If you move the couch at home and change your mind, that's one thing.

But change your mind about the site you chose for your newly built distribution center, and well, you'll probably have plenty of time to spend on the couch after that.

There are a couple of levels of detail—and criticality—to the subject of locating a new DC. It's one thing to be concerned about which side of the tracks to build a new facility. That's as close to a slam-dunk as things get these days. If you've been charged with deploying facilities in a complex supply chain—or even a global operation—that's a different story. And even when the network appears settled, in today's dynamic business environment, it may not be settled for long.

A number of situations can plunge you into a distribution network design/redesign—an acquisition (or being acquired), a line of business added or dropped, expansion into new geographical markets, a shortage of space, a change in strategy toward centralization (or decentralization), or radical sourcing changes, to name but a few.

If none of those sounds like a reasonable prospect, read no further—you're not in the game. If any of them seems likely, press on.

There are really two components to the question. One is obvious, "where." The other, and first, question is "why." Distribution network structure is fundamental to the ultimate strategic commitment to customer service levels. The strategic view of the network provides the basis for distribution locations, facility missions and inventory deployment.

The devil enters the room
Of course, once the magic word "network" has been invoked, there's a tendency to get overheated about the subject of network modeling. Truth is, once you get beyond a couple of locations, modeling can be enormously helpful for number crunching and quick assessments of alternatives. But be warned: There's an unfortunate tendency at this point for geeks and executives alike to get all caught up in the esoterica of modeling. This is unfortunate, because there's so much more to the question that modeling packages aren't equipped to address. In fact, the modeling tool only begins to answer the critical questions in the overall assessment; other tools and analyses are needed to get at things like strategic approaches to markets or channels, facility size and cost issues, the potential to outsource at least some operations, and questions of whether to automate and to what degree. Modeling does not necessarily provide solid, realistic total cost analysis, implementation planning, or business case development.

In addition, while a number of modeling tools are available for network and facility design, the choice of which to use is far less important than how it's used. The model is merely a tool, ultimately sensitive to the quality of questions posed to it, the reliability of data employed in the solution, and the business context of the exercise. The real keys to successful network modeling lie in asking the right questions, using the right data and aggregating them to the right level, and having enough data and auxiliary tools to evaluate the complete solution.

It's also important to keep in mind that there is often a big difference between an optimal solution and a practical one—and models can't make that distinction. They'll change a solution for a one-dollar cost advantage. Further, it's often true that a majority of savings or benefits come from a sub-set of the modeled solution—and models don't know how to fragment their solutions to find the "bang for the buck" payback. In short, they can't edit or interpret their work, which places a powerful burden on the user.

In addition, models are often tough to build, difficult to verify, and consume data as if an information famine were imminent. Some newer products are somewhat more user- and data-friendly, but modeling is not an exercise for the faint of heart or the resource-constrained. That said, modeling tools are indispensable in solving complex network/facility location questions. Just remember they're only tools, not oracles. They can't answer questions that you can't ask; they can't solve problems that you can't define, and they can't think outside the box.

Selecting the right site for a facility adds another major set of considerations to DC network design. Those include such things as tradeoffs between inventory and transportation. The Von Thunen theory suggests that high-value items, such as gemstones, can be shipped relatively economically from almost anywhere in the world. The inverse is that low-value commodities, such as salt, need to be shipped from quite near the point of consumption. Another element is the potential for postponement, in which it might turn out that the best location for finished-product shipment is not the best location for postponement execution.

House hunting
Once the strategic elements are in place, the process of specific site selection begins. But be sure to have requirements defined before launching the search. Making them up based on what you're seeing is self-delusion of a dangerous kind. As Lewis Carroll reminds us, "If you don't know where you're going, any road will get you there."

The site selection process begins with a Requirements Definition, which prioritizes the factors that are important in a new facility or construction site. Examples include access to specific transportation modes (such as rail sidings or water transport), the labor environment, tax issues, and the ability to expand. Don't forget some other important factors, such as community attitudes—little in business life is more fearsome than the NIMBY lobby. And financing alternatives can radically affect build-or-buy decisions.

Site/facility selection is one activity in which taking all the outside advice you can get is probably a good thing. You can save time and leverage the experience of many advisors—and you can preserve anonymity, a very good thing in a run-up to negotiations. Some sources of help include real estate brokers or developers' consulting divisions, warehouse sales representatives, carrier representatives (especially rail), chambers of commerce, state and local development agencies, and consultants.

When you're ready to make the final pick, be sure to check multiple sources for information and opinions. Aggressively look for indicators of potential trouble, such as floods, seismic activity, soil problems, and access difficulties.

In short, be organized, be creative, and document, document, document.