Grapevine, TX—The most important consideration in designing a fulfillment center is choosing the correct weapon, according to Lawrence Dean Shemesh, president of OPSdesign Consulting. In a Monday morning session at the National Conference for Operations and Fulfillment here, he discussed how to do just that.
Most companies, said Shemesh, take what he called the rifle approach: “multiply old information by questionable information and let it age.” Like a rifle, this tactic can be powerful but inaccurate. Collecting historical data is important, but you shouldn’t rely exclusively on it when trying to determine future needs. There are two many variables and a lack of control over critical criteria—SKU proliferation and changes in order profiles such as a trend toward smaller, more frequent orders—to forecast precisely based on company history. Compound those issues with the lag time between conception of a warehouse design and actual construction and implementation, and it’s little wonder, he said, that so many companies need to retrofit their distribution centers almost as soon as they build them.
Shemesh advised opting for a “shotgun approach,” a solution that is not as sensitive to variables. This tactic enables you to design for flexibility and scalability. Not that the shotgun approach is a license to be overly vague and sloppy, he added. Rather, it enables you to construct multiple options, taking “what ifs” and worst-case scenarios into consideration, so that you can better determine how much to invest to safeguard against unexpected changes.
Like so much else in life and in business, Shemesh said, it comes down to a need for balance, such as balancing storage density vs. accessibility. He cited the example of a nutritional supplements supplier that was trying to gauge how much warehouse space it needed. The difference between its best-case option, which assumed there would be minimal changes regarding SKU proliferation and the like, and the option that would accommodate its worst-case scenario was roughly 30,000 sq. ft.; because the cost of the additional space was relatively low, the supplier chose to go to err on the side of caution and chose the latter option.
Part of this “sensitivity testing” of examining various scenarios involves identifying suspect data. Shemesh provided another example, of a company that had one particular SKU that far outperformed all its others. A big of digging showed that the SKU wasn’t a product at all, but rather the code for UPS shipping.
“Any design initiative, to do it right you need to calculate the product cube per unit,” Shemesh continued. Toss the number of turns, picks, and hits of the SKU into the equation, and you get cubic velocity by SKU, which Shemesh said “is probably the most important metric.” Figuring this out helps determine “where the product wants to live”–in flat shelves way in the back as opposed to floor stacks near the loading docks, for instance.
Another tradeoff in the warehouse, Shemesh said, involves automation vs. flexibility. The two have “an inverse relationship. People are ineherently much more flexible and scalable than automated systems. If you can’t predict the future very well, you need to be prudent in your selection of automation.”
Shemesh advised reviewing your distribution center flow and placement roughly every four weeks. This enables you to be sure that certain formerly slow movers haven’t become popular items that need to be relocated closer to the front, for instance. But while warehouse management systems can provide vital statistics regarding traffic, turns, and the like, they can’t tell you why. A product may have suddenly become a fast mover because it was heavily featured in the most recent catalog, for instance. Such scenarios illustrate why operations and marketing need to remain in close communication.
In fact, Shemesh said you should be sure that the sales, marketing, and planning departments sign off on all projections, such as the types of products that will be sold, that you intend to use in creating a warehouse. “Otherwise they’ll forget about their forecasts,” he warned, “and four years later they’ll wonder why you can’t ship on time.”