A Large Cable and Telecommunications Company Tackles Customer Churn

What if you could predict customers with the highest risk of defecting to a competitor and contact them before they decide to make a change?

This large cable and telecommunications company taps into a stream of subscriber information, mining analytics and identifying patterns that help executives recognize the indicators of customer churn and mobilize resources to prevent it.


Even as telecommunications companies relentlessly try to lure customers away from each other they all know it costs much less to keep existing customers than recruit new ones. Wireless customers, for their part, notoriously lack loyalty, needing nothing more than the right incentives - be it service or price - to move to a competitor. Increasingly then, the goal among providers is to model, predict and influence customer behavior by proactively addressing issues impacting satisfaction. This large U.S.-based provider wanted greater insight into why customers churn and which of its customers were most likely to churn in the future. Such information could be used by those responsible for monitoring the customer base to identify subscribers who may, by virtue of their usage pattern or any number of other variables, be forecasting intent to migrate to another provider. However, the company’s existing systems were able to take only a “sample set” of customer data, and over time the data revealed that limited samples were not adequately enabling the company to fully mine historical customer data for its predictive value.

Business Analytics and Intelligence

Predicting the future by analyzing the past, this telecommunications company was able to increase subscriber retention, in part by instituting more proactive and informed customer service practices and by offering customers pricing and service plans and perks more closely tailored to what they want and need. The IBM PureData System for Analytics allows the company to leverage 100% of the data capture and gain insights never seen before. Vast amounts of data are connected to data models that automate the process of profiling customers and predicting their actions months and years into the future. The company now targets individual customers with offers and individualized attention that can mean the difference between loyalty and churn. No other system/technology and investment model was able to take 100% of the data and perform analytics without limitations.

Real Business Results

• Reduces customer churn by increasing customer satisfaction

• Connects vast amounts of customer data into the data models that facilitate accurate profile scoring and segmentation

• Incorporates predictive modeling into its customer satisfaction initiatives

• Automates analytics, reducing complexity and cost

Solution Components

• Software: IBM Netezza (IBM PureData System for Analytics), IBM Cognos, IBM SPSS

• Services: Micro Strategies Architect, Design, Development and Program Management services