Using Support Vector Semiparametric Regression to estimate the effects of pricing on brand substitution1

María Pilar Martínez-Ruiz and Miguel Ángel Gómez-Borja
University of Castilla-La Mancha

Alejandro Mollá-Descals
University of Valencia

José Luis Rojo-Álvarez
University Rey Juan Carlos

INTRODUCTION

Retail marketing managers are primarily concerned with the allocation of scarce marketing resources, such as sales promotion, for improving the market and profit performance of their assortments. The quality of their decisions depends greatly on their understanding of how customers will respond to these strategies (e.g. Blattberg & Neslin 1990; Chen et al. 1999; Dekimpe et al. 1999). In grocery retailing, sales promotions represent an important percentage of the marketing mix budget and, as such, retailers now spend more money on promotions than on advertising. In particular, temporary price reductions can be beneficial to grocery retailers, especially since some of them are not available through other marketing tools (Farris & Quelch 1987). The price promotions literature cites several reasons for the use of temporary price cuts by retail managers (Blattberg & Neslin 1990), which are: