Most businesses already answer the same questions again and again: Do you deliver? How long does it take? What does it cost? An FAQ section puts those answers on a page. FAQ schema goes one step further. It adds a short piece of code that tells machines, in an agreed format, exactly which text is the question and which is the answer.
What FAQ schema is
Schema.org is a shared vocabulary for describing things on web pages. One of its types is FAQPage, which describes a page that lists questions with answers. Each question is a Question and each answer is an Answer.
The usual way to write it is JSON-LD: a script block placed in the page that visitors never see. It is one form of structured data. A reader of the page sees your FAQ section as normal. A machine reading the code sees the same content, labelled.
A small example
Here is a made-up block for the imaginary bakery used in the llms.txt guide. The business and its answer are invented for illustration.
<script type="application/ld+json">
{
"@context": "https://schema.org",
"@type": "FAQPage",
"mainEntity": [
{
"@type": "Question",
"name": "How far ahead should I order a custom cake?",
"acceptedAnswer": {
"@type": "Answer",
"text": "Please order at least three days ahead."
}
}
]
}
</script>The same question and answer must also be on the page itself. If you would rather not write the block by hand, the free FAQ schema generator turns a list of questions and answers into valid markup in your browser.
What it can and cannot do
What it can do is remove guesswork. A page with ten headings and several paragraphs leaves a machine to work out which lines are questions. Markup states it outright. That is useful to any tool that chooses to read it, and it forces you to keep each answer short and self-contained, which is good practice anyway.
What it cannot do is make anyone use your answers. Google has limited its FAQ rich result, the expandable questions that once appeared under a search listing, to a small set of well-known government and health sites, so most businesses should not expect it. AI engines publish no rules saying they rely on FAQ markup, and Omnishown does not claim to know which do. And markup cannot rescue a weak answer: a vague reply is still vague when it is labelled.
When to use it
Use it on a page that already has a real FAQ: a service page, a product page, a pricing page, a contact page with practical questions. It suits questions with short, stable answers. It does not suit a forum thread where users post their own answers; schema.org has a different type, QAPage, for that.
Skip it where there is no FAQ, or where the answers change so often that you will not keep the markup current. Markup that disagrees with the page is worse than none.
Rules that keep it honest
| Rule | What it means |
|---|---|
| Match the page | Every question and answer in the markup appears on the visible page in the same words. |
| One set per page | Mark up the FAQ that belongs to that page. Do not copy the same block onto every page. |
| Real questions | Use questions customers actually ask, not keyword phrases written to be found. |
| Plain answers | Start with the direct answer in the first sentence. Keep links and detail for after it. |
| No advertising | An FAQ is for answering, not selling. Do not use answers as a place for promotional copy. |
How to add and check it
- Write the visible FAQ first. Five to eight questions in your customers' own words, each answered in the first sentence.
- Generate the markup from the same list. Use the generator, or build it from the same data that renders the page, so the two cannot drift apart.
- Place it on the page it describes. Paste the script block into that page, in the head or the body.
- Check that it parses. Run the page through a schema validator, such as the Schema.org validator or Google's Rich Results Test, and fix any error it reports.
- Compare with the page. Read the visible FAQ next to the markup. Every question and answer should be in both.
Keeping it true
The risk with FAQ schema is the same as with any hand-kept file: it goes stale. A price changes on the page, the markup still says the old one. Build it from the same data as the visible FAQ if you can, and recheck it after each release.
That is the sort of check Omnishown Check is built for. It reviews each push to your repository for SEO, AEO and GEO issues, and Omnishown Fix sends any proposed change as a pull request that a person approves. Omnishown Diff shows what changed after each release. To see whether AI engines name you, read how to check by hand, or look at Omnishown Watch. For crawler access, see the robots.txt guide. The AI visibility guide has the wider checklist, and you can join the pilot to try Omnishown on your own site.