Google Killed FAQ Rich Results 33 Months After Quietly Strangling Them

Google confirmed on May 8, 2026 that FAQ rich results stopped appearing in Search as of May 7. Support in the Rich Results Test and the dedicated FAQ rich result report disappear in June 2026, and Search Console API support is removed in August. The lost SERP real estate is being absorbed by AI Overviews and forum citation modules, not redistributed to the next-ranking organic pages.

A 33-month walkdown, not a sudden cut

This was the formal funeral, not the death. Google quietly killed FAQ rich results for almost everyone back in August 2023, when it restricted FAQPage rich results to "well-known, authoritative government and health websites" and removed HowTo entirely. Search Engine Journal's tracking at the time logged the actual visibility collapse before Google's announcement, with FAQ rich results disappearing from mobile around April 5, 2023 and from desktop around May 4. Most non-authority sites lost the SERP real estate within weeks.

What officially ended on May 7, 2026 was the residual carve-out for gov and health sites, plus the supporting tooling. Google's updated documentation puts the timeline in plain language: FAQ rich results no longer appear in Search as of May 7, the FAQ search appearance and rich result report drop in June, the Rich Results Test loses FAQ support the same month, and Search Console API support ends in August.

Search Engine Land's coverage of the final deprecation is the cleanest single read on the timeline. The rise-and-fall retrospective they ran earlier is the better read on why this happened: Google never wanted FAQ schema to become a CTR-stuffing tactic, and once it became one, the half-life of the feature was short.

The redistribution problem most teams will misread

If you ran FAQPage schema and saw a real CTR lift, the easy assumption is that removing the markup is the work. It isn't. The harder question, and the one your traffic is about to answer for you anyway, is where the lost click went.

For a non-trivial slice of FAQ-driven queries, the click isn't going to "the page that ranks below mine." It's going to the AI Overview module that increasingly answers short, recurring questions inline, plus the Reddit and Quora citation modules Google now surfaces in expert advice boxes. We covered the 50-domain citation lock inside AI Overviews two weeks ago, and the read here is the same: FAQ schema and AI Overviews compete for the same intent. Google just decided the model should answer it instead of the markup.

That's a worse outcome for most sites than "we lost two SERP boxes." Losing rank to a competitor is fixable with the usual playbook. Losing the click to an AI Overview is a different fight, and the lever you reach for isn't a better page. It's eligibility to be cited inside the Overview at all.

The audit isn't "remove the markup"

The first GSC pull isn't a list of pages with FAQPage JSON-LD. It's the list of pages that previously logged FAQ rich result impressions, sorted by how much CTR they were earning before the May 7 cut. Compare each one's CTR for the trailing 30 days against the 30 days prior. The mobile drop usually shows up first, with desktop a few days behind, the same pattern that played out in 2023.

Once you have the affected list, the meaningful triage is whether each lost query now shows an AI Overview, a forum-thread cluster, or a clean ten-blue-link SERP. Each one wants a different response.

For queries where AI Overviews swallowed the rich result, the lever is structured-data hygiene aimed at citation eligibility, not visual rendering. Clear question-and-answer formatting inside body copy, named entity references, and source-able numbers Google's grounding model prefers. We covered how Bing's AI grounding index reads similar signals last week, and the requirements roughly converge.

For queries that became Reddit-and-Quora boxes, the only honest response is to participate in the forum where the question lives. Google's $39M licensing deal with Reddit in Q1 isn't going to reverse. For the genuinely blue-link queries that survive both, you compete on titles, meta descriptions, and the first 60 words of body copy where Google snips the snippet preview.

From what I've seen, most teams will skip the second step entirely and just delete their FAQPage JSON-LD. That's cheap, but it's solving the wrong problem. The lost CTR isn't going to your competitors, so cleaning up your own markup doesn't get any of it back.

Should you actually remove the markup?

Google explicitly said you don't have to. Invalid FAQPage schema won't trigger a penalty, and unused valid markup is just unused. There's a soft argument for keeping it. Schema with question-and-answer structure is still a citation signal for Bing's grounding index and probably for the AI Overview model itself, even if the visible rich result no longer renders. Removing it cleans up your codebase but probably loses you a small amount of AI citation surface area. Hard to tell exactly how much, since neither Google nor Bing has published a breakdown of grounding signal weights, and outside testers are working from inference.

The one exception worth acting on this month: if your FAQPage schema currently throws errors in the Rich Results Test, fix it or remove it before June. Once the report disappears, you lose the easiest debugging surface for any structured data you choose to keep.

What changed for the gov-and-health carve-out crowd

The roughly 0.5% of sites that retained FAQ rich results between August 2023 and May 2026 were almost entirely .gov, .edu, and major health publishers. They lost something real on May 7, where most marketers had nothing left to lose. If you operate one of those sites, the GSC drop on May 7 is going to look brutal in isolation, but the long-tail health and government queries the carve-out covered tend to retain decent CTR even without rich formatting because the answer source matters more than the visual treatment.

Where the lost CTR actually went

FAQ schema's formal end is the third visibility lever Google has retracted in the last twelve months, after HowTo (gone), the slow squeeze on snippet length, and the steady expansion of AI Overview surface area into queries that used to be ten-blue-link. None of those changes individually moves a healthy site's traffic that much. Stacked, they describe a SERP where the rich-result toolkit pre-2023 marketers built CTR strategies around is mostly gone.

The honest read on May 7 is that the tactic stopped working in 2023, and Google just stopped pretending the gov-and-health carve-out kept the feature meaningful. The teams that already moved structured-data work into AI citation eligibility have a cleaner runway. The teams still treating schema as a SERP-real-estate multiplier have some catch-up to do, and the catch-up isn't trivial.

Anyway, the markup is the smallest piece of this. The harder work is figuring out which queries are still yours to compete for at all, and which ones Google quietly handed to its own answer engine while everyone was busy auditing JSON-LD.

Notice Me Senpai Editorial