The AI Visibility Index

We measure who AI assistants actually name

When a buyer asks an AI assistant for a supplier, a handful of companies get named and the rest do not. We ask the same buying questions a buyer would type, in one category and one state at a time, and record every company the answers name.

88%

of the suppliers we track were never named when buyers asked AI for a recommendation.

Measured across 2,497 companies in 33 published cells of the Index.

Open a cell

Choose a category and a state

One category crossed with one state is one cell of the Index. Choosing both opens that cell's public report.

Hydraulic & Tube Expanders

Loading states…

Choose a category and a state to continue.

Method

How a cell is measured

Three steps, run the same way for every category and every state.

  1. Step 1 of 3

    Five buying questions

    Each category is written five questions a buyer would actually type, with the state filled in. Every company in the cell faces the same five, which is what makes the result a ranking rather than an anecdote.

  2. Step 2 of 3

    The answers are read

    Every company named in every answer is pulled out and matched against the Blue Book record. A company named under three spellings is one company, counted once.

  3. Step 3 of 3

    The cell is scored

    A company's score is how many of the measured answers named it. The cell's score is how many of the suppliers it lists were named at all.

Reading a cell report

Cell
One category crossed with one state is one cell of the Index — for example, Metal Stamping in Ohio.
Buying questions
The questions a buyer would type, with the state filled in. Every company in a cell is measured on the same ones.
Score
How many of the measured answers named a company — shown as, for example, “8 of 15 answers”.
Never named
Tracked in the cell and asked the same buying questions as everyone else, but not named in a single answer.