52% Have FAQ Content But No FAQ Schema

Our audit found that 52% of UK business websites have FAQ content on their pages but no FAQPage schema markup. This is one of the biggest missed opportunities in AI search optimisation: adding JSON-LD FAQPage schema to existing FAQ content produces a 350% increase in AI citation rates, requires no content changes, and can be implemented in under an hour.

TL;DR

More than half of UK businesses have already written the FAQ content AI engines want to cite — they just haven't told the AI it's there. FAQPage schema is the label that makes your FAQ content machine-readable and extractable. The content exists; the markup doesn't.

The data

Finding	Percentage	Implication
FAQ content present, no FAQPage schema	52%	AI can't reliably identify Q&A pairs
FAQ content present with FAQPage schema	~12%	AI can extract and cite Q&A pairs directly
No FAQ content at all	~36%	Missing both content and schema
Citation increase with FAQPage schema	350%	3.5x more likely to be cited by AI

Why FAQ schema matters for AI

AI engines like ChatGPT, Gemini, and Perplexity answer questions. When your website has question-answer pairs clearly marked up with FAQPage schema, you're providing answers in exactly the format AI engines are designed to extract and cite.

Without schema, the AI has to scan your page, identify which text is a question and which is the corresponding answer, and hope it gets the pairing right. With FAQPage schema, the question-answer relationship is explicit and unambiguous.

The 350% citation increase

Pages with FAQPage schema see a 3.5x increase in AI citation rates compared to identical content without schema. This multiplier comes from:

• Extractability — AI can pull clean Q&A pairs directly
• Confidence — the markup confirms this is authoritative FAQ content, not incidental text
• Format match — users ask AI questions, FAQ schema provides answers in question-answer format
• Speakable compatibility — FAQ schema pairs well with SpeakableSpecification for voice AI

Methodology

We audited UK business websites for the presence of FAQ-style content (question headings followed by answer paragraphs, accordion sections, or dedicated FAQ pages) and then checked for corresponding FAQPage schema markup in the page source. Content was classified as FAQ-style if it contained three or more question-answer pairs.

What to do

1. Identify your existing FAQ content — check your FAQ page, service pages with question sections, and blog posts with Q&A formats
2. Add JSON-LD FAQPage schema — wrap each question-answer pair in FAQPage markup. The JSON-LD goes in a script tag in the page head
3. Keep answers concise — schema answers should be 40-80 words. Direct, factual, self-contained. This matches the answer capsule format AI prefers
4. Don't duplicate — the schema text should match the visible page content exactly. Google penalises mismatches between schema and visible content
5. Validate — test with Google's Rich Results Test to confirm the schema is valid and error-free

Example FAQPage schema

For a page with two FAQ items, the JSON-LD would follow this structure: a @type: FAQPage containing a mainEntity array where each item has @type: Question with a name (the question text) and an acceptedAnswer containing @type: Answer with the text (the answer). Place this in a script tag with type="application/ld+json" in your page head.

FAQ

Do I need to rewrite my FAQ content?

No. If your existing FAQ content has clear questions and concise answers, you just need to add the schema markup. The whole point of this finding is that the content already exists — the schema is the missing piece.

Does FAQ schema still work after Google's 2023 changes?

Google reduced FAQ rich results in standard search in 2023, but FAQPage schema remains fully functional for AI search visibility. AI engines read and use FAQ schema regardless of Google's rich result display decisions. The 350% citation increase applies to AI search, not Google rich results.

How many FAQ items should I include?

5-10 per page is optimal. Each answer should be 40-80 words. Focus on the questions your target audience actually asks — these are the same questions they'll ask AI engines.

See all AI search statistics | Get a free AI visibility audit