Adding a FAQ section can look like a simple way to make a page more useful to buyers—and perhaps more visible in AI answers. Those are separate questions. Relevant, visible answers can address buyer concerns, but the supplied evidence does not establish that FAQ content earns more AI citations. FAQ schema, a way to describe question-and-answer content in structured data, is not a proven shortcut either. Google's rules for FAQ rich results concern a separate search display, not AI citations.
Do FAQ content or FAQ schema improve AI visibility?
Useful visible question-and-answer content can address real buyer needs, but the supplied evidence does not establish a direct FAQ-to-AI-citation effect. FAQ schema has no demonstrated AI citation advantage in the supplied sources; Google's historical 2023 FAQ rich-result restriction and subsequent 2026 discontinuation concern a different search feature. Neither answer means FAQs cannot help users or that schema has no value in every context.
Intervention or comparison condition | Buyer-facing effect | Google FAQ rich-result eligibility | Demonstrated AI-citation effect in supplied evidence |
|---|---|---|---|
Visible FAQ content without FAQPage markup | Can answer specific buyer questions and reduce friction when relevant and accurate | FAQ rich results no longer appear; useful visible content does not depend on markup | No direct effect established |
FAQPage structured data matching visible content | Does not replace or improve the visible answer itself | Google stopped showing FAQ rich results on May 7, 2026; the government/health restriction was its earlier 2023 policy | No FAQ-specific effect established |
Other JSON-LD structured data | Usually describes page information for supported search features and systems | Eligibility depends on the type of markup and current policy | One matched vendor study did not find a major uplift from schema generally; it did not isolate FAQ schema |
Comparison baseline: no added FAQ or schema | Leaves the page's existing buyer information unchanged; no new FAQ answer is added | FAQ rich results no longer appear under any comparison condition | Serves as a baseline in a test; it does not establish an independent effect by itself |
These distinctions matter because a result can be useful without earning a special Google display, and a markup implementation can be technically valid without causing an AI system to cite the page. Treat user value, rich-result eligibility, and AI citations as separate outcomes. For example, a support team could assess whether the FAQ resolves a recurring question, while an SEO team separately tracks organic clicks and a content team records citations across a defined set of queries. Those observations answer different questions; combining them into a single “visibility” result would obscure what changed.
What changed for Google's FAQ rich results in 2023?
In August 2023, Google said FAQ rich results would appear only for well-known, authoritative government and health websites. Other sites would no longer see them regularly. Google also said those sites did not need to proactively remove existing FAQ markup. This was a change to a Search result display, not a test of visible FAQ content or AI answer behavior. Google's August 2023 announcement
The 2023 policy did not mean every government or health site automatically qualified. Losing a rich-result display also does not make on-page answers pointless: visitors can still use them. Keep the user and search-display questions separate when deciding what to maintain. Google later discontinued FAQ rich results. Its May 2026 changelog says they stopped appearing on May 7, 2026; a June 15 entry records removal of the documentation. The 2023 restriction is historical context, not current eligibility guidance.
Should a B2B SaaS site remove existing FAQ markup?
Google's 2023 guidance did not require a B2B SaaS site to remove FAQ markup solely because FAQ rich results were unlikely to appear. FAQ rich results no longer appear for any site, so assess remaining markup for a separate, documented use. Keep it only when it accurately describes visible page content and complies with the applicable current rules; check Google's current documentation before publishing or changing an implementation. Do not treat markup as a promise of a rich result. Google's August 2023 announcement Google's AI-feature guidance
For a broader comparison, evaluate structured data and AI citations separately from visible FAQ content.
Does adding structured data cause more AI citations?
A matched Ahrefs vendor study of pages adding JSON-LD did not find a major citation uplift, but it did not isolate FAQ schema. The study compared 1,885 pages that adopted schema with 4,000 controls. AI Overview citations changed by -4.6% relative to controls. AI Mode (+2.4%) and ChatGPT (+2.2%) changes were statistically indistinguishable from zero. These are reported relative comparisons, not proof that schema caused a change or an FAQ-specific estimate. Ahrefs' matched JSON-LD study
JSON-LD is a format for expressing structured data in a page's code. The study is relevant as a reason not to assume that adding structured data alone will increase citations. It cannot answer whether FAQ markup, visible FAQ copy, or a particular implementation changes citations, because those questions were not isolated in the reported analysis. The study pooled schema types on pages already receiving substantial AI citations and measured 30 days after JSON-LD adoption; it does not establish an effect for previously uncited pages.
The practical conclusion is narrow: do not present FAQ schema as an established AI-citation tactic. If structured data serves a separate, supported purpose, assess that purpose and implementation on its own terms. The same caution applies to AI citation rate measurement: a markup type alone does not establish a citation effect.
What does Google require for AI-feature eligibility?
Google's guidance says sites are subject to the usual Search eligibility requirements for its AI features and that there are no additional technical requirements specifically for those features. That does not guarantee appearance, and it does not establish FAQ markup as an AI ranking factor. Google's AI-feature guidance
In practice, this points teams toward crawlable, indexable pages with accurate, useful content—not an invented “AI schema” checklist. Structured data may help describe content for supported search features, but it is distinct from whether a page can be found, selected, or cited in a particular AI answer. Technical eligibility is a condition, not a placement guarantee.
When is an FAQ useful on a B2B SaaS page?
Add a visible FAQ when it resolves a specific buyer concern with an accurate, concise answer. Do not add generic questions merely to create more markup. Keep answers accessible, current, and consistent with documented product behavior; where a question depends on a plan, integration, or configuration, state that condition.
For example, an illustrative procurement question might be: “What information should our security team review before approving this platform?” The answer should point to the actual, approved security documentation and describe any limits that apply. This is a hypothetical example, not a customer question or a claim about Nortav's product.
Use this checklist before adding a question:
Question evidence: Is it grounded in buyer research, sales conversations, support themes, or a recurring task?
Direct answer: Can the page answer it accurately without hiding a material condition?
Ownership: Who is responsible for confirming the answer remains correct?
Update trigger: What product, policy, pricing, or process change should prompt a review?
Accessible rendering: Can users read the answer in the page itself, including with assistive technology?
If those checks are difficult to satisfy, the question may belong in a maintained support resource—or may not need a standalone FAQ at all.
What makes a good answer-first FAQ response?
Answer the exact question in the first two to four sentences, then add a relevant condition or limitation and link to supporting detail where appropriate. Define technical terms when they first appear, and use self-contained bullets when three or more conditions need explanation. This format makes answers easier to scan; it does not mechanically earn AI placement.
How should teams test FAQ content and FAQ markup separately?
Compare the effect of visible FAQ content separately from the effect of structured data, and track buyer behavior, organic clicks, and AI citations as distinct outcomes. A change in Google rich-result display is not evidence of an AI-citation gain. Because the supplied sources do not establish the independent causal effect of FAQ content or FAQ schema on citations, treat a test as an investigation with uncertainty, not as a guaranteed growth experiment.
Identify buyer questions. Use documented customer research, sales or support themes, and on-page search behavior where available. Record why each question belongs on the page.
Choose comparable pages. Match pages on topic, intent, and relevant traffic or performance characteristics. Record differences that could affect the comparison.
Separate the interventions. Compare pages with and without visible FAQ content. If testing valid FAQ markup, vary that separately and only where the page and current policy make it appropriate.
Validate implementation. Confirm that visible answers are accurate and accessible, markup matches the visible page, and any applicable structured-data rules are met. Do not assume a validation tool predicts citations.
Observe for a fixed period. Set the period and measurement method before reviewing results. Track organic clicks, buyer task completion, and AI citations separately, using consistent denominators and repeated observations where feasible.
Audit answer accuracy. Check product, policy, and other time-sensitive details during and after the test; remove or update stale answers.
Report uncertainty. Note sample limitations, policy eligibility, and confounding changes such as site releases or shifts in demand. Do not attribute every change to the FAQ or markup treatment.
A useful report should make clear what was counted, which prompts or queries were observed, and how repeated runs were handled when evaluating AI citation rate measurement.
The practical decision
Publish FAQs when they help buyers resolve real questions, not to manufacture markup. Use structured data only when it accurately describes visible content and is appropriate under current rules; treat rich-result eligibility and AI citations as separate. Measure buyer outcomes, organic clicks, and AI citations independently, and report uncertainty. A FAQ format test can inform the decision, but the evidence here does not establish a citation shortcut.
There is also a broader search-traffic trade-off to monitor. In a follow-up analysis of 300,000 keywords, Ahrefs reported that searches showing an AI Overview were associated with a 58% lower average desktop click-through rate for the top-ranking page than its comparison estimate. The analysis used aggregated Google Search Console data for December 2023 and December 2025. This is an observational, cross-keyword result—not an estimate of FAQ effects or the value of being cited. Assess it alongside your own traffic and conversion data rather than treating it as a forecast.
Frequently Asked Questions
Do FAQs improve SEO?
A relevant FAQ can answer a buyer's question, but the supplied evidence does not establish that FAQ content by itself improves rankings or AI citations. Add one when it resolves a real user need, and measure outcomes separately from markup or rich-result eligibility.
Is FAQ schema still relevant?
Google stopped showing FAQ rich results on May 7, 2026, according to its official changelog. The government/health restriction was the earlier 2023 policy. The supplied evidence does not establish an AI-citation benefit from FAQ schema; assess any other implementation purpose separately.
How to optimize your content for AI search?
Start with accurate, useful answers to real buyer questions, make the content accessible and maintain it when facts change. Google describes its AI features as using ordinary Search eligibility requirements, with no additional AI-specific technical requirements; neither answer formatting nor structured data guarantees a citation.
How Nortav Can Help
For B2B SaaS teams deciding whether to add visible FAQs or structured data, Nortav can help separate buyer needs from technical eligibility. We can identify evidence gaps and plan measurement that keeps user outcomes distinct from AI citation observations.
Our approach starts with understanding your current search visibility, identifying gaps against competitors, and developing a strategy focused on driving qualified traffic, generating leads, and delivering measurable business value.
Book a Discovery Call →
Sources
Google Search Central, “AI features and your website.”
Ahrefs, “We Tracked 1,885 Pages Adding Schema. AI Citations Barely Moved.”
Google Search Central, “Changes to HowTo and FAQ rich results.”
Ahrefs, “Update: AI Overviews Reduce Clicks by 58%.” -
Google Search Central, “Documentation updates: FAQ rich-result deprecation and documentation removal (May–June 2026).”
