bayesQR (version 2.0)

Churn: Customer Churn Data

Description

This dataset is a stratified random sample from all active customers (at the end of June 2006) of a European financial services company. The dependent variable in this dataset is the churn behavior of the customers in the period from July 1st until December 31th 2006. Here a churned customer is defined as someone who closed all his/her bank accounts with the company. Note that all predictor variables are standardized. This dataset is a small subset of the dataset used in Benoit and Van den Poel (2013). The dataset is structured as a dataframe with 400 observations and 5 variables.

Usage

data("Churn")

format

The data frame has the following components:

• churn: churn (yes/no)
• gender: gender of the customer (male = 1)
• Social_Class_Score: social class of the customer
• lor: length of relationship with the customer
• recency: number of days since last purchase

source

Benoit, D.F. and Van den Poel, D. (2013). Quantile regression for database marketing: methods and applications. In: Coussement, K., De Bock, K.W. and Neslin, S.A. (eds.). Advanced database marketing: Innovative methodologies and applications for managing customer relationships. Gower Publishing: London (UK), (in press).

References

Benoit, D.F. and Van den Poel, D. (2013). Quantile regression for database marketing: methods and applications. In: Coussement, K., De Bock, K.W. and Neslin, S.A. (eds.). Advanced database marketing: Innovative methodologies and applications for managing customer relationships. Gower Publishing: London (UK), (in press).