FAQ Rich Results Are Gone: Should You Delete FAQ Schema?
The decoration in search results may be gone, but you still need the power to answer questions
Why Google scaled back FAQ rich results and how to respond: whether to delete FAQ schema, what to do instead, and bottom-of-page FAQs vs. in-body Q&A blocks.
With Google scaling back FAQ rich results, FAQs need to be redefined not as decoration in search results but as a Q&A evidence structure within the body. Google sharply reduced where HowTo and FAQ rich results appear in August 2023 [1], and the trend has only become clearer over time. For practitioners, the important question isn't "Should we keep writing FAQs?" but "For what purpose, and where, should FAQs go?"
What changed, and when
This change isn't sudden news; it's an extension of a trend that was already underway.
| Period | Change in search results | What it means for practitioners |
|---|---|---|
| Before 2022 | FAQPage schema worked as a way to claim SERP space | Markup alone could be expected to improve click-through rate |
| August 2023 | Google officially scaled back HowTo and FAQ rich results [1] | Practitioners began to feel that markup wasn't as effective as before |
| Now | FAQs are redefined as a content structure issue, not SERP decoration | FAQs need to be redesigned as trust signals in the body |
Users still ask questions. What's changed is where those questions are seen. The center of gravity has shifted from accordions under search results to question recombination inside answer-style interfaces such as AI Overviews, PAA, ChatGPT, and Perplexity.
Don't throw out structured data; split its roles
You don't have to delete FAQPage schema outright. But the explanation that "adding FAQ schema makes an accordion appear in search results" is no longer accurate. It's safer to re-split the purposes and decide for each.
| Purpose | Current judgment |
|---|---|
| Showing up as Google FAQ rich results | Hard to expect anymore |
| Helping Google understand page meaning | May still matter as a supplementary signal, but doesn't guarantee visibility [2] |
| Korean search such as Naver | Consider keeping it depending on the nature of the site |
| Internal search and content reuse | Well-organized FAQ data helps CS, sales, and internal search |
| AI search | Strengthen the questions, answers, and evidence visible in the body rather than relying on JSON-LD alone |
The key isn't deletion; it's priority. The right order isn't to add markup first and fit the body to it, but to build real questions, answers, and evidence in the body first and then attach structured data as a supplement.
What makes a good Q&A evidence block
Old FAQs usually looked like this.
> Q. Does adding FAQ schema help with search? > A. Yes, it can appear as a rich snippet and help improve click-through rate.
That answer is now risky. It creates an expectation that no longer holds, and it gives readers no real criteria for making a decision. A good Q&A evidence block needs at least the following elements.
| Element | Role |
|---|---|
| A real question | Close to the wording customers actually search or ask in consultations |
| A short answer | So the conclusion is clear from the first sentence alone |
| Reasoning | Explains why you can say so |
| Source | States whether it relies on official documentation, data, or internal standards |
| Update reference | A date so readers can tell whether the answer is outdated |
| Next action | Links to a related post or product page |
Rewrite the same question by these criteria and it changes like this.
> Q. Should we keep using FAQPage schema? > A. If the goal is to appear as Google FAQ rich results, its priority has dropped. Google has officially said FAQ rich results are no longer shown in search results [1]. FAQ-style content itself still matters, though. If you keep the schema, redefine its purpose around supplementing page meaning and internal search, not search result visibility.
The difference isn't sentence length; it's whether there's information you can make a decision with.
What to fix first, by page type
Rather than overhauling the entire site at once, it's more efficient to sort pages with FAQs by type and review them.
| Page type | What to check first | How to revise |
|---|---|---|
| Blog posts | Whether the bottom FAQ duplicates the body | Move important questions up into body sections |
| Product and pricing pages | Whether questions that block conversion are hidden | Answer pricing and adoption conditions directly mid-body |
| Customer support docs | Whether answers are too short or outdated | Add update dates, reasoning, and exceptions |
| Comparison content | Whether questions read only like brand claims | State objective criteria and limitations together |
Example: turning a bottom-of-page Q&A into an in-body evidence block
Say a B2B intro post has an FAQ like this.
> Q. Can this work for our company too? > A. Yes, it can be applied across a wide range of industries.
This answer lacks evidence, so neither readers nor AI can reuse it as is. Move it up into a body section that explains adoption conditions, and it changes like this.
> Q. Can this work for our company too? > A. Whether it works depends more on your data structure and operating processes than on your industry. If your customer inquiry data is organized and you have approval criteria for answers, you can adopt it quickly. On the other hand, if your data is scattered across several departments, you need to organize your information structure first.
With this change, the FAQ stops being decoration and becomes a small diagnostic tool readers use to judge their own situation. This structure points in the same direction as GEO research finding that content with inline citations and reasoning is chosen more often for AI citations [3].
In the Korean market, interpret this a bit more carefully
Carrying Google's change straight over as the conclusion for the Korean market can lead to a shortsighted decision. Korea's search environment doesn't end with Google. Naver search results, brand searches, and the blog and cafe ecosystem all move together. Even for B2B services, users often look on Google first, search the company name again on Naver, and then go through blog posts before reaching out.
So the conclusion that "FAQ rich results are over on Google, so FAQs are useless" is risky. The more realistic conclusion is this: FAQs have lost power as search result decoration, but as a question structure that builds trust, they've become even more important. The gap is especially large for Korean-language content. Advice that reads like a straight translation of English SEO docs, such as "Add an FAQ" or "Apply schema," points the right way but isn't enough for readers to act on right away. It only becomes a usable standard when you get down to which questions your customers actually ask before buying, whether you can answer each in one sentence, and whether the evidence for that answer is visible in the body.
What this change means for the whole team
This work isn't just for the content team. The sales team can cut down on repeated questions in consultations, and the customer support team can standardize answers. The product team can see where customers get stuck through FAQ data, and the content team can connect those questions to search intent and internal link structure. Instead of leaving the FAQ as bottom-of-page decoration, treat it as a question database the whole team shares. It then becomes an asset that stays useful far longer, regardless of whether rich results appear.
Common mistakes
- Removing FAQs altogether: Some conclude that "rich results are over, so FAQs are over too." More precisely, it means "the shallow tactic of chasing search result decoration is over."
- Leaving copy that still promises rich results: Be careful with explanations like "adding FAQ schema makes an accordion appear in search results."
- Leaving important questions at the bottom: Core questions that block conversion should be answered first in the relevant body section, not the bottom FAQ.
The disappearance of FAQ rich results doesn't mean the questions disappeared. If anything, now is the time to prove, through body structure, how well-grounded your answers to those questions are. If you want to know whether your site's Q&A is actually being used as evidence in AI search, you can check your current citation status with TRAIL Search.
For related reading, see Does schema markup increase AI citations? The problem with JSON-LD that doesn't match the body, What is GEO (Generative Engine Optimization)? Which pages to start with, and in what order, and A four-step unified framework for SEO and GEO: split the teams or combine them?.
Frequently asked questions
Should I remove FAQPage schema entirely now?
You don't have to. But you do need to change its purpose. If the goal is to show up as an accordion in Google search results, that's hard to expect anymore, but it's still useful for supplementing what the page means, for other search environments such as Naver, for internal search, and for reusing content.
Won't moving the FAQ into the body make the post too long?
You don't need to move all of it. Pull only the core questions that block purchase decisions and the questions that come up repeatedly into the flow of the body, and it's fine to leave the other secondary questions in the bottom FAQ as before. The key is to separate priorities.
How does this change affect AI search optimization?
AI search has always tended to rely more on the questions, answers, and evidence actually visible in the body than on JSON-LD markup. So the scaling back of FAQ rich results actually points in the same direction as AI search optimization: putting a clear Q&A structure in the body.
References
Summary
- Google scaled back FAQ and HowTo rich results in August 2023, and the trend became even clearer in 2026 [1].
- You don't need to delete FAQPage schema, but you should redefine its purpose: not search result decoration, but supplementing page meaning, internal search, and other search environments.
- A good Q&A evidence block has six elements: a question, a short answer, the reasoning, a source, an update reference, and a next action.
- It's safer to move the core questions that block purchase decisions out of the bottom FAQ and into the relevant body sections.
Produce the content this post covers from your brand standard
TRAIL Studio registers the brand standard once and produces cardnews, detail pages, product photos, shorts and blog posts from it. Generated output is reviewed and edited by a person before it ships.
More posts
Best AI Search Optimization Tools 2026: 6 GEO Tools Ranked
We compared six tools that check whether ChatGPT and other AI answers mention your brand, on measurement, action, engines, Korea support, and price.
Why Does ChatGPT Cite Listicles? Format vs. Hierarchy
Listicles split items cleanly, so AI can extract just what it needs. How to write citable listicles, whether format alone is enough, and how to build hierarchy.
What is GEO? Which pages to start with, and in what order
GEO structures content so generative search like ChatGPT and AI Overviews cites it. Does it replace SEO, what order to run it in, and how to measure it.
Keep reading on this topic
- Topical authority: why pillar-cluster beats a single postTopical authority comes from content clusters, not single posts: the pillar-cluster model, how to design one, why it wins in AI search, and when it's enough.
- A 4-step framework for SEO and GEO: split teams or merge?SEO and GEO work best on one content quality foundation. Whether to split or merge teams, a 4-step framework, why unified plans fail, and one shared metric.
- SEO and GEO top spots: high rankings, missing in AI answersReaching the top now covers how AI answers mention you, not just rankings. What changed, how to manage both tops, and why competitor comparison matters.
This post is part of the Content category, which collects all 41 posts on the topic. See all posts in Content