Developing a dynamic S&OP process for third-party logistics

Editor’s Note: The SCM thesis Developing a Dynamic S&OP Process for Third-Party Logistics was authored by Richard A. Elmquist and Luis Dávila and supervised by Dr. Ilya Jackson ([email protected]) and Dr. Jafar Namdar ([email protected]). For more information on the research, please contact the thesis supervisors.

The preservation of food characteristics and quality within a temperature-controlled environment presents a complex challenge in the food supply chain. Third-party logistics (3PL) companies have a significant opportunity to assist producers, wholesalers, and retailers in managing this complexity. Cold chain warehousing providers face the task of determining appropriate capacity requirements in terms of time and location, which entails allocating capital expenditures to construct the necessary infrastructure. Furthermore, accurate prediction of client needs enables the allocation of operational expenses to manage flexible and cost-effective supply chains.

Our research was initiated by a challenging question posed by the second-largest cold chain 3PL provider worldwide: How can the company balance operational costs and service levels while meeting both present and future demand? This project aimed to establish a dynamic sales and operations planning (S&OP) process by developing a scalable and accurate warehouse inventory forecast. The forecast is utilized as an input in the proposed S&OP process, where subject matter expertise is employed to enhance forecast accuracy through a deep understanding of the business.

To answer the question, we leveraged a comprehensive dataset spanning a duration of four years, encompassing inventory positions for each customer and product within a designated warehousing facility belonging to the company. The significant number of possible combinations between customers and products, coupled with customer churn, posed a challenge in determining the appropriate level of granularity and data grouping. To address this, we developed a segmentation model based on a two-by-two matrix that incorporated average inventory on the y-axis and ease of business on the x-axis. Additionally, we established discrete segments based on temperature ranges for product storage: freezer, cooler, and ambient. This allowed for better forecasting at a level of granularity that was actionable for the company.

After segmenting the data, we generated reliable forecasts using different forecasting models, such as SARIMA and Facebook Prophet, that provided key information for the company. Our forecasts revealed the need for additional freezer capacity in the next six months, as well as underutilized space in the cooler segments that could be repurposed in the next six months. Converting a room from cooler to freezer can add significant benefits to the bottom line of the site, but expert judgment is required before any actual decision can be made.

Finally, we recommended an S&OP framework that enables the company to scale efficiently across its 240-plus facilities globally, including the integration of subject matter experts and the establishment of a feedback loop for forecasts. In the specific warehouse, it was deemed necessary to execute the changes identified by our forecasts. This framework at scale allows 3PL companies to better understand their customers and proactively make changes to their network to ensure reliability and reduce cost of operations.

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.

SC
MR