Three Requirements for Reducing Customer Churn
Source: www.spss.com
Copyright SPSS, Inc. 2004


Retaining existing customers can be one of the most effective ways to increase organizational profitability. Here, we outline the three key requirements for an organization to create a successful customer churn program.

In most markets, retaining customers is a necessity. The proliferation of customer data warehouses and sophisticated analytical tools is making technological solutions to this problem increasingly popular. However, few organizations are successfully using technology to reduce customer churn, due to failing in at least one of three key areas.

Having the Right Data. Churn analysis software looks for patterns of behavior or characteristics in customer data that indicate loyalty. Hence, if the data is bad or incomplete, the analysis may be worse than useless, as organizations spend money arriving at wrong answers and then more money acting on them. Out-of-date, incomplete or inaccurate data is an issue that organizations are taking steps to resolve independently of specific applications, but this problem will never be completely fixed. Organizations cannot wait for data cleansing to be completed before initiating a customer churn project. The challenge for IS and business units is to decide when the data is “good enough” to be used. 

Incomplete data is problematic because it is difficult to ascertain which data predicts customer loyalty, so performing analysis without a key data element (e.g., customer buying channel), may cause the analysis to be flawed. The choice is between collecting as much data as possible (expensive and time-consuming) and missing key determinants of customer loyalty (expensive and time-wasting). The best practice for many organizations is to purchase extensive third-party data on a small group of customers that can be used to build churn models. Once the important elements are identified, key variables can be purchased for the entire group to be scored.

Having the Right Tools. Many applications can be used for customer churn analysis, and the tool chosen should depend on the users’ skill set and the range of analyses they perform. The three most-common approaches are generic data mining workbenches (e.g., SAS Institute’s Enterprise Miner, SPSS’s Clementine and IBM Intelligent Miner); application specific tools (e.g., Churn/CPS from slp Infoware); or working with an analysis service provider. The key to success is selecting tools that will actually be used, either by a hard core of power users supporting the entire business or by marketing analysts focused on specific issues. Customer bases and behaviors change over time, so analysis of their behavior is unlikely to be a one-time affair. An application that people are willing and able to use regularly will be more useful than one employing the newest and most-sophisticated data analysis techniques. 

Having the Right Strategy. The final requirement to generate ROI on customer churn analysis is to actually use the analysis, which is also the most under-considered topic within organizations. The results of churn analysis cannot be applied in isolation — other important variables such as customer profitability must also be considered before deciding on a response to a possible customer defection. Organizations tend to undertake a combination of three strategies in response to customer churn analysis.

1. Focus on Determining Variables: Having identified the characteristics of defectors, the organization changes its relationship with those segments. Typical behavior would be no longer targeting that group or running local marketing activities to counter a specific competitor. 

2. Target the Individual: Organizations immediately communicate with high-risk customers to offer specific incentives or promotions in order to retain their business. 

3. Change the Interaction Profile: The customer’s loyalty score is treated as just another variable during interactions. Therefore, it is considered when offering new products, deciding on the service level they should receive through a call center or resolving service disputes.

Bottom Line: Managing customer churn will be an increasingly important task for organizations. Assuming that data analysis is the best strategy to achieve this (which is not always the case), organizations must ensure they do not wait too long for the right data, do not choose a powerful tool that will be used once and forgotten and, most importantly, have a clear plan for how analysis results will be used to retain customers.