Abstract:

Determining the number of lines per range requires assortment decisions on the number of items and facings displayed in the stores. The constraint of limited shelf space and demand and cost interactions among the various products of the assortment must be considered. In many cases, retailers try to solve the problem by trial and error. An alternative is to conduct controlled experiments using experimental designs. They allow definite statements that have a known and controllable probability of being correct and consequently provide a reliable basis for decision-making. The testing procedure is illustrated by two case studies using latin square designs to answer the fundamental questions of adding/deleting a product and changing shelf space allocation.