A clearer view of song performance

Blending DJ creativity and AI-driven analysis to help improve the listener experience.

Who are you?

Global is one of the world’s leading media and entertainment groups, with a portfolio of some of the most respected and recognized media brands and radio stations. We’re Europe’s largest radio company, reaching over 25 million listeners per week. Our brands include the likes of Capital FM, Capital XTRA, Radio X, Classic FM, Smooth, Heart, and LBC, amongst others.

What was your challenge?

We’re always looking at ways we can better personalize the content we provide and improve the Global experience for our listeners. With this in mind, we wanted to gain a clearer understanding of media asset performance – mainly songs – across some of our major commercial stations.

We needed to be able to predict when certain songs or advertisements would lead to listener dropoff/churn, and use this information to influence our future strategic thinking and programming – combining the creativity and expertise of our DJs and programmers with data-driven insights and learnings.

In short, we were looking to provide our listeners with more of the content they want to hear, at the right time, to help improve station retention rates and offer an exemplary listener experience.

Working with Peak has given us a unique, data-led view of the performance of our song catalogue and playlists, allowing us to better predict when songs will reach the end of their popularity lifecycle and lead to listener disconnect.

Mark Crawford

Head of Brand and Audience Insight, Global

What did Peak do?

Peak used its AI platform and Customer Intelligence suit of applications to help us make more of our data and gain a better understanding of song lifecycle and popularity decline.

This gave us a clearer view of song performance, in terms of when a particular song started to result in listeners switching off from our stations, and how factors such as time of day or day of the week impacted these listening trends.

The machine learning models Peak built leveraged the listening data of each song on our current playlists, and allowed us to see the number of times that song was played each week and how many people (and when) would disconnect during that song.

We were then able to map these trends over time, identifying the optimal play count on a song-by-song basis, and predict future song performance and trends.

These insights are being used to help inform our future strategic thinking, bringing together our presenters’ industry expertise and a powerful, AI-informed analysis to help us provide our listeners with a better radio experience.

Mark Crawford

Head of Brand and Audience Insight, Global

What’s the upshot?

Our work with Peak is providing us with a number of new ways to analyze our stations and the songs our presenters and DJs are playing.

We’re now able to use these new, invaluable metrics to inform our programming decisions; we can rank songs in terms of predictive bounce and disconnect performance, bringing together the creativity of our team and an AI-driven analysis to help improve the listening experience of Global station fans.

Simply put, it enables us to provide our listeners with more of the media content they’re interested in, at the right time, to drive brand loyalty and offer a more tailored, optimized experience for listeners of our major commercial radio stations.

What could AI do for your business?

Get in touch with our team of experts to start your own AI journey with Peak.

AI | Technology

State of AI report

AI is a powerhouse, and is changing the way we work.

Sign up to the Peak newsletter

Get the latest Peak news and AI insights delivered straight to your inbox