Share of voice as a KPI: what it does and doesn't tell you

A marketing manager at a SaaS company showed us a dashboard last year. Share of voice sat at the top: 18 percent. Below it, an upward arrow. When we asked what that 18 percent meant, the room went quiet. The number came from a tool, the tool compared five competitors, and nobody knew which five or why those.

That is the problem with share of voice as a KPI. It is a valuable signal and an easily abused figure. The difference lies in how you define it and which decision you attach to it.

Where share of voice comes from

The idea is old and simple: if your share of total marketing communication in a category exceeds your market share, you grow over time. If it is smaller, you shrink. That relationship shows up in various media studies, particularly in FMCG and other categories with frequent repeat purchases.

That origin matters. The original measurements covered advertising spend in mass media, within sharply defined categories, against a measured market share. If you work in B2B with a subscription model, long buying cycles and no reliable market share figure, you are copying a model into a context where the assumptions do not fully hold. That is allowed, but you need to know it.

Three kinds of share of voice that get mixed up

In conversations we notice people using the same word for completely different measurements. That produces discussions without conclusions.

• Share of spend. Your advertising budget relative to the category total. Reasonably estimable for large players with public figures, close to impossible in small B2B markets.
• Share of search. Your portion of brand-name search volume within the category. Well measurable with Google data, and a usable proxy for demand. Only works if people actually search for brand names.
• Share of conversation. Your portion of mentions on social, in media or on review platforms. Interesting as a qualitative signal, sensitive to noise and to one viral post.
• Share of voice within a channel. For example the percentage of impressions you capture on a set of keywords. This is a channel metric, not a brand metric, however often it is presented as one.

Four measurements, one name. If you do not make this explicit in your reporting, you end up debating a number with your board that nobody reads the same way.

What share of voice does tell you

As a directional signal over the medium term it is useful. It tells you something about your position in the market's attention, not about your position in next quarter's pipeline.

In practice we use it for three questions. Am I more or less visible than six months ago? Does my visibility move with the category or diverge from it? And if a competitor scales up, do we see it here before we see it in our own revenue? The answer to that last question is often yes, and that is precisely the value.

Share of search has a practical advantage: you do not need expensive research. You need brand names, a search volume source and consistency in your method. For a small team that is a day to set up and half an hour a month to maintain.

—Share of voice is a thermometer, not a steering wheel. It tells you something is going on, not what to do about it.

What it does not tell you

This is usually where it goes wrong. Four things share of voice does not measure:

• Quality. Being mentioned ten times for the wrong reason counts the same in most tools as ten mentions for the right reason.
• Audience relevance. Visibility among students and visibility among buyers with budget look identical in the data.
• Cause and effect. Higher share of voice can be the result of growth rather than its cause. Companies that sell well get mentioned more.
• Direct revenue. In subscription B2B there are often six to eighteen months between visibility and signature. You cannot line this figure up against monthly results.

The hardest issue is category definition. Who are you competing with? A contract management tool competes with other contract tools, with broader legal software, and with Excel. Choose a narrow category and your share of voice looks great. Choose a broad one and the number is small with barely visible movement. Both are defensible. Neither is objective.

How to set it up with a small team

1. 1Pick one definition and write it down. In B2B we usually start with share of search, because it is reproducible and does not depend on budget estimates.
2. 2Define the competitive set together with sales. Ask which names come up in deals. Five to eight brands is workable; thirty makes the figure stable but meaningless.
3. 3Measure monthly, judge quarterly. Monthly noise leads to panic decisions. Look at the line across four to six quarters.
4. 4Put it next to two other numbers: direct and branded demand, and the share of deals that name you first. Together those three tell a story one figure cannot.
5. 5Map campaign periods onto the timeline. Without context you cannot tell whether a rise was your work or a competitor pulling the media plug.
6. 6Agree in advance which movement triggers action. Otherwise it becomes a decorative number on a dashboard.

Teams underestimate point six most often. A KPI without an agreed threshold is a reporting line, not a steering instrument. Agree on something like: after three quarters of declining share of search with stable or rising category demand, we revisit brand investment. That is a usable agreement.

The ratio is the real question

Share of voice is popular because it makes brand investment defensible. That is legitimate. But the figure alone says nothing about the right split between building brand and harvesting existing demand. A company with a short runway and a proven channel should split differently than a company with a full bank account and an unknown name.

What we do in practice: get the foundation right first, meaning definitions, method and competitive set on paper. Then set direction, including the ratio between brand and pipeline. Then execute. And recalibrate every quarter based on what the numbers actually show, not on what makes the dashboard look good. At REX.digital share of voice sits in the reporting, but never as the only brand indicator.

Start small. Pick your definition, measure consistently for six months, then check whether the number produced a decision you would not have made without it. If not, you are measuring the wrong thing or the wrong category. That is a perfectly fine outcome, as long as you act on it.