Customer churn is defined in most of companies as an expected loss of clients. Recent studies have increased the importance of this loss by comparing the cost of attracting a new customer versus maintaining one, being the former more costly.
Customer churn includes the loss of future sales and resources that were expended in attracting them as customers. A recent study of a consultant company concluded that the cost of attracting a new client is ten times the expense of maintaining one, which derives in the prediction that by 2015 telecommunications companies would have to increase in more than 50% their present budget focused on customer retention. (Alacer Group 2012)(Sarausad 2013)(Stamford 2012)
Customer churn, which is the number of clients that have abandoned the company in a period of time, is quantified easily by companies that keep a contract with their clients, but in a company without contracts (i.e. shipping and online retail) an additional system to keep track of their clients is needed. For that purpose, companies have implemented fidelity programs to motivate clients in providing their information and identify their purchases. (Cuenco et al. 2011) (Scanlon 2002)
Companies have implemented methods to identify customer churn based in their behavior level of risk and value. Clients with high probability of churning and high level of profitability are usually the ones that the company focuses their retention efforts.
Clients likely to churn can be identified by studying their activity (i.e. frequency of purchases) using control charts. Control charts display the clients’ activity in a timeframe, which can be diary, weekly or monthly, for at least thirty gaps and each observation is compared to an approximation of a normal behavior by assigning “normality” limits. These limits can be based in the behavior of other clients or of himself by extracting a time-lapse when the client had normal activity (which should have between 8 to 50 sequential observations with at least one purchase in average). Based in this limits, the company can identify either the client is having a normal behavior or has a probability of churning. (Cuenco et al. 2011)
The next figure shows an example of a control chart used for a client divided in two periods; the first one is used as the normal behavior of the client and the next defines that the client has probably churned. The evaluation of the chart to define this loss possibility can be performed visually or using a defined criteria like the following:
CHURNED CLIENT (CUENCO ET AL. 2011)
Retention efforts are usually focused over the relevant clients. In order to identify the area of focus a company can adopt two common tools based on customer segmentation. The first of them is the segmentation by dimensions, where these dimensions are related to the client (i.e. their attitude by analyzing their demography) and their activity (i.e. frequency of purchases, level of satisfaction, and profitability). Segmenting clients with valuable characteristics for the company using these dimensions will allow focusing the retention efforts to more profitable clients or that the company can’t afford to lose. (Borna 2000) (Alacer Group 2013)
The segmentation by quadrants is similar of the dimensions but more visual. This segmentation can be displayed in a matrix like the following one, where one axis defines the client’s profitability and the other its churn probability. Each dimension is quantified to allow the comparison between clients. Four quadrants can be obtained initially depending in their grade:
The profitability of the client is computed based on their customer lifetime value, which uses the customer’s average purchases, percentage of profit per purchase and average number of purchases per period. On the other side, churn probability is computed by comparing the characteristics of the clients with the characteristics of the people that have churned in the past. (Alacer Group 2013)(Harvard Business School Publishing 2007)
MATRIX OF SEGMENTATION BY QUADRANTS (ALACER GROUP 2013)
Other indications used for churn studies it the churn rate, customer lifetime, acquisition rate, and others, which are calculated based in the company’s requirements. Companies that have implemented activities to reduce the customer churn had as a result a reduction in publicity expenses and expected loss of income, improved retention systems and increase of the number of relevant clients and competitiveness. (Borna 2000)(Sukow and Grant 2013)
References:
Alacer Group. 2012. “Alacer Shows How Big Data Analysis Can Reduce Wireless Churn.” Alacer. http://www.alacergroup.com/wireless-churn-infographic/.
———. 2013. “Using the Profitability Factor and Big Data to Combat Customer Churn.” Alacer Group. July. http://www.alacergroup.com/wp-content/uploads/2013/08/Alacer-Telco-Churn-white-paper.pdf.
Borna, Claude. 2000. “Combating Customer Churn.” Telecommunications, Americas Edition 34 (3): 83–85.
Cuenco, Michael, Crystal Shi, Balaji Padmanabhan, and Alan Hevner. 2011. “Challenges With Churn.” ASQ Six Sigma Forum Magazine 11 (1): 20–26.
Harvard Business School Publishing. 2007. “Customer Lifetime Value Calculator.” Harvard Business School Publishing Corporation. http://hbsp.harvard.edu/multimedia/flashtools/cltv/.
Sarausad, Ed. 2013. “Reality Check: Using Profitability to Determine Churn Strategies.” Mobile and Wireless News – US. http://www.rcrwireless.com/article/20130611/opinion/reality-check-using-profitability-determine-churn-strategies/.
Scanlon, Mavis. 2002. “The True Cost of Churn.” Cable World 14 (28): 32–40.
Stamford, C. 2012. “Gartner Says Organizations That Integrate Communities Into Customer Support Can Realize Cost Reductions of Up to 50 Percent.” February 21. http://www.gartner.com/newsroom/id/1929014.
Sukow, Anthony E. R., and Rebecca Grant. 2013. “Forecasting and the Role of Churn in Software-as-a-Service Business Models.” I – Business 5 (1A): 49–57.