> The full list of bisica's docs pages is at https://docs.bisica.com/llms.txt

# Your report

What every number in your AI visibility snapshot means and how we got it.

- [The four numbers in your report: mentioned, recommended, cited and wrong](https://docs.bisica.com/your-report/the-four-numbers/index.md): Your report opens with four counts of answers. They show how many mention you, how many call you the pick, how many list your site as a source and how many say something false about you.
- [Read the leak count](https://docs.bisica.com/your-report/the-leak-count/index.md): The first page of your report is the leak count. It sorts the answers where AI is costing you into four lines (unnamed source, rival named, missing facts and wrong statements) that always add up to the total.
- [Which questions and AI tools a snapshot uses](https://docs.bisica.com/your-report/how-a-snapshot-asks/index.md): A snapshot puts 50 of your customers’ questions to ChatGPT, Claude, Perplexity and Google AI Overviews. Each question goes to each tool three times on its own with web search on, for 600 asks in all.
- [Why your own ChatGPT answer can differ from the report](https://docs.bisica.com/your-report/why-answers-differ/index.md): AI answers change each time you ask, your app remembers your earlier chats, answers vary by place and the models get updated. A snapshot counts 600 answers instead of trusting one.

---

Source: https://docs.bisica.com/your-report/ · Updated November 2, 2026