Segment customer demand to "rightsize" inventory

The goal of any supply chain is to match supply with demand, and in so doing make a profit. Traditionally, manufacturers have tried to accomplish this by building up inventories to handle spikes in demand and smooth out production cycles. But that's no longer a viable strategy, and even before the economic downturn forced manufacturers to cut inventory to free up working capital, many had been trying to "rightsize" the amount of stock in their warehouses. Finding the right balance between product mix and inventory has always been tricky, however. If the items customers want are not in stock, then companies risk losing sales.

One way to rightsize inventory is to segment demand. Ken Koenemann of TBM Consulting Group Inc. in Durham, N.C., suggests this method: divide products into four categories based on sales volume and demand variability. Products with a high sales volume and low demand variability would be assigned to one quadrant. Products with a high sales volume and high variability would be placed in another quadrant; low volume and low variability in another; and low volume and high variability in the final sector.

The next step is for companies to take a good look at which quadrants their products occupy and equate them to customer types. If a customer reliably places a steady stream of large orders, then it probably is a profitable customer, and it makes good business sense to maintain a sufficient quantity of those products in stock. On the other hand, if a customer orders only small amounts of product and doesn't do it very often, then the additional costs for short manufacturing runs and holding inventory may make it unprofitable to do business with that customer.

Once companies have segmented products and customers in this way, they can then devise a sound supply chain strategy that better matches service levels and inventory to demand. In this regard, Koenemann suggests examining individual customers for their revenue contributions, days of inventory required to serve them, and the costs of holding that inventory to meet their order requests. With that data in hand, companies can then determine whether the volume and frequency of sales justify the associated supply chain costs.

All of this leads up to one thing: correlating service levels in the supply chain with the cost to serve various classes of customers. It makes sense to do what it takes to meet demand from the steadiest and most profitable customers. But not every customer is profitable, and in some cases, segmenting may show that companies should take drastic action—ceasing business with customers that don't order enough product frequently enough to cover supply chain costs.