Site Selection – Part 2 - A Site for Sore Eyes

When you get right down to it, all logistics (like all politics) is local. HP maintains 88 distribution hubs throughout the world. IBM Corp. has at least one major logistics site on every continent in the world except for Africa, and 28 in all. The Gillette Co. has four distribution centers in the United States and 60 total worldwide. In the United States alone, retail behemoth Wal-Mart has 128 distribution centers strategically located in 38 states.

And yet, there's a feeling that the site selection process is more art than science, more luck than strategy. Determining exactly where in the United States a company should locate its logistics and distribution centers requires a study of many factors beyond just transportation costs (although transportation is a major factor in the decision).

For many years, Expansion Management, a magazine that specializes in site selection, has teamed up with Logistics Today to produce the Site Selector—a tool that offers an objective ranking of the 362 major U.S. cities (i.e., metropolitan statistical areas, as defined by the U.S. Office of Management and Budget).

FINDING THE RIGHT PLACE

The Site Selector was designed to help companies find the right city or region for their distribution needs. Because virtually every company uses motor carriers at some point in its distribution network, access to good roads is an important factor, but it's not the only factor. The city of Trenton, New Jersey, for instance, was ranked at number one (the highest rating) for road infrastructure in the 2005 study, which is not too surprising given the city's proximity to major highways and turnpikes. Trenton is also strategically sandwiched between two major metropolises—Philadelphia, Pennsylvania, and New York, New York.

However, the condition of its roads is not very good at all. Trenton's rank in that category was number 355 out of 362. Taking a look at some of the other categories, we find that Trenton placed well for taxes and fees (number 30), and fair-to-middling for rail access (number 151) and transportation and distribution industry (number 154). All things considered, Trenton finished nationally at number 68, which puts it just inside the top 20 percent.

But because most site selection decisions focus on a region of the United States rather than the entire country, it's also helpful to identify how well a city does compared to other cities within the same region. Trenton, for instance, ranks as the 15th most logistics-friendly city within the U.S. Northeast. The condition of its roads is far less of a factor for companies with supply chains in the Northeast because, frankly, none of the roads in that part of the country are in very good shape, relatively speaking. The one-two punch of congestion and Mother Nature accounts for the perpetual epidemic of orange cones on highways in the Northeast. As a result, road condition is almost a nonfactor for companies making site selection decisions centered on the Northeast.

Chicago Consulting undertook a study to determine the best warehouse networks in the United States, with best indicating the lowest possible transit lead times to customers, based on population patterns. Using that criterion, the best place for a company managing one distribution center would be Bloomington, Indiana. The average distance to a customer would be 803 miles, with an average transit time of 2.28 days. For a company operating two DCs, the optimum locations would be Ashland, Kentucky, and Palmdale, California.

When The Gillette Co., a manufacturer of personal care products, batteries, and other consumer packaged goods, launched its North American Network Study in 2002, the goal was straightforward: Identify the best distribution network that would allow the company to deliver excellent customer service at the least cost. As solutions manager for the company, Louise Knabe's job was to figure out how many DCs Gillette should have and where they should be. Least cost was an important consideration, Knabe points out, because if Gillette's goal had been simply to provide the best possible customer service, the network study could well have suggested putting a distribution center in every state.

"From a logistics and distribution perspective, Gillette measures customer service by order cycle time (time from when the customer places the order until they receive the order) and on-time delivery performance (percentage ol shipments that arrive on time)," Knabe explains. "The strategic DC network design affects the order cycle time because the location of the DCs affects the transit time to the customer."

At the time of the network study, Gillette had two DCs located on the East Coast, one near Boston, Massachusetts, and the other near Chattanooga, Tennessee. The Tennessee warehouse stocked only Duracell batteries, while the Massachusetts warehouse stocked everything else. Neither warehouse carried all of Gillette's products.

So why was this a problem? "Our project analysis revealed that this situation made it difficult to deliver top-quality customer service," Knabe points out. "Let's say I was a customer based in Virginia. That meant I was getting a shipment from Tennessee of batteries and shipments from Massachusetts of everything else. So I've got two trucks showing up with Gillette products on it, which was a bit of a nuisance."

The bigger issue for Gillette, though, was that because neither warehouse had all of the company's products, many customer shipments had to be delivered by less-than-truckload (LTL) carriers, a more expensive transportation mode than truckload. The transit times were longer and the reliability was lower than it would have been if Gillette had been able to get all products loaded onto the same truck. Gillette concluded that in order to deliver top-quality service, it needed to find a way to convert as many of those LTL shipments into full truckload as possible.

That's when Gillette got to work on its site selection best practices, with the goal of developing a network that would locate the DCs close to the customer and make it possible to regularly ship by truckload.

COST VERSUS SERVICE

To answer the questions of how many warehouses it needed and where they should be, Gillette conducted a complete theoretical analysis to identify the best locations. The company factored in such considerations as the location of its manufacturing plants and its sourcing points. Equally important, Gillette looked at where its customers were located, and specifically at who ordered what, and in what volume. "You take those two things and then ask: How do I marry them up and how do I figure out where my warehouses should be?" Knabe says.

"In terms of distribution cost, we looked at the freight cost of going from the plant to the warehouse, and then we also looked at the freight cost of going from the warehouse to the customer," she explains. Using an optimization software tool to evaluate every possible scenario, Gillette asked questions such as: If we had three warehouses, where would they be to minimize our freight costs? The company looked at other distribution costs, including real estate, labor, and taxes, and utility costs, such as electricity ("That ruled out Manhattan pretty fast," Knabe notes). Inventory carrying costs were also factored into the plan.

On the service side, the question Gillette asked was: How can we impact customer service when we're designing our distribution network? According to Knabe, there were two ways. The first way was to set up the distribution network so that Gillette could maximize its use of truckload, which meant stocking all products in all warehouses.

The second part of the answer involves order cycle time. "The location of our warehouses affects transit time to the customer," Knabe says, "so we looked at how many warehouses we needed if we had to be able to get to every customer within 48 hours. And then we asked: How many warehouses would we need if we only had to get to 85 percent of our customers within 48 hours? We looked at our network from both of those angles—cost and service—and figured out what made the most sense."