Does your site have FAQ schema markup? If you are not sure, the answer is probably no. And that means you are leaving one of the easiest wins in technical SEO on the table.
We audited a Belgian e-commerce site last quarter. They had 40 product pages, each with a hand-written FAQ section at the bottom -- good questions, clear answers, genuinely helpful. But none of it was marked up with FAQPage schema. Google had no idea those FAQs existed as structured data. Neither did ChatGPT.
We added the JSON-LD markup. Took about two hours for the full site. Within three weeks, 12 of their pages started showing FAQ rich snippets in Google results. Their average CTR on those pages went from 3.1% to 4.8%. No content changes. No new backlinks. Just markup.
Why this particular schema punches above its weight
FAQPage schema has the best effort-to-impact ratio of any structured data type. It is easy to implement (one JSON-LD block in your page head), it produces visible rich snippets that take up more space in search results, and -- this is the part most people miss -- it feeds AI models exactly what they want.
LLMs are trained on question-answer pairs. That is their native format. When your FAQ is structured in JSON-LD, you are handing AI models pre-packaged, extractable answers in the format they prefer. Andrea Volpini, CEO of WordLift in Rome, calls it "the Trojan horse of AI visibility" -- and he is right.
Sistrix (Bonn, 2025) found that pages with FAQ rich snippets have a 27% higher CTR than pages without them. Google AI Overviews regularly pull from structured FAQs. And Perplexity explicitly parses schema markup when building its citations.
The actual JSON-LD code
Here is what it looks like. This goes in a <script type="application/ld+json"> tag, ideally in your page's <head>:
{
"@context": "https://schema.org",
"@type": "FAQPage",
"mainEntity": [
{
"@type": "Question",
"name": "Your question here?",
"acceptedAnswer": {
"@type": "Answer",
"text": "Your answer, clear and concise."
}
},
{
"@type": "Question",
"name": "Second question?",
"acceptedAnswer": {
"@type": "Answer",
"text": "Second answer."
}
}
]
}Five rules that matter:
- Every question in the JSON-LD must also appear as visible text on the page. Google penalises hidden FAQs.
- Keep answers under 100 words. An FAQ answer is not an article.
- 5 to 8 questions per page is the sweet spot.
- Each FAQ must be relevant to that specific page, not copy-pasted across the site.
- Validate your JSON before deploying. A missing comma breaks the entire thing.
How it compares to other schema types
| Schema type | Rich snippet | AI citation impact | Implementation effort |
|---|---|---|---|
| FAQPage | Expandable questions in SERP | Very high | Low |
| HowTo | Numbered steps | High | Medium |
| QAPage | None currently | Medium | Low |
| Speakable | Google Assistant | High | Low |
FAQPage wins on the effort-to-impact ratio. HowTo is great for tutorials but requires more complex markup. QAPage is designed for forum-style content. Speakable is niche but worth watching.
Five mistakes that waste the whole effort
1. Invisible FAQs. Your JSON-LD references questions that are nowhere in the visible page content. Google detects this and ignores the markup.
2. Copy-pasted FAQs. The same five questions on every page of your site. Each FAQ needs to be unique and relevant to its specific page. We see this constantly on WordPress sites using plugins that auto-generate identical FAQ blocks.
3. Answers that are actually articles. If your FAQ answer is 500 words long, it is not an FAQ -- it is a blog post pretending to be one. Stay under 100 words per answer. Be direct.
4. Promotional questions. "Why is our product the best choice?" is not a real question anyone asks. Use AlsoAsked.com or Google's People Also Ask to find the questions people actually type.
5. Invalid JSON. A missing comma, an unescaped quotation mark, a trailing comma after the last item -- any of these silently breaks the entire schema. Always validate with Schema.org Validator or Google's Rich Results Test before deploying.
Where to find the right questions
Three sources, in order of reliability:
Your own clients. What do they ask in emails, on calls, in discovery meetings? These are the real questions. We keep a shared document where we log every client question -- it is the best FAQ source we have.
Google's People Also Ask. Search your target keyword and look at the PAA box. Click on questions to expand more. There are usually 30 to 50 related questions for any broad topic. Free, immediate, and based on actual search behaviour.
AlsoAsked.com. Maps the full tree of related questions for any keyword. The free tier is enough to get started. The visual tree structure helps you see which questions cluster together -- useful for planning FAQ sections across multiple pages.
One more thing. Google reduced FAQ rich snippet display in 2023, and some SEOs declared FAQ schema dead. They were wrong. Google has been bringing rich snippets back for certain sectors, and more importantly, the AI citation benefit of structured FAQs has grown since then. The format is more valuable now than it was in 2023, just for different reasons.
We implement FAQ schema across your site in under a week.
Targeted questions, valid JSON-LD, visible on-page content. Measurable CTR lift within a month.
Get FAQ schema on my site
