Services / Share of Voice

Find every single mention of your brand out there with Share of Voice.

Evaluate only the relevant coverage of one or a whole family of brands, and competitor analysis precisely with our Share of Voice technology. We can also integrate automated sentiment analysis for greater precision.

How it’s done

Counting shares is not just math. It’s reading between the lines. Here’s how we do it.

1 We collect insights for you.

Together we predefine brand search criteria and keywords to look for. Then we evaluate the number of times your brand and products have been mentioned online using Natural Language Processing (NLP). We turn online content into plain text and translate it into English if needed.

2 We can distinguish correlation.

Our technology highlights all pronouns and other contextual synonyms used in the same regard. To do so, our Share of Voice solution analyses written speech and identifies the different parts of a sentence, such as prepositions and predicates. It follows the logical structure of the text and detects everything related to your brand through its consistency.

3 We put it all into a density scale.

After identifying all the relevant content, we assign the brands into a density scale as per the overall number of mentions in all content. You can thus find out what influence each particular brand brings. We reference all collected articles.

Services

Context matters.

A brand mention can have a vastly different influence depending on the contextual scope where it’s reflected. Here are the factors that we observe:

Exact match mentions of the brand
Non-exact match mentions of the brand
Exact match mentions of other brands
Non-exact mentions of other brands
Percent of the sentences that contain a brand mention
Percent of exact matches of the brand towards exact matches of other brands
Percent of mentions of the brand towards mentions of other brands
Occurrence in article titles
Occurrence in first and last sentence

It’s important because

Machines can optimise SoV processes

To complete the described evaluation process, a person needs to fully read all articles and highlight relevant mentions manually. Aside from being extremely time consuming, this process allows room for error. People assess content differently and often assign different scores to the same articles. A machine, on the other hand, would never read diagonally, skip a mention or put personal bias into the analysis.