Share of voice: one phrase, three different sums
Two tools can show the same brands with opposite winners, and both be right. A four-prompt worked example shows how coverage, mention share and search-weighted share of voice can rank rivals in reverse.
“Share of voice” is the number clients and bosses love most, because it fits on a slide: you versus them, as a percentage. In AI visibility it can be calculated at least three different ways, and the method can change who “wins”.
The worked example
Made-up, deliberately small. Two rival brands, North and South, tracked on four prompts. Each prompt has a monthly search volume. Most answers mention one brand; prompt B mentions neither.
| Prompt | Search volume | North mentioned? | South mentioned? |
|---|---|---|---|
| A: “best budget option” | 5,000 | No | Yes |
| B: “best for teams” | 3,000 | No | No |
| C: “easiest to set up” | 1,000 | Yes | No |
| D: “best integrations” | 1,000 | Yes | No |
Three sums, three answers
1. Coverage: share of prompts that mention the brand
- 50%
- 25%
2. Mention share: each brand’s share of all brand mentions
- 67%
- 33%
3. Search-weighted share: mentions weighted by each prompt’s search volume
- 29%
- 71%
On the first two measures North leads comfortably. On the third, South leads by more than two to one, because its single mention is on the question people ask most. None of these is wrong. They answer different questions:
The rival list changes everything
Mention share and weighted share only count the brands you tell the tool to track. Add a third rival that appears in the saved answers and every share shrinks, even though nothing about your own visibility changed. Remove one and yours grows. Ahrefs warns that share of voice “can change a lot depending on how you’ve set up the entities”. Before comparing share of voice between tools, or between months, check that the rival list is the same.
Which one to report
Our piece on what AI visibility tools measure shows how each of these is built from the same saved answers, and trend or wobble explains why small samples like this example swing so much.
Sources
- OtterlyAI: Brand report KPI definitions viewed 11 Nov 2026
- Semrush: AI SEO metrics viewed 11 Nov 2026
- Ahrefs: AI visibility metrics viewed 11 Nov 2026
- Scrunch: Metrics guide viewed 11 Nov 2026
Directory figures come from Hyperfind’s profiles of 34 tools, not the whole market. Prices follow each provider’s published offer on the date shown in its profile.
