A decision support model for supplier portfolio selection in the retail industry

In this paper we propose and test a staged knapsack model to create a supplier portfolio for retailers. The first two stages in the model use judgement criteria related to supplier characteristics and supplier fit, gathered from managers to create a portfolio that optimizes supplier rank and fit scores. In the third stage we optimize cost by considering landed costs for each supplier in the portfolio to arrive at the final set of suppliers. The results indicate that the criteria used, and the staged refinement of the supplier pool resulted in a better fit than just considering the landed costs of the final pool of suppliers without rank and fit score constraints. The proposed method will enable purchasing managers to strategically create a supplier portfolio that will suit the retailer's needs more effectively.