You spent an afternoon last year wiring up FAQ schema on your service pages because Google rewarded it with that expandable blue-link snippet sitting right under your title. Then one week in May 2026, the snippet just stopped showing up, and the question we keep getting from Las Vegas business owners is whether all that markup work was wasted. It wasn't. The snippet disappeared, but the job FAQ schema actually does, making your content unambiguous to machines, matters more now than it did when it was just decorating the SERP.
In this guide:
- What Changed When Google Removed FAQ Rich Results in May 2026
- Why AI Overviews Still Reward Clearly Structured Q&A Content
- FAQ Schema vs Other Structured Data: What Still Earns Its Keep in 2026
- How FAQPage Markup Helps ChatGPT and Perplexity Even Without a Google Snippet
- What We Actually Recommend for Las Vegas Small Business Sites Right Now
- A Technical Note on Diagnosing What's Actually Hurting Your AI Visibility
What Changed When Google Removed FAQ Rich Results in May 2026
Google retired the visual FAQ rich result on May 7, 2026, according to Google's own update notes on the change. The reasoning tracks with everything else Google has shipped since 2024: AI Overviews had already taken over that chunk of SERP real estate for most informational searches, so the old expandable accordion was redundant next to an AI-generated answer sitting in the same spot.
Here's the part that trips people up. FAQPage schema itself was never deprecated. Google only stopped rendering the specific blue-link snippet in classic search results; the structured data markup behind it is still valid, still crawlable, and still parsed the same way it always was.
If your only reason for adding FAQ schema was the cosmetic snippet, you lost that specific benefit and nothing else. You didn't lose the underlying machine-readability, and that's the piece most site owners conflate.
Losing a rich result is not the same as losing structured-data value. Google changed what it renders in the SERP, not how it parses your page.
Why AI Overviews Still Reward Clearly Structured Q&A Content
Google's documentation has said for years that structured data doesn't guarantee an AI Overview citation, but it does help Google understand eligible page content more reliably, and that guidance hasn't changed with the 2026 rich-result cleanup. Schema was never a ranking shortcut. It's a clarity tool.
That clarity still matters because of how AI Overviews actually pull content. The system favors pages where a question and its answer are unambiguously paired, and FAQPage schema is a direct, machine-readable signal of exactly that pairing. Without it, the model has to infer where a question ends and the real answer begins from plain prose, which is a harder, noisier job than reading a tagged pair.
There's outside evidence this correlation holds beyond Google's own surface. A 2026 analysis reported that 82% of domains cited by ChatGPT use schema markup of some kind, which suggests structured markup tracks with AI citation even on engines that have nothing to do with Google's rich results at all.
One industry guide went further and reported a 30 to 40 percent AI visibility uplift, but tied it to Article, FAQPage, and HowTo schema working together, not FAQ markup sitting alone on a page. That's a meaningful distinction for anyone deciding where to spend an afternoon of dev time.
FAQ Schema vs Other Structured Data: What Still Earns Its Keep in 2026
Not every schema type lost the same amount of value when Google pulled the visual snippet. Some never depended on it in the first place. Here's how we'd prioritize structured data for a Las Vegas small business site heading into next year.
| Schema type | Still shows a Google rich result? | Relevance to AI Overview / chatbot parsing | Priority for local small business sites |
|---|---|---|---|
| FAQPage | No (removed May 2026) | High, pairs questions with answers unambiguously | Medium, use where content genuinely fits |
| Article | Limited | High, establishes authorship and topic context | High, foundational for most content pages |
| HowTo | Reduced in some verticals | High, structures sequential steps clearly | High, for process and tutorial content |
| Service | Yes, in relevant verticals | Medium, clarifies what you actually offer | High, for service pages |
| LocalBusiness | Yes | Medium, anchors entity and geo data | High, for local SEO and GBP alignment |
The pattern across this table is consistent: nothing on it works as well alone as it does stacked. FAQPage paired with Article or Service schema gives an AI engine both the surrounding context and the direct answer pairing, which is closer to what that reported 30 to 40 percent uplift was actually measuring.
How FAQPage Markup Helps ChatGPT and Perplexity Even Without a Google Snippet
This is the point that gets lost in the "Google killed my rich result" panic: ChatGPT and Perplexity never rendered Google's rich results to begin with. They crawl and parse your page directly, which means Google's cosmetic removal has exactly zero effect on how those two surfaces read your site.
What those surfaces do care about is whether your question-and-answer blocks are clean enough to lift and attribute. A large language model scanning a page full of prose has to do extra inference work to figure out where a genuine answer to a genuine question starts and stops. A tagged FAQPage block hands that structure over for free.
This is the real reason FAQ schema's value outlived the Google snippet: the schema was always a parsing aid, and parsing matters more to AI engines than it ever did to classic SERPs. The rich result was a bonus. The parsing assist was always the actual job.
What We Actually Recommend for Las Vegas Small Business Sites Right Now
We're not telling clients to rip out FAQ schema, and we're not telling them to bolt it onto every page either. Here's the short version of what we're actually doing on client sites right now:
- Keep FAQPage schema only on pages where the visible content is genuinely structured as questions and answers; don't force it onto a service page that isn't actually written that way.
- Pair it with Article schema, and with Service or LocalBusiness schema where it fits, instead of treating FAQ markup as a tactic that stands on its own.
- Write the visible answers at 40 to 90 words, specific and self-contained, in the same plain language you'd want an AI engine to quote back verbatim.
- Reset your expectations on what "working" looks like. You're not chasing a rich-result bump anymore. You're chasing a parsing and citation bump that shows up as referral traffic from AI tools, not a screenshot of a blue-link accordion.
That last shift in expectations matters more than any of the technical steps. If you're still measuring FAQ schema success by SERP screenshots, you're measuring the wrong thing in late 2026.
A Technical Note on Diagnosing What's Actually Hurting Your AI Visibility
Schema is necessary, but it is not sufficient. If the pages carrying that careful FAQPage markup are slow, unindexed, or buried three clicks deep behind bad internal linking, no amount of structured data fixes the underlying problem.
We saw this directly on one of our own client indexation projects: 39 of 51 submitted pages reached indexed status once we cleaned up robots directives and fixed sitemap errors. Word count alone didn't close the remaining gap; internal linking did. That's the kind of groundwork our AI search readiness work focuses on before we even touch FAQ copy.
The same logic applies to page speed. If your mobile experience is sluggish, AI crawlers and human visitors both bounce before your well-marked-up FAQ ever gets read, which is a Core Web Vitals problem our web design services page is built to address. For a deeper look at what structured data Google and AI engines actually pull versus what gets ignored, our post on what structured data actually gets pulled into AI search results walks through the mechanics in more detail.
FAQ
Does FAQ schema still help SEO after Google removed FAQ rich results in 2026? Google removed the visual FAQ rich snippet on May 7, 2026, but FAQPage markup is still valid and still helps Google and other engines parse your content accurately. It won't earn extra SERP space anymore, but it can still support AI Overview and chatbot citation by making question-answer pairs unambiguous to machines.
Why did Google remove FAQ rich results in May 2026? Reporting on the change indicates Google removed FAQ rich results because AI Overviews had already taken over that visual real estate for most informational searches, making the old accordion snippet redundant in the SERP. The underlying schema type was not deprecated, only the rendered snippet.
Should I remove FAQ schema from my website now that it's not shown in search results? No. Removing valid FAQPage markup doesn't help you and may reduce clarity for AI crawlers that still parse structured data directly. Keep it on pages where the content is genuinely structured as questions and answers, and pair it with Article or Service schema for fuller context.
Does FAQ schema guarantee my page gets cited in an AI Overview? No source, including Google's own documentation, claims schema guarantees AI Overview inclusion. It's one input among many that helps Google and other AI systems understand page content; ranking well, answering clearly, and page speed all factor in alongside structured data.
What schema types matter most for AI search visibility in 2026? Industry guides point to Article, FAQPage, and HowTo schema working together rather than any single type in isolation, often cited alongside a reported 30 to 40 percent AI visibility uplift when combined. LocalBusiness and Service schema add further context for geo-specific small business pages.
If you're not sure whether your own FAQ markup is still doing its job or just sitting there unread by anything, run it through our AI visibility audit and we'll show you exactly which pages are parsing clean and which ones are quietly invisible to the engines that matter now.
