Optimizing vendor selection in a two-stage outsourcing process

The decision processes surrounding outsourcing are complicated by the very nature of uncertainty involved in the outsourcing process and by poor vendor management. In this study, we focus on vendor selection, one of the two basic issues of vendor management in outsourcing. Due to the limitation of the classic one-stage vendor selection model, we propose a two-stage vendor selection research framework in outsourcing. The first stage is a trial phase that helps the client to find the best match between the vendor and the outsourced project. In the second stage, the client employs the chosen vendor for the full implementation of the project. We formulate this selection decision under the two-stage framework as a combinatorial optimization model. We analyze the complexity of the problem and develop a solution procedure to find the exact optimal solution. By applying this model to numerical case studies, we demonstrate that benefit to adopt two-stage process to the vendor depends on information improvement in the first stage and the client's ability to adapt to updated knowledge. We also argue that the selection of vendors for the first stage testing is more about creating a good vendor portfolio than simply picking the frontrunners.