Guide
How to get cited by ChatGPT: 7 signals that actually matter
Getting cited by ChatGPT comes down to making your business easy to identify, easy to verify, and easy to quote. The 7 signals that matter most are: AI crawler access, entity clarity, schema markup, answer-ready content, third-party corroboration, llms.txt, and review presence. The free Visus audit checks all 7 and tells you which gaps apply to your site.
ChatGPT doesn't recommend businesses at random. It recommends businesses that its training data — and, when browsing is enabled, its live web index — can confidently identify and verify. That confidence comes from signals. The more signals you send, the clearer and more consistent they are, the higher the likelihood of a citation.
Here are the 7 signals that matter, with specific fixes for each.
AI crawler access
If GPTBot — OpenAI's web crawler — is blocked in your robots.txt, ChatGPT's training data may not include your website at all. This is the most overlooked barrier to AI citation, and the easiest to fix.
Check your robots.txt file (visit yourdomain.com/robots.txt). Look for any Disallow rules that include GPTBot, or a catch-all Disallow: / that would block all crawlers.
Entity clarity
ChatGPT recommends businesses it can identify without ambiguity. If your website doesn't clearly state your business name, what you do, where you are, and who you serve — in plain prose that a human could read and understand immediately — AI systems struggle to build a confident picture of who you are.
Many businesses have a homepage that leads with a tagline ("Excellence in every project") but buries the actual facts. AI can't work with taglines.
Schema markup
Schema markup (JSON-LD structured data) removes ambiguity by explicitly declaring what type of entity your page describes. Rather than ChatGPT inferring you're a local business from your prose, your schema states it directly.
The schema types that matter most for ChatGPT citation: LocalBusiness or Organization (with name, address, phone, URL), FAQPage (question-answer pairs AI can quote directly), and Person (for founder/expert credibility).
Answer-ready content
ChatGPT constructs answers from passages it can quote or paraphrase. Content written in vague marketing language — "we deliver exceptional results for our clients" — gives AI nothing usable. Content written as direct answers to buyer questions gives AI exactly what it needs.
Think about the questions your customers ask you before buying. Add a FAQ section that answers each one clearly, in 2–4 sentences. "How much does [your service] cost?", "How long does it take?", "What area do you cover?", "Do you offer [specific thing]?"
Third-party corroboration
AI is more confident recommending businesses it can verify from multiple sources. If your business exists only on your own website, AI has to rely entirely on what you say about yourself. Third-party mentions — reviews, press, directories, trade associations — provide independent verification.
The corroboration sources AI systems trust most: Google Business Profile, Trustpilot, Yelp, industry-specific review platforms, local press, trade body directories, and professional association websites.
llms.txt file
llms.txt is a plain-text file at the root of your website that tells AI crawlers directly about your business. It's the AI equivalent of a robots.txt — except instead of controlling access, it provides structured, machine-readable context about who you are, what you offer, and where to find accurate information.
Most businesses don't have one. Adding a well-structured llms.txt is a fast way to ensure AI systems have a direct, authoritative source of information about your business.
Review volume and recency
Reviews are a form of third-party corroboration (signal 5), but they deserve separate attention because of their weight. AI systems — particularly Perplexity, which indexes the live web — give significant weight to review presence when deciding which local businesses to recommend.
A business with 50+ Google reviews and a 4.5+ rating is far more likely to appear in AI answers for "best [service] in [city]" queries than a business with 3 reviews or no reviews. Review recency also matters — reviews from the last 6 months signal that the business is active.
Which of these 7 signals are you missing?
The free Visus audit checks all 7 and returns a score out of 100 with your specific gaps. Under 60 seconds, no credit card required.
Run free auditHow long until ChatGPT cites you?
Fixing signals 1–4 (on-page changes) typically shows results within 2–4 weeks as AI crawlers re-index your pages. Signals 5–7 (third-party) take longer: 4–12 weeks as reviews accumulate and directory listings propagate.
Perplexity usually responds faster than ChatGPT because it indexes the live web. Most businesses applying all 7 fixes see Perplexity movement first, then Google AI Overviews, then ChatGPT.
The Fozias restaurant in Liverpool saw their AI visibility score move from 23 to 80 in 90 days. Read the full case study.