Abstract: Increasing reliance on outsourcing hasmade the issue of supplier selection even more critical to the success of a supplychain, or more precisely a supply network. Nowadays, the market competition isbased on the whole supply network, rather than an individual supplier.Nevertheless, suppliers of different components jointly pay an important role toaffect the attritubes of the finished product as well as the performance of thewhole supply network, in terms of some key coupling attributes criteria, suchas price, quality, delivery reliability and delivery lead-time of finishedproduct. This reveals that the best individual suppliers working together may notproduce the best finished product, or give rise to the best supply network. However, this fact has been ignored and muchattention has been paid to single supplier selection in the literature. Thispaper presents a supplier combinatorial selection method to investigate theissue of multiple component supplier selection. There are two steps in thismethod. First, a prequalification step is done by AHP/DEA to delete the non-paretoindividual suppliers. Second, the best supplier combination is selected basedon some coupling attributes. This paper also compares the results between individualcomponent supplier selection and supplier combinatorial selection. Finally, a scenarioanalysis is given.
Key words:supplier combinatorial selection; data envelopment analysis; coupling attribute
1 Introduction
Intoday’s fierce competitive environment, which is characterized by thin profitmargins, high consumer expectations for high quality products and shortlead-times, companies are forced to take advantage of every opportunity tooptimize their business processes. In order to remain competitive, a companyhas to work with its supply network partners to improve the supply network’stotal performance. The purchasing function is an important strategic andoperational issue in organizations. Buyer and supplier relationships inmanufacturing enterprises have received a great deal of attention in theliterature. Also inindustrial companies,the share of purchasing in the totalturnover typically ranges between 50-90% (DeBoer et al., 2001). Therefore, making decisionsabout purchasing strategies and operations primarily determinestheprofit.
Thesupply network aims to provide the required product for its customers efficientlyand effectively. To reach this aim, a company must choose the best suppliers inorder to produce the best finished product. Most of the relevantpapers in the literature select individual suppliers. They assume that the bestsupplier network is composed of the best suppliers of different components.But, this assumption is not valid because (1) the objective for total product/supply network is sometimes notthe sameas for individual suppliers; (2) the effect on finished product/ supply networkfrom suppliers of different components is nottotally independent of each other. The attributes of thesuppliers of each component will affect the attributes of the finished product sophisticatedly.