Monitoring usage behavior in subscription-based services using control charts for multivariate attribute characteristics

Industries such as telecommunication, cellular phone services or online content providers supply varieties of services to their customers for a specific subscription fee for a given duration of time. Controlling churn rate is an intricate operation for such industries because of the multiplicity of customers, diversity of services and various usage patterns. It may be recorded, in its simplest form, for each customer whether she/he uses different types of services during each data gathering period. Hence, usage behavior of an individual customer could be represented by a binary vector, which may be statistically modeled as a multivariate discrete distribution. This paper presents how to employ control charts for multivariate attribute characteristics to monitor systematically the mean of the customers¿ usage pattern along the time. It is shown that applying the appropriate method allows the firm to detect quickly significant changes in usage behavior and so prevent from customer churn by remedial marketing policies.