|
Activity Number:
|
280
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Wednesday, August 14, 2002 : 8:30 AM to 10:20 AM
|
|
Sponsor:
|
Section on Statistics and Marketing
|
| Abstract - #300550 |
|
Title:
|
A Bayesian Statistical Model for Retail Category Management
|
|
Author(s):
|
Michel Wedel*+ and Jie Zhang
|
|
Affiliation(s):
|
University of Groningen and University of Michigan
|
|
Address:
|
P.O.Box 800, Groningen, International, 9700 AV, Netherlands
|
|
Keywords:
|
Category Management ; Bayesian Statistics ; Pooling ; Nonparametric Smoothing ; Retyailing
|
|
Abstract:
|
The core topic of this study is to offer a set of tools for implementing category management (CM) as a response to the current needs of the retail industry. We develop a Bayesian statistical model that offers a comprehensive description of price-demand relationships and competitive price reactions for a single retail chain. By using a Bayesian pooling method, our model avoids the problems of collinearity and instability of estimates that result from estimating all cross-price demand and reaction effects individually. Price endogeneity is accounted for in the demand functions. Furthermore, we account for heterogeneity in store sales levels and price settings through a parsimonious random effect specification. Finally, our model includes a flexible smooth function to control for time-varying trends in sales levels. Based on parameter estimates of the model, we then provide normative guidelines on how to set retail markups of individual brands to maximize the retailer's category profit. We examine long-term profit implications of adopting category management for retailers based on sales data in the toothbrush category.
|
