What an AI visibility audit should include
An AI visibility audit should measure how often ChatGPT and Google AI recommend you on a fixed set of buyer questions, then find which of five gaps is causing the result.
The measurement
10 to 20 questions your buyers ask, each asked at least three times in each engine, logged out, from your market. One screenshot of one answer is not a measurement: answers vary from run to run.
The five gaps, in order
- Access: can Google, Bing and OpenAI's crawler read and index your site?
- Identity: do AI and the web agree on who you are, what you do and where?
- Sources: are you on the pages AI cites for your category?
- Content: does your site answer each buyer question plainly?
- Reputation: do your reviews and mentions hold up next to the competitors AI names?
What you should get
A written fix plan: each action with the reason from your own answers, its priority, who does it and how you'll see it's done. Then the same questions measured again, the same way.
Red flags
- A promise of guaranteed placement in ChatGPT or Google.
- Results shown from a single run of each question.
- llms.txt files or special AI markup sold as the fix. Google says neither is needed to appear in its AI features.
- Keyword stuffing. Research on generative engines found it doesn't help.