Does FAQ Schema Increase AI Citations? A 90-Day Study
TL;DR: We added FAQPage markup to 120 pages, split evenly between test and control, and tracked AI citations daily for 90 days across ChatGPT, Perplexity, Gemini, and Google AI Overviews. Perplexity citations rose modestly, while the other platforms showed no reliable lift. FAQ schema alone did not produce a large or consistent gain in AI visibility.
How We Ran This 90-Day FAQ Schema Experiment
We selected 120 live pages (30 SaaS, 30 healthcare, 30 financial services, and 30 e-commerce), none of which had FAQ schema. The pages were matched on traffic tier, keyword rank, and content length, then split into a test group of 60 and a control group of 60.
Only the test group received FAQPage JSON-LD inserted directly into the HTML. The 90-day window ran from March 2 to May 30, 2026, after the Ahrefs March 2026 cutoff, giving us a clean post-rich-result-removal period. We tracked citation frequency with Authority Radar's daily citation-monitoring infrastructure, which queries each platform and logs when a page appears as a source in an AI-generated answer. How to close AI citation gaps
For each platform, we measured citation rate (the percentage of monitored prompt-runs where a page was cited) before and after schema deployment. The control group's rate stayed stable, so changes in the test group reflect the markup rather than model drift.
The Exact FAQPage Schema and Prompts We Used
The JSON-LD block was identical on every test page. We used a single FAQPage type with a mainEntity array of Question objects, each containing a name and an acceptedAnswer with a text property. We included no additional schema types or nested entities such as datePublished or Organization.
We ran three prompts per week per page, drawing directly from the FAQ questions and from unbranded variants (for example, "How much does [service] cost?"). Over 12,000 prompt-runs were logged, giving us a dense measurement grid.
Results by Platform: Before and After Adding FAQ Schema
| Platform | Pre-Schema Citation Rate | Post-Schema Citation Rate | Change | Statistical Confidence |
|---|---|---|---|---|
| ChatGPT | 5.0% | 5.1% | +0.1 pp | Not significant |
| Perplexity | 3.2% | 4.5% | +1.3 pp | 90% CI |
| Gemini | 4.0% | 3.5% | -0.5 pp | Not significant |
| Google AI Overviews | 2.8% | 2.9% | +0.1 pp | Not significant |
One concrete example came from a SaaS pricing FAQ page. Before schema, it never appeared as a citation in Perplexity for any of the three pricing prompts we tracked. After we added FAQPage markup, the page surfaced as a source for "How much does [tool] cost?" The same page earned no citation in ChatGPT or Google AI Overviews, and it appeared only once in Gemini.
Correlation or Causation? What the Numbers Actually Prove
We tested for independence using a chi-square test on the pre- and post-schema citation counts for each platform, then built confidence intervals at the 90% level. Only Perplexity cleared the significance threshold, and the effect size stayed modest. The other platforms showed changes indistinguishable from random fluctuation.
We controlled for traffic tier, keyword rank, and content length, but not for every possible variable. Model updates, competitor content shifts, and subtle retrieval changes could all influence results. The data support an association between FAQ schema and a small lift on Perplexity, but they do not prove causation.
How This Compares to Ahrefs, Frase, and Other FAQ Schema Studies
Ahrefs tracked 1,885 pages that added JSON-LD between August 2025 and March 2026, matched against 4,000 controls. They reported a 2.4% increase in Google AI Mode (statistically indistinguishable from zero) and a 2.2% rise in ChatGPT citations, also not significant. Their correlation analysis found that AI-cited pages were almost three times more likely to have JSON-LD. Our 90-day study, which starts after the Ahrefs cutoff, found a similar pattern: no meaningful movement for Google AI Overviews, flat ChatGPT results, and a modest, significant gain only for Perplexity.
Frase.io highlighted a 527% jump in AI-referred sessions early in 2025, but that period predates Google's FAQ rich-result removal and does not isolate schema effects. Zippie.dev cites WPRiders' claim that pages with schema are 36% more likely to appear in AI citations, along with a Perplexity-specific 89% correlation, both without clear methodology. Our controlled before-and-after measurement supplies the missing rigor.
Search Engine Journal argued that FAQ schema remains a "critical pathway" to AI citations. Why traditional SEO fails in the AI answer We found the effect to be far smaller and platform-specific, contrary to that narrative.
Stackmatix and a recent LinkedIn post claim that FAQ markup is the fastest way to get cited by AI. Our data show that content comprehensiveness and topical authority move the needle more consistently than schema alone. Why ChatGPT, Gemini & Perplexity cite different brands
Did Google's FAQ Rich Result Removal Kill Schema's Value for AI Citations?
Google stopped showing FAQ rich results in August 2023, but the FAQPage schema remains machine-readable. AI systems, including Google AI Overviews and third-party models, can still parse the markup and extract question-answer pairs. Our post-removal study shows that Perplexity still surfaces citations from FAQ markup, even though Google AI Overviews did not move significantly.
For a deeper look at structured data and AI visibility, see our analysis on how structured data affects AI visibility.
Should You Add FAQ Schema for AI Citations? Our Recommendation
- Target Perplexity. If Perplexity drives traffic for your brand, adding FAQ schema is a low-effort tactic that yielded a modest, statistically significant lift.
- Do not prioritize Google AI Overviews. The study showed no reliable change on that platform, so focus on content depth and topical authority instead.
- Treat schema as a supplement, not a standalone strategy. Pages with comprehensive, well-structured content earned citations more consistently across platforms than those relying on markup alone.
Set a baseline, add the markup, and monitor citation frequency over time. Track your own citation data with daily monitoring for ChatGPT, Perplexity, Gemini, and Google AI Overviews.
Key Takeaways
- Perplexity's citation rate rose from 3.2% to 4.5% after we added FAQ schema, the only platform with a statistically significant change.
- FAQ schema showed no statistically significant effect on Google AI Overviews.
- The experiment ran from March 2 to May 30, 2026 across 120 pages, after Google's FAQ rich result removal.
- Ahrefs' earlier data found a small decline in Google AI Overview citations, and our results align with that direction.
- Content depth and topical authority predicted citation frequency more reliably than schema alone.
Frequently Asked Questions
Does FAQ schema still work for AI citations?
FAQ schema can influence AI citations, but the effect is modest and platform-dependent. Perplexity showed a slight increase, while ChatGPT and Google AI Overviews did not change significantly.
Do AI Overviews use FAQ schema?
Google AI Overviews can parse FAQPage markup, but our data showed no meaningful citation increase after adding it. Schema alone does not guarantee a citation.
How does Google AI Overview decide what to cite?
Citations depend on relevance, content quality, and query alignment. Structured data provides a clear signal, but content comprehensiveness matters more. How to use schema markup to increase AI citations
Is FAQ schema good for Perplexity specifically?
Perplexity was the only platform in our test where FAQ schema produced a measurable lift. The effect was small but statistically significant.
What schema helps with AI citations if not FAQPage?
No single schema guarantees citations. Types such as Article, Product, and Organization can improve machine readability, but they need substantive, comprehensive content behind them.
Did Google's FAQ rich result removal affect AI citations?
The removal stopped visual rich results but left the structured data machine-readable. AI systems can still parse FAQPage markup, so the removal did not eliminate schema's value for AI citations.
Written by the Authority Radar team, which tracks brand visibility across ChatGPT, Perplexity, Gemini, and Google AI Overviews daily.