Predictive modeling of fulfillment supply chain for delivery performance improvement

Thesis: M.B.A., Massachusetts Institute of Technology, Sloan School of Management, in conjunction with the Leaders for Global Operations Program at MIT, 2017.S.M. in Engineering Systems Massachusetts Institute of Technology, School of Engineering, Institute for Data, Systems, and Society, in conjunction with the Leaders for Global Operations Program at MIT 2017Cataloged from PDF version of thesis.Includes bibliographical references (pages 51-52).NIKE, Inc. Direct-To-Consumer (DTC) represents sales from NIKE-owned retail stores and internet websites. DTC is experiencing tremendous growth and is projected to account for 32% of NIKE's revenue by 2020. DTC Digital (Nike.com) alone is estimated to grow at more than 46% year over year. To support such rapid growth, exceptional delivery performance is critical. As consumers become more demanding about delivery precision and as competitors offer increasingly faster options, NIKE is also focusing on initiatives around delivery service to achieve its targets and remain the industry leader. The internship aims to generate business recommendations to improve the consumer-focused delivery experience. The trade-offs between supply chain (cost) and delivery service (speed, reliability) are explored. Additionally, an analytical framework in the form of a simulation model of the digital delivery supply chain is developed to evaluate the quantitative impact of each recommendation. The following approach is used: 1. Understand current delivery issues and business requirements by learning about the delivery process, system capabilities, available data, and relevant stakeholders 2. Conduct research on consumer experience and how it compares to the industry benchmark 3. Connect with geo representatives and global stakeholders to provide learnings on short-term metric alignment (ensure common "language" in delivery service measurement) 4. Formulate recommendation hypotheses for root cause issues with the largest impact 5. Build a proof-of-concept simulation model that predicts impact on internal operational performance (distribution center processing capacity utilization) as well as the consumer-facing metric (Estimated Delivery Date, or EDD) 6. Scale the simulation model to include dynamic perspective and reflect seasonality, weekdays, time of the day, and unexpected high demand The simulation model, which emulates the North America digital business, shows significant improvement (4-10%) in delivery reliability (EDD) for each of the following recommendations: " Shape demand to level daily order volume " Proactively add two days to period with high daily demand " Prioritize orders of critical consumer segments in the distribution center The internship provides a findings summary and fact-based point-of-view to support leadership on short-term improvement decisions, and additionally produces an analytical framework using predictive modeling methodology to investigate further future long-term strategy around delivery.by Lizhong (Lilly) Cao.M.B.A.S.M. in Engineering System