It’s a dilemma nearly every distribution center manager faces: To remain competitive, costs have to come down. But a number of factors are pushing costs in the opposite direction.
Fortunately, tools and techniques developed during the past few years can help DC managers make their facilities substantially more efficient and keep them running that way. But implementing requires the willingness to dig deep into operational data—down into the details of SKUs and order lines and even further, into what we’ll call sublines.
The inefficiencies in many DCs are a direct consequence of an inability to analyze and understand the complete data of how inventory actually moves through a facility as the orders are fulfilled. Half of a DC’s labor costs are spent on picking, and 65% of the average picker’s time is spent walking to and from the pick face. Making things worse is the persistent problem of stock-outs and the costly effort it takes to overcome them. Another 25%-40% of DC labor costs are spent on replenishment—a cost kept high as a result of poor design of reserve and forward-pick locations.
Long-term and short-term trends in orders are compounding the problem. The number of one-line orders is on the rise. Customers are asking for more value-added services. These issues push up the unit cost of fulfilling orders—and magnify existing inefficiencies.
Reducing imbedded inefficiencies and cutting per order costs is possible, but only after an intensive analysis of SKU and order-line data to reveal true costs per line or order. Several processes need to be understood in order to gather the data to make intelligent decisions. There are no shortcuts: You need to take all the processes and steps outlined here before the data can help your team plan to reduce costs.
Getting started
The best approach is to have one full year of SKU and order data. It is how these two databases are used that allows for accurately understanding the differences between the storage and throughput needs of the various seasons. It is also necessary to determine fast-moving and medium-moving SKU volatility.
The process starts with profiling—that is, the analysis of orders, lines, sublines, and SKUs based on velocity, activity, and cubic flow. Profiling accomplishes a number of things. First, it provides an analytic understanding of daily processing volumes, including the mix of eaches, full cases, and pallets for each line. It provides the data to compare lines, quantities, and complete orders to SKU velocity.
Understanding that profile is essential to determining facility design options such as the possible number of divert lanes for sortation, opportunities for a dedicated one-line order fulfillment area, and developing a high-level wave-planning approach. In addition, it offers insight to the effects of seasonality business.
Profiling includes both SKU velocity analysis and order velocity analysis. Analysis of the historical and forecasted SKU velocity is the start, albeit an important start. And at the very outset, clarity and agreement over terminology is important. For instance, the term “A mover” has different meanings for different businesses. Better phrases to use are “very fast mover,” “fast mover,” “medium mover,” “slow mover,” “no mover” (a SKU that is on hand with no movement for a year), and “no SKU” (a SKU that is not on hand but is in the database). These categories should be set in everyday terms so that the definitions become real to management. Examples such as “three lines per day on average” or “one line per week per quarter “begin to make sense to the impact to order picking.
True SKU velocity analysis uses a database file containing the total number of SKUs; the activity of each; the item, case, and pallet dimensions; SKU class and description; and many other variables set by the team such as sales dollars and replenishment quantities for each. There are many more variables and use- defined variables in the database that ensure all critical business information is tied back to orders during the analysis.
Moving on
Remember that analysis of the historical and forecasted SKU velocity is a great start. It is an important start. Stopping at the SKU velocity analysis, however, is like having a distribution center without a roof. The other critical link in the data analysis is the order velocity analysis.
While SKU velocity analysis helps define equipment allocation and slotting, order velocity analysis defines the order picking requirements; the storage media; and the effect of orders on the building footprint, replenishment trips, and labor requirements for picking and replenishment.
Order velocity analysis requires a database file containing the actual discrete order information, the SKU IDs, order quantity, the dates the orders were received, the shipping dates, the wave identities, the carriers, the order dollar values, and other user defined variables set by the team.
This analysis is critical. It provides the data for understanding the effect of seasonality; analyzing one-line orders; and determining a host of other crucial operational details. Those can include the need for separate forward eaches and forward full-cases areas, throughput from picking zones, allocating orders to pick module or reserve picking areas, and the adequacy of forward-pick locations. Determining the throughput from pick zones also helps plan rates for conveyor systems and aisles in order to eliminate potential bottlenecks in merges, sortation, and print-and-apply devices. Determining the allocation of orders to pick module or reserve picking areas prevents making forward-pick areas bigger than necessary and allows for planning to pull solid pallets from reserve to eliminate multiple touches.
The productivity gains that can be derived from order data are dramatic. Without it, a facility’s design could have costly contingencies such as extra storage media, which can lead to a facility that is as much as 35% larger than needed. That in turn causes longer travel times in picking.
The order velocity analysis provides data needed to choose storage media based on cube or productivity considerations, to understand the effect of replenishment on labor costs and unexpected stock outs in forward-pick locations, to judge the financial benefits and costs of investing in higher-productivity picking media, and to understand seasonality and its impact on the building and operations.
Drilling deeper
The next step—and a vital one to truly understand operational costs—requires drilling deeper into orders.
We all know that customers’ orders consist of lines and that a line is a single SKU in various quantities. In reality, order lines consist of smaller groups that must be picked at a defined pick face. These smaller units are called order sublines. That is a breakdown of a customer’s order into the number of eaches, full cases, and/or full pallets of each SKU, plus any possible combinations of these. It is really these order sublines that define the order picking requirements, the storage media requirements, the impact of orders on operations, the building footprint, the replenishment plan, and the labor needed for key functional areas.
This is where the cost of picking and replenishment is located. Looking at the picking activities to this level makes sure full pallets are optimally picked from reserve, full cases are optimally picked as cases, and eaches are picked as eaches or inner packs. This subline data also helps in designing the pick methods.
The subline analysis can also provide a better understanding of the effects of seasonality on operations. By analyzing the data to this subline level, we can distinguish between the movement and storage needs of basic and seasonal items, thus producing a smaller, more efficient operation. By understanding in which seasons a defined SKU is really a fast mover and when it is a medium mover gives you the ability to slot and pick this SKU in a fast-moving method during this prime season. During the slower season, this SKU can perhaps be put away in slower-style media as it is received again into the building. The complete understanding of the differences between peak-season release times and nonpeak release times is important to lowering cost.
Using the analysis
Once the detailed data analysis is complete, the next step is to select the correct storage media in parallel to designing the picking procedure and process. This process, equipment allocation and selection, can benefit from a good distribution software design tool, which will allow “what if” analyses based on businesses variables and operational scenarios.
After you’ve selected the picking processes and the storage media, you can begin to develop the design on paper. Before the drawing is a finished and material-handling hardware is integrated, however, the same data can be used to balance the building. Zone balancing and slotting is the assignment of a specific SKU in a defined quantity (eaches, cases, pallet quantities) to a specific storage location in a type of storage media to balance picking activities.
Such detailed analysis can seem overwhelming. Indeed, this level of analysis could not have taken place as recently as two years ago. But with today’s updates in software and information systems the job is easier.
This process of a data-driven design results in smaller, less expensive, and simpler distribution center with less automation for the distribution center. It also allows for financial evaluations of designs that can be modeled with various storage media. It will ensure a detailed understanding of the order fulfillment requirements with comprehensive order reports, complete with seasonality down to the subline level. It gives the information needed to develop balanced pick zones and pick faces. It provides a unique view of the effect of storage media and orders on the building, the picking, the replenishment, and staffing. Once the layout is locked and the system is changed or installed, the DC is easier to start up and easier to operate. These results are a reduction of overall operating costs and an increase to the bottom line.
David Farmer is vice president of sales for Nashville, TN-based distribution services and software provider Fortna.