Abstract
Efficiency decomposition is being able to know the most effective way to improve a company’s organisational efficiency. This study proposes a network performance evaluation model with non-controllable variables to evaluate dealers with an emphasis on their internal processes, including demand forecasting, sales force and inventory control. The proposed model decomposes the organisational efficiency into a weighted average of internal process efficiencies, where the weights associated with the internal processes are data-driven. This is applied to an empirical case consisting of 27 automobile parts dealers in Taiwan. Based on the analysis results, dealers know the relative contribution of each internal process to the organisational efficiency, and inefficient dealers are assigned targets for enhancing the organisational efficiency as well as the process efficiencies.
Acknowledgements
We are grateful to Dr. William W.H. Lien who is a Senior Special Assistant in the case company and generously offers the data-set to accomplish this research. We also thank the anonymous reviewers for their constructive suggestions, which improved this paper significantly.
Notes
1. Compared with retailers, in terms of the organisational function, dealers have to ‘buy the goods’ and ‘sell the goods in a large area’. Additionally, the consumers to dealers are usually businesses rather than end-users.
2. In practice, the achievement rate (the sales revenue divided by the sales target) majorly depends on consumer needs, the number of competitors and the business cycle. Because the achievement rate is driven by a wide range of factors, Company XYZ always sets reasonably achievable sales targets for each dealer. Thus, from the business point of view, the case company believes that the achievement rate should be considered as a controllable variable in the process of sales force.
3. As requested by Company XYZ, this study cannot disclose any information about the raw data in the case study. Therefore, the slack values of the inputs and outputs are transformed into the percentage terms.