Article ID: | iaor20172071 |
Volume: | 47 |
Issue: | 3 |
Start Page Number: | 230 |
End Page Number: | 243 |
Publication Date: | Jun 2017 |
Journal: | Interfaces |
Authors: | Natarajan Harihara Prasad, Bernales Patxi J, Guan Yongtao, Gimenez Patricia Souza, Tajes Mario Xavier Alvarez |
Keywords: | supply & supply chains, demand, statistics: inference, inventory |
Intcomex, a large global distributor of information technology products, was experiencing severe stress in its supply chain from a rapidly expanding set of products in its catalog. The company proactively partnered with an academic team to develop a data‐driven approach to address this issue. Rather than adopt potentially suboptimal rules of thumb for product selection (and exclusion), the project team developed a composite method for rationalizing stock‐keeping units (SKUs). This approach allows the company to statistically estimate product demand and product substitution using profit‐based optimization of the product selection decision. Intuitively, the approach seeks to leverage the company’s ability to substitute products by eliminating low‐profit SKUs for which substitutes are available. It accounts for all costs incurred over a product’s life cycle, incorporates other important contextual considerations, and delivers tailored recommendations for each product category and geographical market. We implemented the composite method as two software modules–one for statistical estimation and one for assortment optimization. Using the software tools of the composite method on a product category in the company’s Uruguay operations, we generated an 18 percent increase in profits from rationalizing the company’s product catalog. Despite the streamlined catalog, more than 97.5 percent of the product demand was served; in addition, revenues increased because of the better match between supply and demand.