Measuring The Supply Chain

"You can't manage what you can't measure." The old adage rings true. So as companies focus on more actively managing their entire end-to-end supply chain processes, they are also looking for better ways to measure end-to-end supply chain performance.

"The whole idea of supply chain management is that instead of looking at different functions or processes in isolation, companies need to look more holistically and globally at the entire linkage of companies and processes that take products from raw material to the end consumer," observes John Neale, director of research for Optiant Inc., Burlington, MA. What companies need therefore, he says, are "multidimensional measures," that encompass multiple elements at once and are cross-functional, even cross-enterprise.

In practical terms, companies are seeking performance indicators that provide both "bigger picture" views, and at the same time, more granularity in the data reported. They need clearer pictures of both the forest and the trees.

Given the accelerating pace of movement and competition in the supply chain today, companies also want measuring systems that can help them more speedily translate data into action.

John Stelzer, director of industry development at Sterling Commerce, Dublin, OH, notes that retailers have traditionally viewed their world as a collection of three separate spheres: supplier-facing activities, internal operations and a customer-facing side.

"They have tended to implement solutions or programs that deal with each of these worlds separately," he points out.

But if these individual "islands of automation" can't interoperate in a single seamless process, Stelzer says, companies pay one of two penalties, or both: losses in operational efficiency and detriments to the customer experience.

Performance measures which "provide visibility between all three worlds and islands of automation," he points out, make it less likely managers will miss those "failure points" that occur on the border or in the cracks between the three worlds.

Companies also do better when they can take a more granular approach to performance measurement, Robert Byrne, president, Terra Technology, Norwalk, CT, comments. "Today they can look at performance in more detailed ways because there is a lot more daily data available, thanks to larger faster computers," he says.

Much of the work on enhancing supply chain measures centers today on inventory. This makes sense, given how much money is typically tied up there, and the fact that it is the key element around which all supply chain processes revolve.

One focus of Optiant's approach is to help companies better model uncertainty up and down the supply chain, and monitor how uncertainty manifests itself in inventory, Neale notes.

"We want to see how uncertainty propagates and changes as product moves along the supply chain. As a rule you want to hold more inventory where uncertainty is lower. As demand gets pooled, uncertainty diminishes. Typically you're trading off the ability to hold inventory higher up in the supply network with the responsiveness of the supply chain."

To help companies manage the trade-off better, Optiant provides tools to measure inventory performance by cost, service, and asset use. A key difference, he adds, is that historically, each functional area has focused on just the cost components under its control.

"They didn't look at the impacts of their cost minimization tactics up and downstream," Neale explains.

Some of the key measurements that go into its analyses include very detailed, mathematical calculations of forecast error, "because there is a difference between variability and uncertainty," he adds.

"If you can predict variability, that is how things change over time, then you don't need as much inventory.

"We also provide reports that look at inventory by type and by cause."

Drilling Down

Optiant allows companies to drill down, for example, from high-level reports showing days of supply across a company's entire business to look at the different types of inventory that make up that aggregate, such as inventories of raw materials, work in process, finished goods, and product in transit.

In one case, for example, its inventory optimization calculations showed a manufacturer that by increasing its raw material inventories, it could greatly decrease inventories of finished goods, and thus substantially decrease total inventory supply cost—even though costs in one functional area, raw material supply, went up.

Similarly, Terra concentrates on helping companies analyze and drive change through detailed analysis of the "levers" that affect inventory.

"If you want to drive step changes in inventory, then you have to drive step changes through the levers," comments Michael Mastroianni, vice president of North American planning, reliability, and operations support, Campbell Soup Co., which has worked with Terra.

"As simple as that sounds, people don't always get it. But the old days of the CFO 'waving a wand' and saying, 'I want inventory taken down by $15 million,' are over. Because simply taking inventories down to meet the goal could be detrimental to service, and often to total cost as well."

Using Terra's Real-Time Inventory solution, Campbell Soup has "done a lot of work to decompose the components of inventory," says Mastroianni. "Then we go back to the drivers of these components to understand what levers we should focus on. To drive down safety stocks, for example, we might concentrate on driving forecast accuracy up and volatility down.

"We've also shifted toward focusing more on measuring inventory performance in terms of 'days of cover,' instead of just by dollars," he notes.

"With some of the reporting tools from Terra, we've been able to go very precisely to a certain manufacturing line in a particular plant in a particular month and understand that if we influence cycle stock on this line, what effect that will have on a particular lever."

The company also might look at stock in-transit from plant to DC, and analyze its key drivers.

"One driver might be the fact that we're using over-the-road vs. intermodal transit, and the transit time added as a result might be driving inventory up. That becomes an issue of cost trade-off. We have to see whether at the end of the day, the savings achieved by going to a different transportation mode is really net, net the right decision, when you also factor in such elements as inventory holding costs and product obsolescence."

Another concept Terra helps companies examine is inventory "health."

"A company might say 'we want five weeks of supply', and might be on target, but still not really be aligned to where it should be, because one warehouse might have 26 weeks' supply of an item while another has none," Byrne points out.

Action Now!

Companies are also looking for supply chain measures that are more forward -ooking and immediately actionable.

"Instead of waiting until the end of the month to see how the train wrecked, and how to clean it up, people want to look down the tracks and see where wrecks may occur before they happen, so they can be prevented," says Jane Lee, director of supply chain solutions for Wilmington, DE-based Supply Chain Consultants. SCC recently introduced Zemeter Performance Management to help companies do that.

"For example, if we look at inventory and it's too high today, we'll look at forecast-to-demand and see how that's measuring up, and we'll look at the schedule of shipments into the warehouse, to see where inventory is projected to be at the end of the month based on today's numbers.

"We've always had various metrics to serve as early warning systems, but now we actually have an expert system behind the scene that says if this is high, or this is low, chances are it is due to any of several specific causal relationships. Companies can then examine each of the potential causes to see if that's where the problem lies, and take corrective action.

"The better you get at this kind of early warning management, the less likely you are to have problems, because you can trade off time, capacity and inventory against one another, as long as you can plan far enough ahead," she points out.

Types of measures that don't help, Lee adds, are anywhere when the trend goes the wrong way and it doesn't really affect how a company goes about its business.

"If it's not going to help you change anything, you may as well not measure it," she observes.

Loading