Securing revenue with targeted retention initiatives in telecom
Context & Objectives
For several years, a leading Belgian telecom operator was suffering from a structural customer retention problem. With a rate well above 10%, the estimated churn impact cost in the hundreds of millions of euros each year.
In the telecom industry, retention is made more complex by originating from a combination of factors. Some are external, such as increased operator promotional bundles grouping mobile, internet, and TV. Others are internal: core services quality, price, and customer support efficiency in solving issues. The siloed data and their inconsistent trustworthiness made the task of understanding churn even more challenging.
They turned to Agilytic to understand churn drivers, target at-risk customers, and improve customer retention.
Approach
We started with a workshop with client teams to define the relevant KPIs by one of three problem sources:
Administrative (e.g., errors in contracts, untimely changes in pricing formulas, price increases)
Invoicing (e.g., billing error, overconsumption, extras not included in the fixed price)
Technical (e.g., network performance, service interruption)
Next, we consolidated the customer data (CRM) with billing and network data into a homogeneous data warehouse fit for analysis.
We then built a retention model to identify the churn drivers and then score the churn propensity of existing customers. We managed to forecast up to three months before churn’s actual appearance, which proved critical for adequate planning of the retention campaigns.
Each client received a score with two arguments justifying their churn probability. But we did not stop there. We pushed those scores directly into the operator’s CRM tool. This way, front-line support staff operators had all relevant information for inbound and outbound calls.
Wrapping up on the analytics side, we developed the nightly model automatic update that pushed a CSV directly into the CRM.
We ended the project with a comprehensive handover, including training sessions and exhaustive documentation.
Results
For the operator, the results were multiplied:
A much clearer understanding of the root causes of customer churn
Actionable customer segments each assigned personalized next-best retention initiatives
A 10% decrease in annual customer churn (equivalent to ±1% of the portfolio)
The project was a success. Today, churn modeling is an integral part of any modern telecom operator and is considered worthy of continuous optimization.
