pixeltrap - Fotolia


Enterprises save with efficient distribution network design

Although a perfect distribution network doesn't exist, you can optimize for current conditions and build in enough flexibility to adapt.

Companies that ship products to customers across a wide geographic area often start by shipping from a single warehouse at or near the factory or company headquarters. As these companies grow, many will add remote warehouses and distribution centers as product volume or delivery distance to customers increases. Often, this distribution network growth is haphazard or opportunistic, and may not result in the most effective or efficient arrangement of resources. Distribution network design aims to make such expansion efficient.

Distribution network design focuses on two factors: proximity to the customer and cost savings.  When it is impossible, impractical or unaffordable to get product to customers within their lead time tolerance from an existing warehouse or when expanding into new territory, adding an additional warehouse closer to the customer location is the obvious answer. It is also common to add a local warehouse to gain an advantage (i.e., faster delivery) over the competition. The second factor in distribution network design is saving money. Although adding warehouses does add costs in terms of space, equipment, and labor, and more inventory is required, savings in shipping and transportation can outweigh those additional costs. Shipping full truckloads of goods to a remote warehouse is relatively inexpensive, and local delivery from the warehouse to the nearby customer is also less expensive, when compared to the cost of shipping a case or a pallet of goods to a customer a thousand miles away. Thorough analysis and modeling reveals that overall costs decline when warehouses are added -- up to a certain point. After the optimum number of warehouses is reached, additions will increase overall costs in a "bathtub curve," much like the well-known economic order quantity graph.

What's important to realize about a bathtub curve is that it tends to be somewhat shallow near that optimum point, meaning that overall cost doesn't change much. For example, if the lowest overall cost can be achieved with 12 warehouses, the cost would not change by much if that number were 11 or 14. In other words, from a cost perspective, getting close to the right number of warehouse locations is the goal. The customer service side of the equation -- responsiveness and lead time -- is much more important.

Taking these factors into consideration, distribution network design and warehouse location can be very complicated. Software models for optimizing distribution network design should include demand information for every customer and location for each product; lead time requirements; all warehouse costs for each potential site, shipping costs between warehouses and their supplying locations, and from the warehouse to its set of customers; shipping tradeoffs (speed versus cost for air, rail, truckload and light truckload); lot size and inventory considerations; and more.

Once you've determined the number and general location of distribution facilities, the exact location for each becomes the focus. You need to consider whether to rent, buy, or build; the cost of land, taxes or tax incentives; transportation convenience (near an on-ramp or where land is cheaper, for example); availability of space and people, and other factors.

It is obvious that distribution network design and management can be a very demanding task. Not only is it complex to set up for optimum performance, but today's conditions won't be tomorrow's. New products, new customers, demand, and shipping rates -- all are subject to change and all will affect the overall efficiency of the distribution network. So a "perfect" network doesn't really exist. You can optimize only for the current environment, and build in enough flexibility to adapt as things change.

The existing distribution network in a supply chain is the result of past decisions and investments that responded to conditions that are no longer the same as they were when those decisions were made. The pessimist will say that you can never get it right. The optimist sees room for improvement. The realist anticipates the need for constant vigilance and adjustment to get best performance, and builds as much flexibility into the supply chain as possible.

Supply chain and distribution network design software and consulting services are available from a number of suppliers. One such supplier describes its service as a fact-based analysis to help you determine facility location and capacity, customer service requirements, material flow, product portfolios and pricing in order to minimize costs and maximize profits. Supply chain optimization (operations management) software is usually listed as Inventory Optimization and offered by suppliers of warehouse management, transportation management, demand management, and sales and operations planning system suppliers.

Next Steps

Closing the gaps in your supply chain

Your guide to sensor technology

How to account for WIP inventory


Dig Deeper on Supply chain planning and execution