Higher profits due to store-level optimization of retail assortments
January 13, 2014
As a retailer, how do you choose the right assortment? The right assortment determines why a customer opts for a certain store and not for the competitor. Moreover, it determines what consumers buy when they enter the store. But the large – and increasingly larger - range of products and the limited amount of shelf space complicate the choice for an optimal assortment of products. In additon, the problem should ideally be solved for every individual store a retailer owns as stores are likely to differ in terms of size and characteristics of the store regions they serve.
Image: © Nationale Beeldbank
Associate professor Marketing Robert Rooderkerk of the TIAS School for Business and Tilburg University, together with Harald van Heerde of the Massey University in New Zealand and Tammo Bijmolt of the Rijksuniversiteit Groningen, devised an optimization metholody that determines which assortiment a retailer can best keep in stock at a given store.
The study 'Optimizing Retail Assortments' of Rooderkerk, van Heerde and Bijmolt was recently published in the top journal Marketing Science. The method as devised by Rooderkerk and his colleagues yields a 37 percent higher profit for the retailer. If the method is also used to determine the optimal prices of the products included in the assortment, then the profit achieved by using the method increases to no less than 43 percent.
Decision based on a feeling
Nowadays, retailers use various methods to decide which products ultimately end up on the shelf. Some use software, others come to a decision based on their gut feeling. The paper describes an assortment-optimization-model that predicts store-level SKU demand and that also predicts changes in the demand should the assortment change. To retain a parsimonious model of store-level SKU sales Rooderkerk and colleagues built an attribute-based model. Instead of directly comparing products themselves, it compares them as bundles of attributes such as brand, size, and fragrance. In addition, their study consists of a novel heuristic that uses the estimated demand model to search for the assortment that maximizes expected profit.
The methodology (model + heuristic) was applied to three years of weekly store-level scanner data on 61 different laundry detergents sold at a French supermarket chain with 54 stores. The results were encouraging. Using the model to optimize just the assortment caused the expected profits to increase by 37 percent. The price optimization alone led to a profit increase of 7.9 percent. Jointly, the optimization model resulted in a 43.7-percent rise in profits.