Improving customer service level through centralized supply flexibility

Abstract

This thesis explores a combined application of Supply Chain Management theory for centralized and decentralized distribution systems and Customer Relationship Management techniques in data mining to solve the challenges of supply allocations to individual customer entities at Intel Corporation. The relatively long manufacturing lead time compared to order lead time restricts Intel's supply flexibility from responding to rapid order changes by each customer. As a result, the individual customer service level is sacrificed. Therefore, the objective of the research project is to improve customer service level by global optimization for all customers as opposed to the local optimization that each customer currently operates on. Over-ordering occurs in the local optimization system. The proposed solution is to use a scientific and systematic methodology of data mining to guide operational strategies that will provide incentives to encourage customers to order more accurately with Intel. Through this exercise, the supplies can be virtually centralized at Intel instead of being committed out to each customer in a decentralized fashion. The hypothesis is that the new system with more accurate ordering from customers will result in increased supply flexibility and higher effective service level.