it was written to win are the ones that just stopped appearing. For how Rank Math’s other automatic schema holds up against Google’s actual structured-data guidelines, see what Rank Math fills in for Article schema, and if your sitemap hasn’t picked up recent changes at all, start with diagnosing a stale Rank Math sitemap before assuming a schema problem. The underlying judgment call â keep content because it’s useful, not because of what it might trigger in a search result â is the same one covered in rewriting a thin AI-generated page against Google’s helpful content guidelines.
Google will likely keep retiring individual rich-result types on its own schedule, the way HowTo lost its rich result in the same August 2023 round that first narrowed FAQ. The durable move isn’t chasing each one’s eligibility rules as they change; it’s writing the FAQ content, or any content, because it answers something a reader actually needed, and letting whatever rich-result status Google assigns it be a byproduct rather than the goal.
