Who shows up when you do not

• You know you barely show up, and you do not know who takes the place you do not.
• The models recommend a company nobody on your team had heard of.
• Your share of voice is sixty per cent and it turns out to be three mentions.

A share of voice with no denominator is a decorative figure.

1. 1

   Everyone in the same answer

   There is no separate sweep per brand: who appears in each answer of the same sweep is counted. It is the only way the comparison means anything, because the model, the country and the question are identical for everyone.

2. 2

   The ones you did not know are discovered

   Besides the brands you declared, every brand the model names in your category is recorded. That is where the competitors missing from your list appear, the ones eating your ground for a year.

3. 3

   The denominator comes first

   Share of voice is always shown alongside presence rate. Sixty per cent share over five answers and forty over three hundred are not comparable, and presenting them together without saying so is the category's commonest trick.

• Share of voice and presence per brand, on the same row and with the same denominator.
• The emerging brands the model recognises and you had not declared.
• The per-engine breakdown, because winning in one and losing in another is different information that an average hides.

• Anyone who must justify investment against a specific competitor.
• New categories where nobody yet knows who rules.
• Agencies presenting a comparison that must survive the awkward question.

Available from Starter with a small number of declared brands, widened on Pro and Agency. Emerging competitor detection costs nothing extra: it comes from the same sweep you already ran.

See the full pricing

How exactly is share of voice counted?

Mentions of your brand over total brand mentions in the sweep's answers. The formula is published in the methodology with a worked example. The formula is not the important part: what matters is that it always travels with the absolute count, so you can judge whether the percentage supports anything.

Can I watch a company that is not a direct competitor?

Yes, and sometimes it is the most useful thing. In many categories whoever takes the recommendation is not your usual competitor: it is a publisher, a comparison site or a manufacturer answering the question before you do.

What if a competitor has a generic name?

That is the classic problem, and it is solved as with your own brand: aliases and exclusions in its record. A brand named like a common word and left ambiguous contaminates the whole comparison, and the dashboard warns when it detects that risk.

Does it tell me why they get recommended?

It tells you how much and where, not why. The why is in the pages the model cited, and comparing them with yours is another job with its own tool. Confusing the count with the explanation leads to changing things that were not the problem.