Academic journal article Journal of Business Economics and Management

Towards a Dynamic Analysis of Multiple-Store Shopping: Evidence from Spanish Panel Data

Academic journal article Journal of Business Economics and Management

Towards a Dynamic Analysis of Multiple-Store Shopping: Evidence from Spanish Panel Data

Article excerpt

1. Introduction

The literature on consumer behaviour in the retail market is skewed towards studying consumer loyalty and retail patronage behaviour. Most of the empirical contributions to this field of study have focused on brand choice and, to a lesser extent, on analysing store choice. Nevertheless, over the last three decades, variety-seeking behaviour and multiple-store shopping have attracted the attention of researchers in the consumer behaviour area.

Variety-seeking attempts to stimulate purchasing behaviour by alternating between objects of choice. For example, a situation of boredom caused by a non-optimum level of stimulation from purchasing can lead to multiple-store shopping.

When there are alternatives available, households complement their purchases at their first-choice store with purchases at other stores (Kahn, McAlister 1997; Rhee, Bell 2002). In highly competitive markets, such as the fast moving consumer goods (FMCG) markets, multiple-store shopping is ever more widespread (McGoldrick, Andre 1997; Gijsbrechts et al. 2008).

Multiple-store shopping could be considered as observable either at a given time or over the course of time (Pessemier 1985). Both types of variety offer a wide range of research possibilities that may have implications for business management. In this paper, we are going to focus on studying dynamic variation.

A regular store set is defined as the stores in which households regularly make purchases. These stores complement each other and may even belong to the same retail chain. The budget of the household is allocated among the different stores in the regular store set. Within this set, one store will typically capture the greatest proportion of expenditure, i.e., it is the first-choice store (Rhee, Bell 2002). It is worth analysing the composition of the household store set and its variation over time. This variation in household store set should be analysed and taken into account by retailers. As the variation increases, the portion of the overall budget not allocated to expenditures at the first-choice store will be increasingly important.

The specific objective of this paper is to study, from a dynamic perspective, the determinants of the variation in the regular store set at which households do their FMCG shopping. For this purpose, a Bayesian statistical model is built (Rossi, Allenby 2003; Rossi et al. 2005).

From a revision of the specialised literature, we consider different variables that may be having an effect on the dynamic variation of the regular store set. These underlying factors are: (1) shopping pattern variables; (2) demographic variables of households; (3) demographic characteristics of the shopper and (4) geographical characteristics.

The statistical analysis has been carried out from a dynamic viewpoint using Tobit models. The adopted approach is Bayesian because it allows more flexibility and realism in the modelling process, making inferences that are conditional on the data and that do not depend on asymptotic results (Rossi et al. 2005). Due to the complexity of the analysis, the estimation of the parameters uses Markov Chain Monte Carlo (MCMC) methods and the data augmentation technique.

The primary motivation for this study is to generate new insights into the nature of dynamic effects that characterize household store choice behaviour. From an academic perspective, we analyse differences across households in store choice in a study of the dynamic behaviour of consumers and we empirically estimate the model on scanner panel data. From a managerial perspective, we provide several managerial guidelines for retailers interested in maintaining their market share, taking into account the profile of multiple-store shoppers.

The remainder of the article is organized as follows. In Section 2, a review of the relevant literature is carried out and hypotheses are formulated. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed


An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.