Database marketing modelling for financial services using hazard rate models

Abstract

This paper investigates the effectiveness of using direct mail message as an element in the communication mix of a financial retail company. The focus in this study will be on interpurchase time, i.e. duration times between the last purchase of any type of financial service by a customer at that bank and the purchasing of a personal loan.

Our findings indicate that (1) direct mail message have a very significant effect on shortening the interpurchase time, (2) the combined effect of several direct mail messages exhibits diminishing returns to scale, (3) incorporating time-varying variables in the model significantly improves model results, (4) Ownership of certain other services has a significant impact on the probability of purchase of the personal loan, and (5) conditional probability of purchase is found to increase with time elapsed since last purchase. The latter finding shows that the receny effect, known to be very important in most RRM models, may to dependent on the product category.

Keywords:

• Database marketing
• interpurchase times
• financial services
• Hazard rate models
• cross-selling