Skip to main content

Select your location

Revving up the recommendation engine

Media and telecommunications company

User receiving a mobile notification while watching television

Making it personal

  • Category: Data and Analytics Machine Learning

Our client is a provider of subscription television, internet and mobile services. It has an international reach totalling over 22 million customers, and a history of developing innovative products and entertainment content.

We were asked to increase the impact of customer marketing campaigns . The client wanted us to utilise machine learning to give the audience a more personalised and relevant experience in real time.

Using the data

An initial MVP was created in close collaboration with expert promotion planners.

We explored the possibility of a Recommender system, based on collaborative filtering Machine Learning techniques. Using the client’s Cloud infrastructure, our aim was to store data and process it to ensure privacy and protection.

First party viewing data was used to create a personalised model for targeted promotions. We iterated based on feedback from the client to create the most accurate and interpretable model possible. A/B testing was employed to quantitatively show increased performance.

Highly recommended

We delivered a production promotion recommendation engine using Machine Learning to create personalised and relevant targeting.

This intuitive web-based system allowed promotion planners to feedback the model and adjust recommendations. The end result was an accurate and interpretable model based on the client’s own infrastructure for iteration and internal utilisation.

Most importantly, it resulted in a 113% uplift of views for promoted content for targeted customers versus the existing solution. That’s a job well done.

Good data in, good data out. Unlock the value of data.

Contact us