How is AI share of voice calculated?
The simple version counts names. Peec AI documents it as "SoV = (Number of times your brand was mentioned) /(Total number of brand mentions) × 100" (Peec AI docs, read 1 November 2026). In Peec's example, a brand mentioned 4 times in 10 chats while a competitor is mentioned 12 times has a 25% share of voice. If each of those 4 mentions is in a different chat, Peec's visibility formula puts the same brand in 40% of answers (Peec AI). Share and presence are different numbers.
Other tools weight the count. Ahrefs Brand Radar weights each prompt by an "AI-adjusted volume" based on Google search volume and adds that to a brand's estimated impressions when an answer mentions it or cites its site. Ahrefs calls the result "a modeled estimate of exposure, not a count of people who saw an answer" (Ahrefs, 22 October 2026).
| Tool | What it counts | Weighting |
|---|---|---|
| Peec AI | Your brand mentions vs all brand mentions | Every mention counts the same |
| Ahrefs Brand Radar | Estimated impressions of answers that mention or cite you | Prompt search volume times a platform ratio |
| Semrush | Your mentions vs competitors in your market | Not stated on its metrics page |
Read on each vendor's site, 1 November 2026.
Why do tools give different numbers?
Because they track different prompts, engines and competitors, and weight them differently. Ahrefs says it plainly: "two tools can report different AI SOV for the same brand without either being wrong" (Ahrefs).
Answers also change from run to run. In Prefer's study of 960 answers (measured 13 September 2026), two runs of the same question overlapped by 37% in the sites they cited on ChatGPT and 35% on Gemini, against 96% on Perplexity. So run each prompt more than once before you trust a share.
Why does AI share of voice matter for B2B SaaS?
An AI answer to a buying question usually names a handful of tools. AI share of voice tells you how often you are one of them compared with your rivals, across the questions your buyers ask. Tracked monthly with the same prompts, it shows whether that is moving.
What does AI share of voice look like? (hypothetical example)
This example is hypothetical. You track 50 prompts on two engines, three runs each: 300 answers. Your brand is named 60 times; three competitors 120, 70 and 50 times. Total: 300 mentions, so your AI share of voice is 20%. Competitor A has 40%. The next step is to read the sources behind A's answers and see which lists name A and not you.
AI share of voice vs related metrics
- Share of voice is the original marketing measure: your part of all attention in a market.
- Citation share counts links to your pages, not your name in the text.
- An AI visibility score is a single number a tool reports, often built from mentions.
- Prompt tracking is the method that collects the answers all of these are counted from.
Many of the mentions behind your share come from pages you do not own. When Prefer checked 188 answers from ChatGPT, Perplexity, Gemini and Claude in July 2026, 40% of the sources cited for buyer questions were independent lists. Vendors' own sites got 34%, review sites 3%. The LLM SEO guide explains how brands get into those lists.