Share of voice: the portfolio number your rank tracker doesn't show you

What share of voice actually measures

Ahrefs defines share of voice as brand visibility value divided by total market visibility value, then lets the "visibility value" change depending on the channel. For organic search, that value is usually built from estimated traffic share across a tracked set of keywords, which means it depends on two things a plain rank report does not need: the search volume behind every keyword in the set, and the click-through rate assumed for whichever position you hold on each one.

Semrush's own guide to the metric makes the portfolio logic concrete with a real worked example: a domain carried about 8% share of voice in South Lake Tahoe and roughly 4 percentage points less across Los Angeles, San Francisco and Sacramento. A second illustration in the same guide names a client sitting at 5% share in its main city and 2% in surrounding areas, against a competitor at 7% and a stated target of 10%. None of those numbers come from one keyword. They come from however many keywords sat inside the tracked set and however much volume each one carried, which is precisely the information a single average position throws away.

That volume weighting is also where the number gets fragile. Advanced Web Ranking's Dan Popa reported on 23 June 2026 that organic click-through across the first five desktop positions rose 10.54 percentage points in Q1 2026 versus Q4 2025, while mobile's top position fell 2.20 points over the same quarter, and that when an AI Overview renders on a query, the top organic result's click-through is typically cut by roughly half. Any share of voice figure built on a fixed click-through curve is already using a stale weighting scheme the moment that curve moves, and the curve just moved that much in a single quarter.

Why one keyword's rank can look steady while the real picture moves

This is the same blind spot our guide to whether rank tracking still matters describes from a different angle: a blended average position can hide which placement is actually sending traffic. Share of voice attacks the same problem from the portfolio side instead of the per-query side. Semrush's guide puts the asymmetry plainly: a hundred new top 10 rankings on keywords with little search volume can move almost nothing, while a single top position on one or two of a niche's highest volume terms can move the number a lot. A keyword-by-keyword rank report treats both outcomes as a pile of wins. Share of voice, weighted by volume, treats them as what they actually are.

It is still built on the same underlying rank data, though, so the problems that data already has do not disappear. Our guide to why your rank tracker doesn't match what you see covers how a single manual check can diverge from what a tracker reports because of location, login state and personalization. A share of voice score aggregates many of those same individually noisy checks, which smooths some of the noise out, but it does not fix the underlying question of whether any one of those checks reflects a stable, comparable position in the first place.

The real limitations nobody puts in the pitch deck

Ahrefs' own documentation is candid about where organic share of voice runs thin. Rank tracker based visibility, in its words, covers one channel only, and its proportions "don't correspond to overall SOV or market shares." Sites with broad audiences can inflate their apparent share, and a keyword set stuffed with high volume, low value terms can skew the result in either direction, which is why the glossary itself tells readers to take the organic number "with a grain of salt."

Quattr's James Gibbons lays out a longer list of mechanical problems with calculating share of voice for SEO. The core one: share of voice tools "are not connected to brand's specific traffic metrics such as clicks and impressions," so the score is hard to tie back to what actually happened on your site. Combining tools to patch that gap introduces its own mess, since Semrush, Ahrefs and Google Search Console "prioritise different data set[s] such as search volume or search visibility," producing what Gibbons calls inconsistent insight once their numbers get reconciled. Refresh speed is a separate problem: most share of voice tools update weekly or monthly, while Google itself reported roughly 4,725 ranking changes across 2022, close to 13 a day, a pace that makes a monthly snapshot stale before anyone finishes reading it. Local, mobile and seasonal variation tend to get flattened by that same slow refresh cycle, which Gibbons notes most tools do not break out at all.

Worth saying plainly: Gibbons' piece is published by a company selling a competing visibility metric, so its specific pitch for an alternative score deserves the same skepticism as any vendor comparison. The underlying list of gaps, no link to real traffic, inconsistent output across tools, and a refresh cycle slower than the thing it is measuring, holds up on its own regardless of who is pointing it out, and it matches Ahrefs' separate, independently stated caveat about one-channel coverage.

What we could not verify

We could not find an independent, non-vendor study that directly measures the correlation between a change in share of voice and a resulting change in actual revenue or traffic, which is the exact claim a pitch deck built on the metric usually implies without stating. The "roughly half" click-through reduction Advanced Web Ranking reports for AI Overview SERPs is described qualitatively, "typically on the order of half," rather than as one precise percentage, so we have not repeated it as an exact figure. Semrush's 8%-to-4-point worked example dates to 2021 and is presented there as an illustration of the concept rather than a current benchmark; the arithmetic logic holds, but those specific percentages are not current data for any real market. And the Google figure of roughly 4,725 algorithm changes in 2022 is itself several years old, with no equivalent published count for 2026, so the argument that monthly checks lag daily change rests on a dated baseline, even though nothing suggests Google's pace of change has slowed since.

A check worth running this week

1. Split your tracked keyword list by estimated monthly search volume. If your reported ranking gains sit mostly on the bottom half of that list, your share of voice likely moved less than a keyword-count summary implies, the exact pattern Semrush's own low-volume example describes.
2. Ask whichever tool reports your share of voice number what click-through curve it assumes, and how recently that curve was updated. If the answer is that it is not disclosed, the number is less auditable than a plain average rank, not more.
3. Before quoting a share of voice trend to a client or your own leadership, rerun the same keyword set signed out and note whether it holds. The same discipline our rank tracker personalization guide recommends for an ordinary position check applies just as much to the aggregate number built on top of it.

Sources

The share of voice definition and formula, and the one-channel caveat, are from Joshua Hardwick, Share of Voice (SOV), Ahrefs, published 13 August 2020, updated 4 October 2023. The worked city-level examples are from How to Measure Your SEO Share of Voice, Semrush, 16 March 2021. The tool and refresh-cycle limitations are from James Gibbons, Challenges of Calculating Share of Voice for SEO, Quattr. The click-through rate figures are from Dan Popa, Google CTR Stats: Changes Report for Q1 2026, Advanced Web Ranking, 23 June 2026.

Common questions