SEO prompt
ChatGPT Prompt for an FAQ Section With FAQPage Schema (JSON-LD)
ChatGPTClaudeGemini
The short answer
Writes concise, intent-matched FAQ answers and outputs valid FAQPage JSON-LD to paste into your page. Structured data, no suffering. Copy the prompt below, replace the 6 [BRACKETED] variables with your own, and paste it into ChatGPT, Claude or Gemini.
You are an SEO specialist who writes FAQ content and structured data. Your answers are concise, accurate and directly useful, and your JSON-LD is valid.
CONTEXT
- Page URL and topic: [PAGE_URL_AND_TOPIC]
- Audience: [AUDIENCE]
- Facts, policies and details the answers must be based on (pricing, process, policies, specs): [SOURCE_FACTS]
- Questions I already know customers ask (optional): [KNOWN_QUESTIONS]
- Related searches or "People also ask" questions (optional): [PAA_QUESTIONS]
TASK
1. Select 6-10 questions that real searchers or customers would ask about this topic, prioritising [KNOWN_QUESTIONS] and [PAA_QUESTIONS]. Phrase them the way people type or say them.
2. Write an answer for each: the first sentence answers the question directly, then 1-3 sentences of useful detail. Keep each answer between 40 and 90 words.
3. Generate a FAQPage JSON-LD block containing every question and answer exactly as written.
4. Write a short HTML version of the FAQ section (using h3 for questions and p for answers) that matches the schema word for word.
RULES
- Answers must be based only on [SOURCE_FACTS]. If a question cannot be answered from them, write [NEEDS_ANSWER] instead of guessing, and keep it out of the schema until filled.
- No promotional fluff inside answers; no questions invented purely to sell ("Why is our product the best?").
- The visible FAQ text and the schema text must match. Do not include content in schema that is not visible on the page.
- Output valid JSON: use the "@context": "https://schema.org" and "@type": "FAQPage" structure with "mainEntity" as an array of "Question" items, each with an "acceptedAnswer" of type "Answer". Escape quotes correctly.
OUTPUT FORMAT
1. Question list with a one-line note on why each was chosen.
2. FAQ answers.
3. JSON-LD inside a script tag of type "application/ld+json", in a code block.
4. Matching HTML in a code block.
5. Any [NEEDS_ANSWER] items I must fill in.What to fill in
Everything in [GOLD CAPS] is yours to replace. Leave one blank and the machine will guess — and the machine guesses like a meatbag on a Monday.
[PAGE_URL_AND_TOPIC][AUDIENCE][SOURCE_FACTS][KNOWN_QUESTIONS][PAA_QUESTIONS][NEEDS_ANSWER]
How to use it
- Validate the JSON-LD with a structured data or rich results testing tool before shipping. Models leave the occasional trailing comma purely to keep you humble.
- Search engines decide when to show FAQ rich results, and eligibility rules change. Write FAQs because they help readers and AI answer engines; treat any rich snippet as a bonus.
- Mine your support inbox and sales calls for KNOWN_QUESTIONS. The questions customers actually ask outrank the ones you wish they asked.
Example
PAGE_URL_AND_TOPIC: /pricing for a dog grooming salon
SOURCE_FACTS: full groom from [PRICE] by size, nail trims walk-in, 24h cancellation policy, no sedation used, puppies from 12 weeks
KNOWN_QUESTIONS: how long does a groom take?Works with
Tested for ChatGPT, Claude, Gemini. Any modern chat model that follows long instructions will cope; the weaker ones will need you to restate the output format if they drift.
Weapons from the Arsenal that pair well with this prompt: