AI share of voice is your brand's share of all brand mentions across a set of answers for a defined set of prompts. It is only comparable when the prompt set is fixed — change the prompts and the number moves without anything about your visibility changing.
The prompt set is the measurement
Share of voice has no meaning independent of the questions asked. A brand can hold 60% share on branded prompts and 0% on buying prompts. Any share-of-voice figure quoted without its prompt set attached is not a metric, it is a claim.
When we run this for a client the prompt set is harvested from real search demand and frozen, then re-run on the same set. Weighting matters too: our scorer weights prompts by buyer intent so a brand-name lookup cannot inflate the headline.
Open field versus beaten
Two very different situations produce a 0% share. In the first, competitors are named and you are not — you are losing a contest. In the second, the answer names no brands at all — the question is open field and nobody has claimed it.
Open field is the cheaper opportunity and it is common. In our own most recent audit of surfio.net, all 24 recorded citation gaps were open field: the engines answered the question fully and cited no vendor at all. Treating that as "we are losing to competitors" would be simply wrong, and it changes what you build next.
Frequently asked questions
- What is a good AI share of voice?
- There is no absolute threshold. Share of voice is only interpretable against the same prompt set over time, and against the specific competitors that appear in those answers. A number without its prompt set is not comparable to anything.
- Why is my share of voice zero when the answer mentions no brands at all?
- That is an open-field question, not a loss. Nobody has been cited, so the slot is unclaimed. It is usually easier to win than a question where competitors are already established.
- Should brand-name prompts count toward share of voice?
- Not in the headline. Almost every brand dominates answers to its own name, so including them makes the score look strong while telling you nothing about buying questions. Our scorer excludes raw brand-presence prompts from the headline figure.
Method last reviewed . Thresholds and engine list mirror the scorer used in SurfIO's paid audits.