Share of AI voice — The proportion of a fixed set of buying questions where a business is named in the AI answer.
Why a single check proves nothing
AI answers are non-deterministic. The same prompt returns different text at different times and to different people. One run is an anecdote. A number taken from one run and put in a report is, at best, decoration.
How to measure it honestly
- Fix a set of 10 to 20 prompts that mirror real buying questions for your category and city. Do not change them.
- Run them on a schedule, in clean sessions, across every surface you care about.
- Log three things per run: whether you were named, who else was, and which sources were cited.
- Report the trend, not the snapshot.
The citation log is the most actionable artefact in the whole exercise. It is a ranked list of the sources you need to appear on, generated by the systems you are trying to influence.
The number is a floor, not a score
Share of AI voice tells you how often you were in the room. It does not tell you whether the answer described you well, or whether the person acted on it. Anyone reporting it as a single score without the transcripts behind it is asking you to trust arithmetic you cannot check.
See it applied in the free AI Visibility Snapshot. Related: zero-click search.