Adding Risk in Measuring Customer Value Using Bivariate Hierarchical Bayesian Approach

Abstract

Hierarchical Bayesian approach to predict changes in individual customer behavior is deemed successful, but often assume that the irrelevance between purchase interval and money. In many situations, this assumption may not be valid. In this paper we proposed bivariate hierarchical Bayesian approach, which allows correlation between them, and can educe conditional probability density function of purchase interval or money. The model is applied to medical instruments sale data to predict customer changes and shows more precise than traditional models. Based on distribution of customer behavior, the concept of customer risk is brought out, including churn risk, decline risk and fluctuating risk, which can be calculated using probability density curve. This value prediction considering risk can be used managerially as a signal for the firm to use some type of intervention to keep that customer