Early-life, quad play downspin for a major Telco

Services

Recommendation Engine, Data Enablement, Data Modelling

‍

Challenge

A major Telco operator had a 90-day point at which customers could downspin their products. Customers were doing this frequently (and this also led to end of commitment  churn). This was at a point in time where the TiVo service and related VoD products were coming through in the local market (Sweden) and the operator had put a lot of faith and investment into the TiVo product.

‍

Proposal

Our proposal was a build on our previous work where we deliver a structured TiVo usage dataset. We suggested creating an early life customer recommendation programme.
 
This began with a customer segmentation on engagement in content, service and product usage. This included content category, showing the type of content that the household watched.
 
Utilising these, we created a content/feature recommendation model that mapped usage, the adoption step that would increase product engagement (for example using VoD or a premium channel). We used decision logic to then select from content library to create a personalised and highly relevant recommendation set, at an individual customer basis, all fully automated via the BAU campaign operations teams.

‍

Team

Analytics Director, Programme Lead, Senior Data Consultant, Senior Data Modeller

‍

Outcome

The recommendations ran from week 4-12 and led to a significant uplift in the depth of customer product engagement (both premium content and product features) and then, naturally, improved downspin performance for the business. Just as importantly, it was a commercial use of the dataset that we had created for them and showed the importance of good data in running a business.