SEO Share of Voice

There is no standard formula, and two vendors show it. Ahrefs says its share of voice in Rank Tracker is the total clicks your target receives from all tracked keywords divided by the total clicks going to all results on the results pages of those keywords, and that it is equivalent to the “visibility score” other platforms use^[1].

Semrush treats the two as different. In Position Tracking, its Visibility % is a metric based on click-through rate that shows a website’s progress in Google’s top 100 for the keywords of the campaign. Share of voice is a separate metric that goes one step further by taking into account the total combined volume of all the keywords, so it is weighted more by the actual number of searches and estimated traffic than by the number of keywords^[2]. In Semrush’s own example, a site in the top position for ten small keywords and absent for one keyword with 30,000 monthly searches would have a high visibility % and a low share of voice.

So the same words name different numbers. A share of voice from one tool cannot be set against one from another, and it cannot be set against a change of tool. Both rest on estimates of clicks and searches, and Google says third-party tools don’t have access to its internal ranking data, can’t guarantee performance, and that any predictions are their own^[3].

The first-party counterpart is narrower. Search Console reports the impressions, clicks and average position of your own site, and counts position as the average over impressions, so a rank tracker’s position taken on its own schedule and location need not equal it^[4]. It cannot tell you what competitors received, which is why a share of voice needs a tool’s estimate.

A share is also only as meaningful as the set it is taken over. In QueryRev’s study, the share of sites that block an AI crawler by name runs from 44.6% of the sites read on .it to 3.7% on .cn^[5]: the same measure over a different set gives a different share. A share of voice across sixty keywords and across six hundred differ in the same way, which is why the set is part of the number.

Fix the keyword set, the competitors, the countries and the devices, and write them down. A share is a ratio between a numerator and a set, and moving the set moves the ratio without anything changing in search.

Record the tool’s definition with the figure: whether it weights by estimated clicks, by volume or by a click-through curve. Semrush adds automatic notes to its trend graph when keywords are added or deleted from a campaign^[2]; keep a note of your own for every change to the set.

Write the figure as an estimate with its set size and date, and as a range where the inputs are estimated. Then read it beside Search Console clicks and impressions for non-branded queries and beside conversions, which show what the site actually received.

Illustrative: an invented site, to show the method.

An outdoor-gear shop tracks 60 keywords against two competitors in one rank tracker. The shop is first on 24 of the 60, more than either competitor. Competitor A holds first place on the five keywords with the most searches, and the tool weights by volume and estimated clicks.

The tool gives competitor A a share of voice of 31%, the shop 24% and competitor B 9%. The shop is first by keyword count and second by share of voice. Both statements are true. Because the volumes and clicks behind each figure are estimates, the shop reads them as ranges, 28% to 34%, 21% to 27% and 7% to 11%, and plans for the five big keywords.