An Inventory Management Solution for Complex Portfolios of SKUs

Every year, 40 or so students in the MIT Center for Transportation & Logistics' (MIT CTL) Master of Supply Chain Management (SCM) program complete one-year thesis research projects. The students are early-career business professionals from multiple countries, with 2 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 the real-world problems. In this series, we summarize a selection of the latest SCM research.

The SCM research project Joint Replenishment and Base Stock Model for the US Beer Industry is authored by Nate Moison and supervised by Dr. Josué C Velázquez Martinez ([email protected]), Executive Director, Master of Supply Chain Management, and Dr. Sergio Alex Caballero ([email protected]), Research Scientist, MIT CTL. For more information on the research, please contact the project supervisors.

A major challenge for supply chain practitioners in many consumer-packaged goods companies is how to manage inventory for an ever-growing portfolio of stock-keeping units (SKUs). Add a mandatory third tier to the distribution network, and the challenge becomes even more complicated.

Moreover, as consumer preferences continually change SKU counts increase, yet the revenue implications of eliminating products can make enterprises reluctant to cut the number of SKUs.

In such situations, supply chain managers have little choice but to manage the company's bloated portfolio of SKUs. Solving this inventory manage problem supports competitiveness and keeps internal and external stakeholders happy.

Brewers in the United States face this type of problem daily. By law, the industry operates under a three-tier distribution network, and to maintain high service levels must manage inventories across these tiers. MIT CTL researchers worked with a leading US beer company to develop an inventory management solution. The company had taken a strategic decision to open distribution across the United States to handle the complexity of its network.

Brewing a Successful Inventory Policy

The inventory management problem can be broken down into two parts. First, the organization needs to classify or segment the SKUs. Second, it is necessary to develop an inventory policy for each segment. There are many different approaches to determining the classifications, such as Pareto Analysis (the “80/20” rule), that are straightforward to calculate and implement. However, the shortcoming of these approaches is that they are one-dimensional. These methods focus typically on either a cost or volume-indexed classification but fail to link all important characteristics such as production frequency or demand variability to more accurately classify the SKUs.

In our research, we found it advantageous to link SKU classification to the frequency in which SKUs are produced. This method is especially beneficial when an organization is not capable of producing the entire portfolio every week. The classification helps group the individual SKUs into production weeks.

In developing the inventory policy – especially when service quality is a key driver –using the Base Stock model is a best practice. This approach allows the enterprise to balance inventory cost considerations with customer service levels for their inventory policy. To link the SKU classification with the Base Stock model, we recommend using the production frequency as a proxy for lead time. This enables the company to account for the interval between production runs and maintain sufficient inventory to cover the interval.

Bringing It Together

Based on our research, we recommend the following three-step solution to solve the inventory policy problem for complex SKU portfolios.

• First, categorize your portfolio using a comprehensive and holistic approach, such as Joint Replenishment.
• Map your classifications to production intervals. The Joint Replenishment approach is an easy way to solve both the production interval problem and SKU classification.
• Finally, link the production intervals with the base stock model to determine the best way to deploy your inventory.

While this approach will not guarantee an optimal solution, we found that the solution from this heuristic is near-optimal and improves upon the brewer's current strategy. Importantly, the approach reduced the inventory levels of over 80% of the SKUs studied without impacting customer service levels.

SC
MR