At least once in your life, you will be charged with setting up a sizable direct-to-customer fulfillment operation in a short time. If you’re in the dot-com business, this project can take on a scale and urgency approaching epic proportions, seeming no less frightening than Jason’s quest for the golden fleece
Trying to fulfill thousands of direct-to-customer orders – many with intensive value-added services – in a single day is daunting. What’s worse is that there isn’t always an Argonaut around when you need one. Often the typical catalog or Internet start-up doesn’t have a seasoned executive team that can quickly establish a robust operations capability. Usually, one person with minimal experience is selected to develop and direct all operations for the first 18 months of business. I have counseled many a novice who was simultaneously thrilled and terrified about having to create order fulfillment, customer service, and returns capabilities from scratch.
Success in this venture, as in most, is far more certain when the journey begins properly. The first step is to identify the capital investment and resources you need and the operational costs to be incurred while fulfilling orders within the anticipated business requirements. When faced with a start-up, several vital design and planning questions should be addressed: How much volume do you expect? What devices and systems will best handle this load? What is the launch date? To answer these questions, and to set up a distribution operation quickly and effectively, you must pay attention to three core components of planning: business and order volume forecasting, material handling system design, and project scheduling.
Turn up the volume In a start-up situation, it is not uncommon for the operations planner to be challenged with estimating order fulfillment volumes based on scant business information. Merchandise buyers usually have an idea of what is likely to sell, how many items they expect to carry, and how much of each item should move within the product offering. But the truly critical statistics, although less obvious, are line and order fill data. This information is of primary importance and can make the difference between a stable business and one brought to its knees by being based on statistics that are too generalized. The following are six sets of numbers that you must obtain before you begin designing a facility:
1. Orders filled, estimated per peak week
2. Lines filled, estimated per peak week
Estimation technique: Aside from dividing the total annual orders or lines by 52 weeks, you should assign a peak-to-average ratio appropriate for the merchandise in question. These ratios help determine the peak volume week that one can anticipate in a given operating year. Typical peak-to-average ratios in the DTC business range from 2.0 to 8.0 when annual seasonality is taken into account. Consider the example below:
Average lines per week = 52,000 Peak-to-average ratio for business = 4.5 10,000 average lines per week x 4.5 = 45,000 lines filled per peak week
Of course there is no magic to the formula above – anyone can punch it out on a calculator. The difficulty lies in determining the suitable peak-to-average ratio for the business and merchandise involved in the distribution. Don’t sweat it if you can’t run the calculation to the 0.001% level of precision, because your call on peak to average will be accurate only to the plus or minus 15% level at best. Even an established catalog operation cannot predict its next year’s peak to average with complete accuracy. For some common peak-to-average ratios, see Table 1.
3.Orders/lines estimated to fill per peak day
Estimation technique: There are several ways to estimate how a peak week can translate into a specific peak day order volume, but I prefer to use a simple rule of dividing the peak week volume by three days. This equates to allocating orders/lines as if orders arrived with a typical “weekend effect” – this being the tendency of customers to order within a tight three-day period after catalog release, or the effect of a weekend on unit sales. In industries such as pharmaceuticals, Monday can represent 60% of the total order base for the week. In such cases, specific consideration should be given to the circumstances of the business.
4. Lines filled per average order
Estimation technique: Aside from the total workload anticipated, this is the most important statistic to address before beginning a design effort, and is certainly easy enough to calculate. Take the total orders anticipated for the peak month or week, and divide by the total lines to be shipped in that same time frame. The purpose of the data is to aid in the selection of the right order filling technology. Classic bin-shelf and carton flow-rack environments work best in businesses with high SKUs/low lines per order, such as books and videotapes. On the other hand, enterprises with high lines per order, such as pharmaceuticals or fasteners, may achieve a payback by reducing the labor content of order filling by applying an automated order-filling technology. Knowing how to match the right technology to the right order profile is critical for healthy operations.
Even some of the largest dot-com and customer-direct operations in the country have failed to take this important statistic into consideration while planning order-filling techniques. This is best exemplified by a dot-com retailer who installs a tilt-tray sorter to fill orders when supporting an SKU-intensive business with about 1.5 lines per order. In such an instance, a great preponderance of orders are single-line orders; that is, one trip is made to a unique SKU slot to pull the batch to feed into the sortation device. If that pull from that single slot represents only one or two orders, then, in essence, the pick of that order has already been completed when the batch is pulled. Efficiencies are gained only when a pull from stock represents a large number of orders. If the pull represents only one order, then the automated device into which the pull batch is fed becomes nothing more than a very expensive conveyor to the packing station.
5. Percentage of peak week orders that will require value-added services
Value-added services include any modification that the business requires to the item before it is packed for shipping. This includes gift wrapping, engraving or other forms of personalization, insertion of informational material, assembly, and so forth.
6. Productivity targets
Estimation technique: Just as in the estimation of peak-to-average ratios, workstation productivity should be assessed taking into account the specific nature of the business and customer demands. Productivity rates are best gauged by measuring the work content of an average activity as opposed to outside benchmarking.
Outfit yourself The elements of material handling system design are highly dependent on the lines filled per average order, as discussed above. Designs involving automated order-filling technology require the assistance of the supplier of the technology. However, let us assume that you are thrust into a start-up that will not undertake within its first two years enough volume to generate a payback on high-end technology. In this circumstance, and while there is great risk as to the business volume to be achieved, it is wise to take a flexible, effective, and yet conservative approach to facility design. The photographs on page 40 depict some of the more common elements contained in a design supporting a conventional order-filling operation.
Photos A and B represent storage devices common to the basic order-filling business, bin shelving and carton flow rack respectively. In general, for any given SKU, carton flow rack should be installed rather than bin shelving if the number of stock replenishments to a bin exceeds two to three per week.
Photo C represents a classic order-filling conveyor spine, one of whose applications is to sort orders to each successive zone in which a pick is required. If an average order fills out 100% in a single zone, consideration should be given to not having any zone sortation at all; however, let us assume that the typical order in your new facility will require three lines, picked from three separate SKU zones. Photo D depicts a very common and inexpensive routing device called a pop-up wheel diverter. This mechanism, along with the scanning or logical flagging that determines if an order is to be routed, is the primary tool used in achieving a zone sort. There are myriad versions of these devices available on the market, differing mostly by the wheel pattern required to achieve the most effective divert for the carton or tote that the operation employs.
Photo E represents a classic order-packing and value-added services workstation. At these locations, employees perform a number of activities that improve the overall quality, presentation, and customer perception of the merchandise ordered, such as gift wrap and monogramming. I typically recommend that value-added work and ship packing be conducted at separate workstations. The single-activity-oriented workstations keep efficiency and productivity high and workers focused, becoming a better solution financially despite the extra expense required to route products between the stations. Photo F depicts a possible value-added and ship-packing workstation spine. The value-added workstations appear first in the process, to be followed by a conveyor and routing device to sort packages or totes to the ship-packing workstations further down the line.
The cost of such equipment varies, but a good benchmark for a fairly low-tech, 100,000-sq.-ft. start-up fulfillment operation is $8 to $16 per square foot. You will probably not have the time to approach the market, develop a specification, and bid the project to several material handling system suppliers. In this event, you may want to request that an installer conduct an audit-based open book project, in which the supplier will be asked to allow the customer to examine the invoices, time sheets, mark-ups, and other cost components of the project. Table 3 illustrates average material handling costs for a facility similar to that shown in the photographs.
Many times, on a bid-based project, a contracting firm will choose to forgo a portion of its SG&A or even its profit margin, so you may be able to obtain a lower overhead cost component depending on how motivated the supplier is to have your business. However, it is usually wise to allow contractors to make their profit and SG&A on a project. You’ll find it far more cost-effective to be prudent in technology selection and design planning than to be penny-wise and pound-foolish with your material handling supplier. It is all too common to hear of start-up operations in which the customer threw away millions of dollars on inappropriate high-end order-filling technology, yet also asked the supplying contractors to cut their profit or SG&A. A better strategy is to make a portion of the profit contingent on on-time completion of the project or some other positive partnership incentive.
Take time out Project schedules can vary greatly depending on how complex the design is and whether a building has to be constructed. As a general minimum, however, if a facility has already been identified, you can expect a duration of six months of elapsed time from the first steps of business statistical planning and design to the actual start-up of the material handling equipment. If the project is much larger, expect a span of 14 months or more. If the facility has not been constructed, the time will extend to nine months at the very least, and more reasonably 22 months for larger operations.
Start-up planning is never easy, and requires the foresight and flexibility that are the hallmarks of prudent operations executives. Bringing to bear the right individuals and/or consulting resources to enact this process is critical to the fledgling operation’s effectiveness. Using the proper critical factors to plan your operation and employing the suitable level of facility design will allow you to conduct business profitably, supported by happy suppliers, bosses, and above all, customers.