Brand Portfolio Promotions
Published 8/1/2008
Author: Anocha Aribarg and Neeraj Arora
View this content
Executive Summary
Large manufacturing firms, such as General Mills, Procter & Gamble, and Unilever, often engage in promotions that encompass multiple brands and categories. Such promotions are mailed to carefully selected consumers at a predetermined time interval and coordinated across brands that the firm owns. In this article, the authors refer to such a corporate-level promotion as “brand portfolio promotion” (BPP). For example, General Mills mails a magazine-like promotion Qué Rica Vida (www.quericavida.com) and Serving Up Soul (www.servingupsoul.com) to millions of households.
The goals of this article are to (1) develop a modeling framework to examine the effectiveness of BPPs, (2) characterize the nature of BPP effects on sales, (3) quantify effectiveness of BPP relative to other forms of sales promotion (e.g., feature), and (4) provide guidelines to enhance the effectiveness of a BPP program. The authors develop a modeling framework with a system of equations that link a BPP program targeted at individual households to weekly store sales. It captures the magnitude and shape of each BPP’s effect at the brand level using a gamma density function. The model structure is parsimonious yet flexible. It captures heterogeneity in the magnitude and shape of sales lift across brands as a result of BPPs and accounts for diminishing returns and the intertemporal effect of BPP exposure. To incorporate store-level heterogeneity, the authors impose a hierarchical Bayes structure. A particular advantage is that the model could be applied to any form of discrete promotion in which the effect duration is unknown.
The authors test the proposed model using a unique data set provided by a multinational firm that offers brands across more than 50 product categories. The selected data set involves ten brands over a 210-week period and 14 BPPs. The results demonstrate heterogeneity in magnitude and shape of the BPP effect across brands. Evidence in support of diminishing returns for brand exposure and intertemporal BPP effect also exists. The authors demonstrate the generality of the model by using it to estimate feature effects. Unlike a feature that exhibits an instantaneous effect, they find that a BPP has a nonmonotonic carryover effect after the initial launch. Several model-based metrics that allow an objective comparison of return on investment from a BPP versus other forms of promotion are presented. The results suggest that a BPP, when contrasted with feature, leads to higher sales lift per household for some of the brands. The authors develop an optimal exposure allocation procedure based on the proposed model that informs (1) which assortments of brands to promote across multiple BPPs and (2) the exposure level for each brand.
A unique aspect of a BPP is that though it is targeted at the household level, its effect needs to be measured at the market level. The“micro–macro” approach to measurement, modeling, and subsequent optimization in this article naturally extends to a variety of retailing contexts, such as Home Depot and Lands’ End.
Biography
Anocha Aribarg is Assistant Professor of Marketing in the Stephen M. Ross School of Business at the University of Michigan. She has an undergraduate degree in Statistics/Actuarial Science from Chulalongkorn University (Thailand), an MBA from University of Wisconsin–Milwaukee, and a PhD in Marketing from the University of Wisconsin–Madison. Her primary research interests involve integrating econometric modeling, Bayesian statistical inference, conjoint choice experiment, and consumer behavior theories to understand consumers’ individual and group choice behavior and decision-making processes. She aims to apply such understanding to help firms make more effective marketing decisions, particularly in the area of product portfolio management.
Neeraj Arora is Arthur C. Nielsen, Jr., Professor of Marketing Research at University of Madison–Wisconsin, where he also serves as the Executive Director of the A.C. Nielsen Center for Marketing Research. He has an undergraduate degree in Engineering from Delhi University and an MBA and Ph.D. from The Ohio State University. Neeraj’s research focuses on business problems that require econometric modeling of individual or group choice. Most of his research builds on psychological and microeconomic models of consumer choice that use Bayesian statistics. He serves on the editorial board of Journal of Marketing Research and Marketing Science. His research has appeared in Journal of Marketing Research, Marketing Science, Journal of Consumer Research, International Journal of Research in Marketing, and Marketing Letters.
Journal of Marketing Research, Volume 45, Number 4, August 2008
View Table of Contents.