Fabric maker learns forecasting, operational metrics

When senior management at Propex Fibers decided the company's supply chain needed an overhaul, the entire supply chain organization went back to school. More than 35 employees attended a course in supply chain management at the Logistics Institute, Georgia Tech University.

The education proved invaluable, claims Frank Slaughter, director of production planning, scheduling, and inventory management for the Austell, Ga.-based manufacturer of non-woven fabrics. "It meant we all spoke the same language," he says.

At supply chain management school, a number of things quickly became clear. "We always assumed that because we were metric-driven, we were focused on performance," Slaughter recalls. "But we found out that most of our metrics only related to financial performance."

Propex also found things that had been overlooked, such as the structure of its supply chain organization. Multiple departments were blended into a new organization—comprised of customer service, production planning, procurement, warehousing, and transportation—that fell under the direction of a single VP of supply chain (see chart).

This organization then developed strategies to consistently achieve the best possible score on what has become Propex's most important operational metric—delivering the perfect order (see box).

"We came to understand that demand management was the head of the beast, if you will," says Rush Clarke, Propex's VP of supply chain.

Further study revealed accurate forecasting was necessary for proper demand management, another area in which Propex needed to improve. "It seemed like everyone in the business had their own forecasts," Clarke recalls.

That problem ultimately was resolved through the use of Zemeter forecasting and inventory management software from Supply Chain Consultants(SCC). According to Sujit Singh, VP of solutions for SCC, Propex's implementation of Zemeter modules was easier due to the analysis and education that Propex undertook before purchasing the software.

"The fact that they were an educated client was tremendously helpful," Singh says. "They all spoke the same language, and understood what it was all about."

Still, both Clarke and Singh admit not everyone at Propex readily accepted the forecasts generated by the SCC solution.

Salespeople were particularly reluctant at the outset, because they believed direct contact with customers gave them an edge in developing accurate forecasts. Instead of debating that point, Clarke and Slaughter decided to let the salespeople continue submitting their own forecasts while also posting the system-generated forecasts.

Over time, the salespeople realized the computer-generated forecasts were more accurate, and required less work. Propex now relies on Zemeter to generate all of its forecasts, although sales managers still have the authority to override forecasts for specific items.

"There's a single set of numbers for the whole business," Clarke says. "It doesn't matter where you are in the organization—there's just one forecast." Those forecasts also are more accurate. "We're about 50 percent more accurate than the forecasts produced by the account managers on their own," Clarke says.