Most businesses have never seen how an AI assistant describes them to a stranger. They have seen their website, their search rankings, and their own branded results — none of which tell you what happens when a buyer asks an assistant to explain the category, compare the options, and recommend a shortlist.

That gap matters because the answer arrives before the click. When an assistant summarizes a market, names three companies, and cites four sources, the buyer’s perception is largely set before any website is visited. A business can rank well and still be absent from that summary.

This audit closes the gap. It is deliberately manual: no tool currently reports AI visibility reliably across systems, and the manual pass produces something a dashboard cannot — the verbatim language the machine uses about your category, and the specific sources it treats as authoritative.

The procedure below produces a baseline you can defend and repeat. It will not tell you how to rank in an AI answer, because no one can promise that. It will tell you, precisely, where you currently stand and what is causing each gap.