By Ibrahim AlArfaj and Yalcin Arslan
Editor’s Note: The SCM thesis Enhancing Supply Chain Operating Models Through Segmentation was authored by Ibrahim AlArfaj and Yalcin Arslan and supervised by Dr. Özden Tozanlı. For more information on the research, please contact MIT SCM Program Executive Director Maria Jesus Saenz.
Inherently complex and uncertain supply chains—combined with constantly changing customer needs—create many challenges for companies, including supply and demand volatility, stock-keeping unit proliferation, and expanding distribution channels. Given these challenges, one-size-fits-all supply chain operating models cannot meet the distinct requirements of products, customers, and distribution channels. Companies now seek ways to customize their operating models to cost-effectively meet demand at the right time, at the right place, and in the right quantity.
Segmentation is a valuable tool for designing supply chain strategies that meet the unique characteristics and needs of products, customers, and distribution channels. Our study focuses on product segmentation in the food and beverage industry. We clustered products based on varying characteristics and developed supply chain strategies for each segment to improve cost, service level, and sustainability metrics.
Handling the complexity of food and beverage
Food and beverage retailers have a variety of products with different characteristics such as perishability, diverse sources and varying demand patterns. Given this complexity, manually segmenting products requires selecting a few key variables, which can be relatively subjective and incomplete.
With this in mind, we designed a data-driven methodology that integrates data analysis, machine learning and simulation to develop a robust segmentation strategy. Using data from a global food and beverage retailer, we applied our methodology to more than 450 products.
These products are characterized by demand volume, demand volatility, lead time, cost, seasonality, shelf life, and storage temperature. Through this approach, we identified distinct segments; developed customized inventory and forecasting strategies; and evaluated cost, sustainability and service level trade-offs of each strategy.
The slow, fast, and complex
Our analysis resulted in three product segments, each having unique inventory and forecasting strategies.
The first segment is the slow-moving items segment consisting of low-volume, medium-volatility, and non-perishable items. Slow-moving items stay in the distribution centers (DCs) for extended periods due to low demand. Therefore, we recommend just-in-time inventory where the materials are sent directly from suppliers to stores based on each store’s immediate needs. Given their infrequent deliveries, eliminating slow-moving items’ inventories in the DCs reduce inventory costs without impacting total transportation costs or service levels.
The second segment is the fast-moving items segment including high-volume, low-volatility, and highly perishable items. These items should be stored in facilities closest to the stores. Also, forecasts should be disaggregated for these storage facilities to accurately meet the demand of each area. As fast-moving items have high inventory turnover rates, these strategies improve the service level with a small increase in inventory costs.
The third segment is the complex items segment consisting of highly volatile and seasonal items. These items should be pooled in large DCs to reduce inventory risks. Seasonal items can be moved closer to customers in their respective seasons. Lastly, forecasting can be improved by aggregating the demand forecasts at larger storage facilities and for a longer period. These demand and inventory aggregation strategies reduce inventory risks and decrease costs.
Applying the product segmentation strategy can positively impact the entire supply chain.
First, the number of stock-keeping units in each DC will be reduced, enabling more streamlined warehouse operations.
Second, additional supply chain strategies such as building micro-fulfillment centers for fast-moving products and signing Vendor Managed Inventory contracts for slow-moving items can further improve cost and service level metrics.
Finally, modifying the segmentation strategy in different markets facilitates product introductions to new markets and enables focused growth. These strategies can reduce inventory, improve service levels, and decrease distance traveled to enable long-term improvements in supply chain efficiency and responsiveness.
Every year, approximately 80 students in the MIT Center for Transportation & Logistics’s (MIT CTL) Master of Supply Chain Management (SCM) program complete approximately 45 one-year research projects.
These students are early-career business professionals from multiple countries, with two to 10 years of experience in the industry. Most of the research projects are chosen, sponsored by, and carried out in collaboration with multinational corporations. Joint teams that include MIT SCM students and MIT CTL faculty work on real-world problems. In this series, they summarize a selection of the latest SCM research.