How to Get Your Site Cited by ChatGPT: 7 Proven Tactics

ChatGPT does not rank websites the way Google does. It synthesises answers from training data and live web retrieval, pulling in passages that are authoritative, clearly structured, and directly relevant to the query. If your site is not being cited, it is usually because one of a small number of signals is missing, not because your content is bad.

Here are the 7 tactics that consistently move the needle for ChatGPT citation, based on analysis of hundreds of sites that are and are not cited across different categories.

1. Allow GPTBot in your robots.txt

This is the most common and most fixable reason sites are not cited. OpenAI's crawler is called GPTBot. If your robots.txt blocks it (even accidentally, via a wildcard Disallow), ChatGPT cannot index your content for retrieval. Check your robots.txt right now.

# Allow ChatGPT to crawl your site
User-agent: GPTBot
Allow: /

# Also allow other AI crawlers
User-agent: ClaudeBot
Allow: /

User-agent: PerplexityBot
Allow: /

User-agent: Google-Extended
Allow: /

2. Add FAQPage schema to your key pages

FAQPage schema is the single highest-impact structured data type for AI citation. It tells AI systems: "here is a question, and here is the authoritative answer." ChatGPT's retrieval system is specifically optimised to surface pages with Q&A structure because they directly match the format of conversational queries.

The FAQ does not need to be visible on the page. It can be injected into the HTML head as JSON-LD and remain invisible to users. What matters is that the machine-readable markup exists.

3. Write definition-first paragraphs

AI retrieval systems extract passages, not full pages. A passage that opens with a clear, direct definition of the topic is far more likely to be cited than one that builds context for three sentences before getting to the point.

4. Add Organisation and Person schema

ChatGPT has a strong preference for citing sources it can verify as real organisations with real people. Organisation schema with a name, URL, and description (and Person schema for your authors) signals credibility at the machine level. Without it, your content looks anonymous.

5. Write in question-and-answer format

The queries ChatGPT receives are conversational. "What is the best project management tool for a 5-person startup?" is a typical query. Content that is structured as answers to specific questions (not just keyword-optimised paragraphs) aligns with how the retrieval system matches content to queries.

Use H2 and H3 headings that are phrased as questions. "What does X mean?", "How does Y work?", "When should I use Z?" These headings signal to retrieval systems that the content directly answers a conversational query.

6. Keep key passages under 200 words

AI retrieval systems extract passages, not pages. A 3,000-word article is only as good as its best 150-word passage. Structure your content so that each section contains a self-contained, citable answer, not a continuation of a long narrative that only makes sense in context.

7. Build topical authority through content clusters

ChatGPT is more likely to cite a site that has multiple pieces of content on a topic than a site with a single page. If you want to be cited for "project management software", you need content on project management methodologies, team productivity, software comparisons, and implementation guides, not just a homepage.

This is where AEO and SEO fully align: topical authority built for Google also builds citation credibility for AI engines.