Redesign of the demand and supply match process in Nike SCM

Nike is currently working in an overall strategy that aims to transform the marketplace in order to increase the customer experience and become the chosen brand when the customer decides to buy. This strategy implies changes in different areas and processes of the supply chain. One of these processes is called Demand & Supply Match (DSM), which main objective is to assure the optimal use of Nike inventory taking into account performance parameters such as time delivery, order coverage and service level. This report presents the outcome of the Logistics Design Project carried out for Nike Inc. This project has as a main goal the redesign of the DSM process by incorporating concepts such as market information and contracts segregation in order to improve demand & supply matching performance and provide more inventory flexibility. Three DSM models are introduced as alternatives to the current DSM process. The first one is a first needed first served (FNFS) model that matches orders based on their needed date in the market. The second one is a mixed model, which combines historical and actual market information to create forecasted product-based consumption pools that are updated each three weeks based on the market behavior. The last model combines the two previous models to create a logic that allows products that are covered in full on time be matched by the FNFS model. Meanwhile, products that are not covered in full on time are matched using the mixed model. In this project, a DSM decision framework based on workflow techniques is also introduced as a guidance tool so that the user can decide based on the current demand and supply conditions which model suits best to match demand with supply. Moreover, a set of rules to identify Load-in and Fill-in orders is given and implemented. This identification is quite critical not only for this project but also for the coming Nike projects because it is expected that in the future orders will be managed in terms of Load-In and Fill-in orders. The results obtained in this project lead us to a number of conclusions: 1. It has been shown that the mixed model improves if sufficient market information (sell-through) is available. In fact, it is recommended the use of the mixed model when sufficient market information is available and the percentage of contract orders is a significant portion of the dataset that will be matched. 2. The FNFS model provides the best coverage in full on time when two conditions are presented: It is applied at the start of the season, when little demand information is available and when the supply is larger than demand. 3. Better coverage in full on time is obtained when orders are selected and matched based on multiple CCD month pools compared to the case where orders are selected and matched in only one season pool. 4. A hybrid approach can improve the DSM performance. This approach means to start the season with the base model. After the first month or until sufficient market information (sell-through information) is available, and based on the current coverage, a decision is made. If high coverage is presented, then the based model or the FNFS model can be used. However, if low coverage is presented, then the mixed model is used. Despite that the current results indicate that the base model is still outperforming the proposed DSM models, there are strong motives, presented in this project, to believe that better performance in terms of coverage and DSI can be obtained if sufficient and reliable market information is rapidly captured and translated to the DSM models.