AI Share of Voice: The Metric That Matters Now
AI share of voice explained: how it's calculated, why it beats a single visibility score, what moves it, and how often to check it. With a worked example.
July 3, 2026 · 8 min read
Photo by Thắng-Nhật Trần on Pexels
A visibility score answers 'am I doing okay?' Share of voice answers the more useful question: 'okay compared to whom?' In a world where an AI answer typically names just two or three brands for any given question, your share of voice — how often you're the one named, relative to your real competitors — is the number that maps most directly to whether you're winning or losing deals you never even hear about.
Key takeaways
- Share of voice is the proportion of AI answers naming you versus each tracked competitor, across engines and over time
- It beats a single visibility score because a rising score can still mean you're losing ground if rivals rise faster
- It surfaces exactly which competitor is winning which question, turning a vague problem into a specific one
- Crawl access, structured data and third-party mentions are the biggest levers that move it
- Weekly checks are a reasonable minimum; daily once you're actively making changes
How is AI share of voice actually calculated?
For a set of tracked questions, share of voice is the proportion of AI answers that name you versus each tracked competitor, aggregated across engines and over time. If you're named in 4 of 10 checks and a rival is named in 6, your share of voice is roughly 40% against their 60% for that basket of questions. Run across five engines and a dozen real buyer questions weekly, this produces a genuinely useful trend line rather than a single noisy data point.
| Brand | Times named | Share of voice |
|---|---|---|
| Your brand | 4 of 10 | 40% |
| Competitor A | 6 of 10 | 60% |
| Competitor B | 2 of 10 | 20% |
Note the percentages don't need to sum to 100 — a single AI answer can name more than one brand, or none at all, so share of voice measures presence against your own tracked basket of questions, not a fixed pool that gets divided up.
Why does share of voice beat a single visibility score?
- A rising score in isolation can still mean you're losing ground if competitors are rising faster over the same period
- It surfaces exactly which competitor is winning which specific question, which turns a vague problem into a specific, assignable one
- It's the number that maps most directly to real buyer decisions, since most AI answers are explicitly comparative by nature
What actually moves it?
- Crawl access and structured data — fixing a blocked crawler can shift share of voice within a single re-check cycle
- Fresh, specific content on the exact questions being tracked, rather than general brand content that doesn't match buyer phrasing
- Independent third-party mentions, which many engines weigh alongside your own site content
- A competitor's own activity — their new content, price changes or product launches can move your share of voice even if you've changed nothing yourself
Share of voice turns 'we should probably work on AI visibility' into 'we're losing the pricing question to Competitor B, specifically on Perplexity.'
A composite example
A B2B software company we've seen was tracking a flat 35% overall AI visibility score for two months and assumed things were stable. Breaking it down by competitor showed the real picture: their share of voice against one specific rival had dropped from 50% to 20% over that same window, offset by gains against two smaller, less relevant competitors. The overall score hid a genuinely serious problem with their most important rival. Once they saw the breakdown, they could trace it to that rival publishing a strong new comparison page — and respond with one of their own.
How often should you check it?
Weekly is a reasonable minimum for most brands; daily makes sense once you're actively making changes and want to see whether a specific fix moved the number before the next planning cycle. Because AI answers can shift quickly — sometimes within days of a competitor publishing new content — a stale share-of-voice number is close to useless for actual decision-making.
Is share of voice the same across every AI engine?
No — and that's exactly why it needs to be tracked per engine, not just as one blended number. A brand might hold 60% share of voice on Google AI Overviews (where organic SEO strength carries over) while sitting at 15% on Perplexity (where live retrieval and crawl access dominate). Blending those into a single figure hides which engine actually needs attention.
See your share of voice against up to five competitors, tracked across every major AI engine.
See how competitor tracking worksOnce you're watching share of voice regularly, the next question is usually why a specific competitor keeps winning a specific question — which almost always traces back to one of a handful of fixable causes.
Read why AI recommends a competitor instead of you, and how to fix it.
Diagnose your visibility gapFrequently asked questions
It's the proportion of AI-generated answers that name your brand versus each tracked competitor, aggregated across engines and questions over time — the comparative version of a raw AI visibility score.