Multi-echelon inventory strategies for a retail replenishment business model

Abstract

The mission of the Always Available retail replenishment business at NIKE is to ensure consumer-essential products are in-stock at retailers at all times. To achieve this goal, NIKE has developed a forecast-driven, make-to-stock supply chain model which allows retailers to place weekly orders to an on-hand inventory position in a distribution center. The challenge facing the business is how to design an inventory strategy that achieves a high level of service to its customers while minimizing inventory holding cost. Specifically, safety stock holding cost is targeted as it accounts for the majority of on-hand inventory and can be reduced without significantly impacting the underlying supply chain architecture. This thesis outlines the application of multi-echelon inventory optimization in a retail replenishment business model. This technique is used to determine where and how much safety stock should be staged throughout the supply chain in order to minimize safety stock holding cost for a fixed service level. Provided a static supply chain network, the ideal safety stock locations and quantities which result in minimal total safety stock holding cost is determined. For this business, the optimal solution is to stage lower-cost component materials with long supplier lead times and high commonality across multiple finished goods at the manufacturer in addition to finished goods at the distribution centers. Safety stock holding cost reduction from component staging increases significantly when the distance between manufacturers and the distribution center decreases and for those factories producing a variety of finished goods made from the same component materials due to inventory pooling. Forecast accuracy drives the quantity of safety stock in the network. The removal of low volume, highly unpredictable products from the portfolio yields significant inventory holding cost savings without a detrimental impact to revenue. By deploying the optimal safety stock staging solution and by removing unpredictable products, this analysis shows that finish goods safety stock inventory would be reduced by 35% for the modeling period (calendar year 2012) while only decreasing topline revenue by 5%.