Alleviating demand uncertainty for seasonal goods: An analysis of attribute-based markdown policy for fashion retailers

Abstract

This study develops a model for pricing seasonal goods, helping retailers better cope with demand uncertainty. Specifically, to improve price markdown policies for fashion apparel retailers, we uncover the relationship between fashion product characteristics and consumers’ within-season product adoption behavior. We develop an aggregate demand model and estimate it using a finite mixture model on data from a leading specialty apparel retailer. The demand model identifies two latent classes of products based on the evolution of demand within a product’s lifecycle (i.e., sharply deteriorating vs. stable demand), and accounts for unobserved heterogeneity where mixing probabilities are defined as functions of fashion product attributes. We then run hundreds of counterfactuals to evaluate pricing policies in terms of: (1) timing and (2) depth of price markdowns. Our findings show that the retailer should implement middle-of-the-season price markdowns for products that have high initial prices, are introduced in the summer/fall, or are darker in colors. For other products, markdowns should be shallower and earlier in the season. We show that ignoring the cross-product heterogeneity in within-season demand could result in a 5.77% reduction in revenues. Our solutions provide managerial implications and enable the retailer to predict products’ demand patterns prior to launching products in the market.