Rationalizing Production Scheduling at Swift & Company

The Problem

Swift & Company, a diversified cattle- and pork-processing business with $8 billion in annual sales, operates in a low-margin commodity industry with complex production management requirements. The perishable nature of the product, constraints imposed by Swift’s large scale of operations, and variability in the quality and quantity yield of each animal being processed place a high premium on the ability to match customer orders with the fast-moving production process. Tight margins require that every part of each animal is sold, but varying demand for different animal parts adds to the challenge of pricing them optimally. Swift’s order management systems were not up to the task of juggling all these variables. As a result, the company was operating on a “production-push” basis, failing to meet customer demand for certain products, and at the same time was forced to under-price other products to sell them before they spoiled.

The Analytics Solution

Swift secured an analytics-based production and inventory management system for its beef- packing operations that would facilitate a change from the “production-push” to a “demand-pull” business model. The system was designed to:

• Give customer sales reps real-time product availability knowledge;
• Provide accurate control of inventories;
• Optimize pricing of less demanded residual products; and
• Enable Swift to adapt raw material usage to satisfy changing demand.

The nalytics-based system built to satisfy these requirements rested on three mathematical optimization models: One for production scheduling, a second for “capable-to-promise” (CTP) capacity determination functionality, and a third that generated a real-time snapshot of available unsold inventory based on the current production schedule. Each model would be updated regularly with the most current inventory, demand and production data. Then each model would be optimized to calculate unsold production from each shift at each of five meatpacking plants, updating sales teams at 15-minute intervals on any undesirable inventory buildups.

The system also generates a shift-level schedule for each processing plant and factors in side constraints to production, including the physical difficulty and time consuming nature of certain productions, as well as business rules that prevent the accumulation of unwanted inventory.

The Value

Purely financial benefits of the new system included a $12.74 million (or 13%) annual bottom-line improvement, primarily from the optimization of the product mix. An estimated $560,000 of that bottom-line benefit is attributed to the avoidance of excess price discounting triggered by the buildup of undesired product inventory. Additional benefits included an 8% increase in on-time shipment performance, the freeing up of customer sales representatives to perform more value-added activities instead of addressing the adverse consequences of inventory management problems, and improved customer perception of Swift & Company.