Retail column: Understanding design criteria

by Edward Stevens

If we were handed a straightforward list of static data, business requirements, product handling characteristics and growth rates, it would be relatively simple to design a cost-effective materials handling equipment, systems and facility model, but retail distribution is far too unpredictable today. Designing a facility for sustainable use is no longer a fixed target. Unless the retailer is highly specialized, a SKU proliferation demands a flexible solution.

When designing for five to 10 years into the future, serious consideration should be given to a phased construction approach which will allow the organization to:

  • Delay spending capital;
  • Fully utilize the new space as phases are constructed;
  • Reduce taxes;
  • Reduce building operating expenses.

Businesses should start building just the space necessary at the moment, even if that only amounts to a third or half of projected total space. Then, at a later date, the remainder of the facility can be constructed just before it is needed. One important lesson learned from experience is not to design and construct a facility for use ten years in the future with very aggressive growth forecasted. This can cause the construction of an oversized facility with operations using only half the facility for the first five years after it is commissioned.

The materials handling equipment design and installation must be a combination of automation (high volume, low labour content) and manual processes (low volume, high labour content). Equally important is the realization or acceptance of the available (inherent) flexibility of each of the choices.

Design criteria

During the process of designing the materials handling equipment of the new DC, one must not lose sight of the business case and customer service objectives established in the program vision. The ultimate question must be asked: “What products will be handled throughout the distribution centre?” Based on our experience, directing this type of question to the wrong level of the organization can result in unproductive conversations.

The responses will vary in detail, depending on the individuals’ areas of expertise, years of experience in the role (or in various supply chain roles), and their personal agendas. Our job is to deconstruct the replies and present the decision-makers with plans based on solid business requirements.

The logical starting point for gathering the needed information is historical data. Current WMS and business systems can provide much of the detailed data. This information will provide a base for developing the projections that will define the materials handling requirements in a new DC.

Creating future projections, however, is another matter. Once again, the importance of the program vision comes into play. Is there a requirement for the new distribution centre to support future volumes, new markets, new customers, or new product offerings? The final design criteria for the project must be consistent with and reflect the vision and objectives of the program.

Before embarking on the design of the materials handling systems, prepare and obtain approval of the design data. This includes detailed numerical data describing SKUs, inventory and throughput. While of great significance, this numerical information is not all that is needed. All assumptions regarding operating conditions—days and hours of operation, customer service objectives, operating cost factors for analysis of options, etc—must be documented.
The most important factor in developing the design criteria is that flexibility is accommodated in the final design. The only sure bet regarding the final approved design criteria is that in some way it will be incorrect in describing the future. Is there way to avoid this, or at least plan for it?

Understanding a product’s weekly profiles and seasonal peaks allows the WMS to recommend product relocation within the facility during the year as seasons change. Fully utilizing more expensive materials handling systems for high activity items may mean that slow moving items are relocated to the appropriate case flow or shelving systems for the slow season. Seasonal reslotting of product can reduce the materials handling equipment needs and system cost for a new facility.

Basic design

At a high level, the distribution centre will support storage and throughput. The materials handling system for a facility designed to store large quantities of a small number of items is vastly different from a facility designed for a large number of very small items. The typical situation will fall somewhere between these two extremes. For retail distribution centres, there will likely be a wide range of movement volumes and, depending on the business type, there may also be a wide range of physical characteristics across the SKU set. These details all impact the final materials handling design.

The most obvious factor in determining storage design is the physical characteristics of the items to be stored. Very large, awkward or heavy items not compatible with pallet storage and handling are typically stored in open floor areas, sometimes with the aid of metal stacking frames. Small or low-volume items that require storage in a carton or less-than-carton quantity are usually stored on solid surfaces such as decked rack or shelving.

The most common situation in retail distribution centres is storage on pallets. The design data must provide the information necessary for selecting and evaluating appropriate alternatives such as height and depth of storage and lift truck type.

Regardless of physical characteristics, the storage design must consider throughput. A product will often be physically received differently from the way it is ships (ie received as a unit load full pallet, but shipped out to a store as a single carton). Here again, the design data will provide the information for determining correct quantities and designs for each situation.

Experience has taught that a rack layout may need one or more drive aisles that are wider to accommodate lift trucks passing each other and that not all the aisles should be the minimum width. In addition to aisles that can accommodate passing, larger than normal aisles may also be required to allow for:

  • Access to high-volume storage areas;
  • Access to double-deep storage areas;
  • Two-way-traffic movement;
  • A pedestrian walkway beside the main traffic aisle;
  • Overhanging pallet loads.

When developing the throughput design we always recommend starting with shipping. Why? This approach will help ensure that your design is compatible with transportation and service level objectives.

Major distinctions in shipping design include shipment volume (trailer load, LTL and parcel), loading characteristics (large awkward items, pallets, and cartons) and delivery system from storage to shipping (manual, lift truck, and conveyor). And in our experience, designing for the outbound volume (characteristic, behaviours) greatly simplifies the design of the rest of the facility.

Just as the storage design must consider throughput, the throughput design must be compatible with the storage system. The forward location replenishment design is the link between storage and throughput. Too often this design aspect is not given sufficient attention. The tendency is to concentrate on order filling without adequately addressing the fact that it is impossible to fill demand from an empty location. Of course order-filling design also requires close attention. The materials handling design must reflect order filling schedules, waves, and the order filling process, whether it is batch, cluster or discrete order filling.


There is something of a renaissance taking place in automation. Any new or retrofit DC project undertaken today should consider the potential for automation, recognizing that it can take many forms—from near complete receiving to shipping application to tactical implementation for a single process.
Some general factors to consider when evaluating automation for a project include:

  • Compatibility of product characteristics;
  • Potential for decreased flexibility;
  • Risk level—tried and true, leading edge or bleeding edge;
  • Vendor alternatives and stability.

The most common justifications for automation are operating-cost reduction and service-level improvement. The reduction in labour that typically accompanies automation can be particularly beneficial in highly competitive labour markets. The biggest trap in evaluating automation is the tendency to get caught up in the wondrous capabilities of modern day materials handling systems. The result may be the implementation of automation for automation’s sake, which could result in poorly applied systems that may not deliver the expected benefits. We have seen many situations where a design is actually dependent on a business model change that never materializes.

Another important lesson is an automated system has a maximum throughput per aisle, per hour and the maximum is really the maximum. An additional process of balancing the activities across the aisles often becomes necessary for maintaining effective system performance. Having too much activity in an aisle may extend building cycle times or force the recycling of the orders to another day.

Evaluation and justification

Often, multiple materials handling designs may seem to be compatible with the program vision. A thorough evaluation of these alternatives will identify the preferred system for implementation. This evaluation must address compatibility of the design with all objectives and fully consider all quantitative and qualitative aspects of each alternative.

The final facility design should consider the flow of product from receiving docks to shipping docks. It must be based on an understanding of the systems’ constraints or limits with the following goals:

  • Reduce the number of times a person touches a product;
  • Be able to pick the largest unit of product that can be ordered—pallet, layer, carton or unit;
  • Use automation or mechanized systems to move product between activities wherever possible;
  • Use the WMS system to track all product movement and balance activities.

Edward Stevens is the pseudonym of a professional who has worked in the Canadian retail supply chain industry for over 30 years, with a strategic focus on the physical distribution of goods and the systems that make up flexible, cost-efficient and effective delivery design.