In traditional warehouse operations supplying retail, we typically buy in pallets and sell in cases. But could it make sense in some circumstances to work with "a lot size of one"?

Most people's initial response to that idea would be that picking one or two pieces instead of a pack of 12 would be suicide in a retail warehouse! If we step back and look at the whole picture, though, we'll see that in the right situation, picking very small lots – even lots of one – could actually be more cost-effective than traditional methods. In fact, if you are being challenged to provide availability of a broad selection of products with the least amount of inventory and the lowest total cost, including cost at the retail outlet level, this approach can help.

Breakthroughs in process flexibility permit companies to manufacture in lot sizes of one. That can bring many benefits, but it still does not satisfy the immediate needs created by purchasers of critical service parts or of impulse purchases from retailers' shelves. Both require inventory control at the point of sale as well as a highly flexible distribution scheme to support them cost effectively.

Much has been done with faster information flow and more responsive transportation services to reduce those field inventories. In spite of these advances, though, for most products we are still tied to purchase-order minimums, price breaks, and packaging quantities that represent weeks or months of demand.

Yet large quantities are not always the best way to go.

When we replenish with a quantity of one the downstream savings can be substantial. The new item can go directly to its place on the shelf. Double handling in the store is eliminated. There is no backroom inventory to manage.

That's great for the retail outlet, but what about the DC? Piece picking may, in fact, be suicide in our warehouses based on our current practices. But rethinking warehouse processes to optimize "each-picking" tasks could yield significant productivity improvements in both the warehouse and the store.

There are several ways to accomplish that. Start by thinking about the final destination of the inventory inside the store. The small orders that are created by increased delivery frequency, for example, can be further divided to match fixtures, aisles or backroom layouts. Several of these "sub-orders" can be picked simultaneously as an efficient batch in a single pick trip at the DC. Loose pieces also can be bagged and labeled with the pick ticket at the DC. As a result, the store or the branch will receive products grouped and identified.

Meanwhile, the pick path for the batch can be greatly reduced if products are slotted at the DC to mimic the way that orders are received from the store. This will challenge us to choose pick-face configurations in the DC that combine efficiency for the fast movers with close proximity for the slower movers.

This may sound old-fashioned. Before we had sophisticated warehouse management systems to optimize DC operations, warehouses and back rooms were often arranged by product family or by vendor, and often in part-num- ber sequence. We can now use these methods to optimize product flow to the store shelf.