Is your business visible to ChatGPT, Gemini and Perplexity? AI visibility checks
Do ChatGPT, Gemini and Perplexity know and recommend you? 12 checks · 10% of the overall score.
A growing share of “who should I hire” and “what is the best…” questions are now asked to AI assistants rather than typed into Google. Those assistants answer from what they can crawl and what they can parse. If your site blocks GPTBot, buries facts in images, or never states plainly what you do and where, you are invisible in that channel.
These checks cover the practical, first-principles parts of generative engine optimisation (GEO): can AI crawlers fetch the site, is there a clear machine-readable summary, does the content answer questions directly, and are there trust signals a model can cite.
The 12 checks
| Check | Why it matters |
|---|---|
| AI search crawlers allowed | robots.txt rules for GPTBot, ClaudeBot, PerplexityBot, Google-Extended and others decide whether you exist to AI search. |
| llms.txt present | An emerging convention: a plain-text summary of the site written for language models. |
| Clear opening summary | The first paragraph should say what the business does, for whom, and where — the sentence an AI will quote. |
| FAQ content | Question-and-answer content is the format assistants extract most reliably. |
| FAQ structured data | FAQPage schema makes the Q&A pairs explicit. |
| Question-style headings | Headings that match how people ask questions get matched to those questions. |
| Semantic HTML | main, article, nav and header tags help models separate content from chrome. |
| About page | An about page is where models learn who is behind the business. |
| Experience & trust signals (E-E-A-T) | Named people, credentials, years in business and reviews are what models weigh when recommending. |
| Outbound references | Citing sources and partners places the site in a web of trusted entities. |
| Page length for AI | Very thin pages give a model nothing to work with. |
| Wikipedia entity | Presence in Wikipedia/Wikidata is the strongest entity signal available. |
Most common fixes
- Allow GPTBot, ClaudeBot, PerplexityBot and Google-Extended in robots.txt unless you have a specific reason not to.
- Publish /llms.txt: two paragraphs on what you do, who for, where, and how to contact you.
- Add an FAQ section with real customer questions and mark it up with FAQPage schema.
- Open the homepage with one plain sentence: “[Business] is a [type] in [place] that [does what] for [whom].”
Questions
What is GEO (generative engine optimisation)?
GEO is the practice of making a business visible and citable inside AI-generated answers — ChatGPT, Gemini, Perplexity, Copilot and Google's AI Overviews. It overlaps with SEO but emphasises crawler access, plain factual statements, question-and-answer content and trust signals a model can quote.
Should I block AI crawlers?
Only if you are a publisher protecting content licensing. For a business that wants customers, blocking GPTBot or PerplexityBot means those assistants cannot recommend you.
What is llms.txt?
A proposed convention (like robots.txt) for a plain-text file at /llms.txt that summarises a site for language models. It is cheap to add and increasingly read.