FAQPage is the Schema.org type that encodes a list of questions and authoritative answers in machine-readable format. For AI citation, it is the single most impactful piece of structured data you can add to a page, because it directly maps to the question-and-answer format that AI retrieval systems use to generate responses.
When a user asks ChatGPT "what is the best way to do X?", the retrieval system looks for pages that have encoded answers to that question. FAQPage schema is a direct signal: "this page authoritatively answers these questions."
The basic FAQPage JSON-LD structure
Implementation in Next.js (App Router)
Writing good FAQ content for AEO
- Write the question exactly as a user would phrase it to an AI engine
- Start the answer with a direct, complete response, not "it depends" or "there are many factors"
- Keep answers between 40-150 words, long enough to be useful, short enough to be a passage
- Use the same terminology in the answer as in the question (for retrieval matching)
- Do not repeat the question text in the answer
- Include 4-8 FAQs per page: more than 10 dilutes quality signals
Where to add FAQPage schema
FAQPage schema should be added to: your homepage (with FAQs about your product/service), your product or pricing pages (with questions buyers ask), blog posts (with topic FAQs at the bottom), your FAQ page (obviously), and any landing pages targeting question-based queries.
Do not add FAQPage schema to pages where the questions are not actually answered on the page. Mismatched schema (schema that claims answers exist but they do not) is a quality signal violation and can reduce citation confidence.