If ever there was an industry that had to balance inventory right, it was contract manufacturing. While most manufacturing companies face challenges in trying to maintain ideal inventory levels to meet ever-changing demand levels (given that forecasts can never be 100% accurate), the challenge is even greater for contract manufacturers. These supply chain "middlemen" accept business from numerous customers and then must interact quickly with their suppliers to get the right materials to the right locations in the right quantities.

Contract electronics manufacturer Plexus of Neenah, Wis. has leveraged an innovative inventory management system called integrated inventory optimization to help it proactively respond to changes in demands. "We are responsible for procuring the components to support the manufacturing of our customers' products," says Scott Brown, manager of supply chain analysis and design at Plexus. "Obviously, this plays a significant part of our total investment from an asset standpoint." In addition, Plexus doesn't really have a lot of direct visibility to the demand stream coming to it, because it is filtered through a number of different processes and hands. "It is often not terribly accurate in terms of its representation of future requirements," he says.

Using a traditional MRP approach has been less than satisfactory in an environment where the forecast is not very accurate, and where customers are expecting the company to be flexible by still responding to demand that was never forecast, as well as to deal with liability in cases where the forecast was overly optimistic. "As a result, we ended up being in a 'place and chase' mentality, with a heavy reliance on expediting and de-expediting, and answering a lot of MRP messages," Brown tells Purchasing. ("Place and chase" involves placing orders and then chasing them with expediting and de-expediting activities.)

In studying the situation, Brown realized that since Plexus didn't have a good handle on future demand it needed to take a different approach. He suggested a strategy that involved being more demand-driven and more flexible, responding to actual demand signals, instead of pushing inventory in and having too much of the wrong material and not enough of the right material.

"The strategy allows us to optimize our investment in inventory based on analysis of past demand, as well as input from the future forecast," he says. This allows Plexus to create a flexible model based on a demand-pull type of approach, a fairly dramatic shift from the traditional MRP push methodology.

Rolling the concept out to Plexus' supply chain organization was critical to its success. "The key to success was education—getting people to understand the difference between the MRP push approach and the demand pull-based approach," Brown says. It involved getting people to realize that they needed to move out of the "place and chase" mentality, and become more proactive about having inventory in place that supports flexibility, and only replenishing it when there is actual consumption. "This was a huge change for people, especially those whose whole training and experience had been in an MRP environment," he says. "We needed to get them comfortable with the idea of strategically placing inventory in the right quantities in the right place to provide flexibility, instead of being reactive when things don't go according to the original forecast."

According to Brown, the Holy Grail of inventory optimization is being able to maintain all of the data points and have a dynamic system for updating the inventory levels, reorder points, and reorder quantities. Plexus has actually developed tools internally to monitor demand variability, supply variability, and to recalculate and update its system with the correct parameters.

Brown says there are many supply chain organizations discussing concepts like demand-pull and vendor-managed inventory (VMI), but he believes most of them aren't doing it effectively, leaving Plexus with few examples to follow.

"When we started talking with suppliers, for example, they were able to begin using modified push models, which were still being driven largely by the forecast," he recalls. For example, they would say they offered VMI, but they meant that they would keep two weeks on hand, and if Plexus didn't use it by the end of the month, they still had to take it. "That's actually still a push model," he points out.

As a result, Plexus began working with suppliers to show them the proper way to do a demand-pull analysis. Still, though, the systems aren't fully in place in the supplier organizations, so Plexus continues to offer support to the suppliers with some of its own systems.

"Currently, we are about 30% of the way through the transition to this model," Brown says.

Recently, Plexus has become familiar with some commercial packages that focus on demand-pull inventory optimization, but hasn't made any decision about adopting any. "We like what some of the packages do. It's just a question of whether the incremental improvements that they offer are worth the investment to us."

As a result of its demand-pull inventory optimization effort, Plexus has been enjoying one of the highest levels of return on capital employed (ROCE) in the industry over the last year, at more than 30%. "This has given us a step up in this industry, and the concept is proving to be extremely popular with customers and suppliers alike," Brown states. "In fact, customers in particular see it as a differentiator."