Product Variety Under Brand Influence: An Empirical Investigation of Personal Computer Demand

Posted: 3 Jun 2002

See all articles by Kai-Lung Hui

Kai-Lung Hui

National University of Singapore (NUS) - School of Computing

Date Written: December 2002

Abstract

Theoretical research on self-selection and cannibalization suggests that brand quality may affect consumers' choice of products from a menu of similar offerings. However, few studies have empirically considered the effect of product variety under brand influence. This paper examines the demand response to the proliferation of personal computer (PC) models. Using both the central processing unit and brand as segmentation variables, I set up a two-level, nested generalized extreme value discrete choice model to estimate PC vendors' brand qualities and product similarities. Based on the estimates, I infer the relative efficacy of product variety for firms which possess different degrees of brand qualities. My results suggest that consumers view PCs from the same firm as close substitutes, and the closeness of the PCs correlates positively with brand quality. This implies that the short-term demand gain of product variety may be somewhat limited for branded multiproduct firms. I also discuss possible causes of brand quality, and explore the role of product line extension in building long-term brand reputation.

Keywords: product variety, brand quality, self-selection, cannibalization, generalized extreme value, discrete choice model

JEL Classification: L10, L63, M10, M31

Suggested Citation

Hui, Kai-Lung, Product Variety Under Brand Influence: An Empirical Investigation of Personal Computer Demand (December 2002). Available at SSRN: https://ssrn.com/abstract=312560 or http://dx.doi.org/10.2139/ssrn.312560

Kai-Lung Hui (Contact Author)

National University of Singapore (NUS) - School of Computing ( email )

Department of Information Systems
3 Science Drive 2
Singapore 117543
+65 874 4483 (Phone)
+65 779 4580 (Fax)

Here is the Coronavirus
related research on SSRN

Paper statistics

Abstract Views
1,160
PlumX Metrics