Share of Voice

Share of voice began as an advertising metric: the proportion of total paid media in a market that belongs to one brand, set against its share of market. The SEO version borrows the logic and applies it to organic search, and the two are often confused, so it is worth stating the distinction plainly before going further.

SEO share of voice (SOV) is the proportion of total available organic visibility, across a chosen set of keywords, that your site captures compared with every other site ranking for those keywords. Where average position tells you how you rank and traffic tells you how many people arrived, share of voice tells you how much of an entire market you own. It is a competitive metric, not an absolute one.

The appeal is that it compresses a whole topic into a single comparable number. A board or client does not want a spreadsheet of 200 keyword positions; they want to know whether the brand’s grip on its market is growing or shrinking. Share of voice answers that.

How do you calculate share of voice?

The calculation turns rankings across a keyword set into an estimate of captured clicks, then expresses yours as a percentage of the total available.

Define the keyword set. Choose the keywords that represent the market you care about. This is the most consequential decision: the number is only meaningful relative to the set it is measured against, and the set must stay stable to be comparable over time.

Apply estimated click-through rates to positions. For each keyword, your ranking position maps to an estimated click-through rate (position 1 earns far more clicks than position 8). Multiply each keyword’s search volume by the CTR for your position to estimate the clicks you capture from it.

Sum and divide. Add up your estimated clicks across the whole keyword set, then divide by the total estimated clicks available across all sites ranking for those keywords, and multiply by 100.

The formula in plain terms: (your estimated clicks across the keyword set ÷ total estimated clicks available across the set) × 100. If 100 target keywords generate 50,000 estimated monthly visits across every site that ranks for them, and your positions capture 15,000 of those, your share of voice is 30%.

In practice, rank tracking tools (Ahrefs, Semrush, and dedicated rank trackers) compute this automatically once you supply the keyword set, applying their own CTR curves. The arithmetic is less important than understanding what the number represents and how the inputs shape it.

Why use share of voice as a KPI?

Share of voice captures something individual metrics miss.

It is competitive, not absolute. Traffic can rise while you lose ground, if the market is growing and competitors are growing faster. Share of voice exposes that: a falling percentage during rising traffic is an early warning that absolute numbers hide.

It summarises a whole market. One number stands in for the combined position of dozens or hundreds of keywords, which makes it suited to executive reporting and to tracking strategic progress against a market rather than tactical movement on single terms.

It frames the goal as market share. Positioning SEO as capturing share of a defined market aligns it with how businesses think about competition generally, which helps when communicating SEO value to stakeholders who do not think in rankings.

What is AI share of voice?

Since 2025 “share of voice” has meant two different things, and which one someone means is usually left unstated. The version above measures your share of estimated organic clicks across a keyword set. AI share of voice measures your share of brand mentions across AI-generated answers. Semrush states the simple form as “(your AI mentions ÷ total AI mentions across all brands in your category) × 100”.1

The two are worth holding apart because almost nothing transfers between them:

SEO share of voiceAI share of voice
Unit countedEstimated clicksBrand mentions and citations
InputRanking positions across a keyword setAnswers across a prompt set
DenominatorEvery site ranking on those SERPsEvery brand mentioned in your category
WeightingModelled CTR by positionVaries by vendor; often mention count and position

A brand can hold a strong SEO share of voice and a weak AI one, or the reverse, because the retrieval that produces an AI answer is not the ranking that produces a SERP. Treat them as two KPIs, not one metric measured two ways. The prompt-set mechanics, tooling and the other AI visibility metrics that sit alongside it (citation rate, prompt coverage, sentiment) are covered in AI visibility measurement; this page owns the concept and the comparison.

What are the limits of share of voice?

Share of voice is an estimate built on estimates, and treating it as precise is a mistake.

It rests on modelled CTRs, and this is the load-bearing weakness. The click-through rates applied to positions are industry averages, not your actual data. Zero-click results, AI Overviews, and SERP features distort real CTR away from the model. See zero-click measurement for why position no longer maps cleanly to clicks. There is a live argument that this now invalidates the metric outright rather than merely softening it: writing in Search Engine Land in June 2026, Dan Taylor of SALT.agency argued that “traditional share of voice (SOV) is effectively obsolete”, on the grounds that “no two users will encounter the same interface, even when entering the exact same query at the exact same moment”, and that AI share of voice inherits the same flaw.2 That is a signed opinion column rather than a study, but it is the strongest case against the metric and worth reading before you build a board report on it.

It depends entirely on the keyword set. A flattering set produces a flattering number. The set must be chosen honestly and held constant; changing it mid-stream breaks comparability and can manufacture an improvement that is not real. Watch branded terms in particular: leaving them in is the easiest way to manufacture a healthy-looking number, since you will nearly always dominate your own brand SERPs. Split them out the way you would anywhere else (branded vs non-branded).

A stable keyword set does not guarantee a stable series. Google’s removal of the num=100 parameter in September 2025 disrupted how rank trackers collect positions, and vendors restored deep collection on their own timelines. Any share of voice series spanning autumn 2025 carries a tool-side discontinuity that has nothing to do with your performance. See rank tracking.

No two vendors calculate it the same way. Ahrefs computes it in Rank Tracker as “Total clicks your target receives from all Tracked keywords / Total clicks going to all result in the SERP receives for all Tracked keywords”.3 Semrush’s model also takes SERP features into account, and its AI share of voice uses mention count and brand position within the answer, factoring topic search volume for ChatGPT.1 Two tools will hand you two different numbers for the same site on the same day. Pick one, and never compare a figure from one tool against a figure from another.

It is a direction, not a precise figure. The trend in share of voice over time is informative; the exact percentage in any single month is soft. Report it as a movement, not a measurement to two decimal places.

The denominator is available clicks, not actual ones. With current zero-click rates, much of the volume in the denominator produces no click for anyone. Share of voice can rise while every site in the set, including yours, loses absolute clicks.

Fix the keyword set before you report the number

The integrity of share of voice lives entirely in the keyword set. Define it once, deliberately, to represent the market you actually compete in, then lock it. The temptation to add keywords you have started ranking well for, or to drop ones where you slipped, will inflate the number and destroy its value as a trend. If the market genuinely shifts and the set must change, treat it as a baseline reset and note the discontinuity, the same way you would when changing any measurement methodology.

How does share of voice fit with other metrics?

Share of voice is a strategic, market-level KPI that complements rather than replaces the operational metrics in how to measure SEO performance. Use traffic and conversions to measure what SEO delivers to the business, rank tracking to monitor specific keyword movements, and share of voice to answer the higher-level question of whether your competitive position in a market is strengthening or eroding. Each answers a different question, and share of voice is the one built for the boardroom rather than the daily dashboard.

Footnotes

  1. How to measure AI share of voice — Semrush 2

  2. The problem with AI share of voice and 3 metrics that matter more — Search Engine Land

  3. What is Share of Voice (SoV) and how is it calculated? — Ahrefs Help Center