Enhancing Supply Chain Operating Models Through Segmentation
Name
SCM05_AlArfaj_Arslan_project.pdf
Size
659.49 KB
Format
Adobe PDF
Checksum (MD5)
fde1ce391826436e69eeb42407f87c11
Author(s)
AlArfaj, Ibrahim
Arslan, Yalcin
Date Issued
June 8, 2022
Abstract
Large Food and Beverage retail chains often manage diverse sets of products and markets, where one- size-fits-all supply chain operating models are insufficient to meet their distinct requirements. In collaboration with a global retailer, the main objective of this study is to identify distinguishable supply chain segments based on product and market characteristics and design an alternative supply chain operating model (SCOM) for each segment. To achieve this, a five-step, integrated, data-driven methodology is designed. First, data is gathered and reviewed for accuracy and completeness. Second, data is analyzed to identify potential segmentation criteria and select the most relevant factors. Third, k- means clustering is applied to create the product segments. Fourth, a SCOM is designed for each segment based on the product characteristics. Finally, the SCOMs are simulated to analyze how they perform in different scenarios. Applying the methodology resulted in three segments differentiated based on the products’ demand volume, demand volatility, shelf-life, item cost, and seasonality. The three segments are slow-moving, fast-moving, and complex items. Each segment was recommended to be managed using different inventory and forecasting policies. Using simulation and scenario analysis, several service level targets were tested to show how they impact inventory costs, transportation costs, and fill rate. As a result, the SCOM for each operating model brings benefits to the overall performance. Focusing on this, slow-moving products are not delivered frequently, hence eliminating their inventories in the DCs is expected to reduce the inventory holding cost without significantly increasing the transportation cost and decreasing service levels. Disaggregating the inventory in the CDCs for fast-moving items is expected to improve service levels for these items, with a low increase in inventory costs. Lastly, aggregating the demand for complex items is expected to reduce the risks of stockout and excess inventory. The methodology in this study can be generalized to other industries with high product variety to enable them to reduce inventory, improve service level, and reduce the total distance traveled.
Subjects
Supply Chain Strategy
Machine Learning
Network Design
Terms of Use
CC0 1.0 Universal
Persistent DSpace Link
