Forecasting the Effects of Non-Linear Shipping and Handling Fees

While the creation of a virtual store eliminates many of the costs incurred by the traditional bricks and mortar retailer, the spatial separation between customers and the retailer’s distribution facility may result in significant distribution costs. In these circumstances, the determination of shipping and handling fee schedules becomes a key element of the marketing mix. This is a non-trivial task since there are many possible variations of shipping schedules. In addition to setting the level of fees, the firm must choose between constant, increasing (in order size), or decreasing schedules. These multiple decisions are important because consumer response to shipping fee schedules may be in terms of both order incidence and order size. The objective of this paper is to understand how shipping and handling charges impact customer order incidence and basket size. We begin with a conceptual model of the process by which consumers select shopping baskets. This model provides a convenient framework for predicting how consumers will alter behavior in response to various fee structures, and also motivates an approach to empirically estimating the impact of nonlinear shipping fees schedules. Our empirical approach generalizes the standard ordered probability model to account for the impact nonlinear and discontinuous shipping fees have on ordering decisions. The model is illustrated using a unique database provided by an online grocer. Our results indicate that shipping fees significantly impact order incidence as well as basket size. Profitability implications of different types of shipping fee schedules for online and other direct marketers are discussed.