Modeling the Relationship Between EDT Implementation and Firm Performance Improvement With Neural Networks.

This paper examines a number of electronic data interchange (EDI) usage and implementation factors and their role in improving a firm's efficiency, productivity and competitiveness. Unlike other studies in the literature that use exclusively linear models, we apply nonlinear neural networks to model the relationship between performance improvement and a set of predictor variables of EDI usage and supply chain coordination activities. A variable selection method is employed to identify key factors to predict a firm's operational excellence due to EDI implementation. In addition, a bootstrap resampling scheme is used to evaluate the robustness of the results. Note to Practitioners-Competing successfully in a highly competitive environment requires firms to coordinate well with their suppliers and customers. !n this paper, we study the issue of how EDI implementation activities and usage affect a firm's performance. Our research has several important implications to practitioners. First, although a firm's performance improvement in several key areas may be attributed to many different aspects of EDI use and coordination activities, only a few key variables are critical. By focusing on a few key coordination activities and EDI implementation issues, managers are able to achieve significant performance improvement in efficiency, productivity, and competitiveness. Second, we find that coordination activities through the implementation of EDI can clearly improve many facets of a firm's performance, developing strategic alliances with supply chain partners is essential in enhancing both a firm's and its supply chain's competitiveness. Third, while using the basic routine EDI can improve efficiency, it is the enhanced use of the EDI system that contributes more to the productivity improvement and firm competitiveness. [ABSTRACT FROM AUTHOR]