Optimizing and Evaluating Retail Assortments for Infrequently Purchased Products

This article develops a methodology for choosing optimal retail assortments for infrequently purchased products and assessing the robustness of such assortments with regard to shifts in customer preferences. The approach incorporates heterogeneous consumer preferences and uses integer programming to determine optimal retail assortments. The robustness of the resulting assortments is examined using alternative consumer choice assumptions and ranges of parameter values. The methodology is illustrated using adaptive conjoint data collected via the Internet and is assessed against the sales results from an assortment offered by a national retail chain.