Audience segmentation for Sky Sports

4 August 2015


We already run a monthly tracking study for Sky Sports to monitor the key metrics of their viewers (as well as their potential viewers). But they wanted to know more, especially which types of sport their viewers are passionate about.

Instead of just interpreting descriptive statistics from a set of classic data tables they wanted to ‘get under their skin’ to find natural groupings and better understand them.


Our data from the first 4 months of the Sky Sports tracker (4,000 respondents) gave us a really robust base size to create stable segments from, and to ensure we measured attitudes and opinions that were relevant to the current market place.

We worked closely with Sky to produce a segmentation model that was more than mere demographics; we kept their business needs at the heart of the process. That’s why we presented several models and several cluster solutions to guarantee the best results.

Cluster analysis was the key part of identifying the segments – we could separate individuals via a distance measure to create a clearer distinction between segments and the most cohesion within segments.


Our 5 segment solution allowed Sky to really understand the different types of viewers that watch their channels, which sports they’re passionate about and the importance of each one in the market. Now, rather than a one-model-view for everyone, they have a base to build new marketing plans which will target individuals more effectively.

Sky have since added this segmentation to several other projects, as well as their monthly sports tracker. It’s allowed them to track changes to segments over time and integrate the information across their research channels.

As an added extra, we also produced a classification model that predicts which segment a respondent belongs to. We’ve run it on data for internal projects, as well as other projects run by external research agencies, and recently we’ve developed it into a programmable model which Sky have started to use independently.

Our in-house statistician commented: “This project is a great example of how clients are increasingly integrating segmentation across their research channels; placing it at the heart of the business to drive decisions and add insight.”