When we founded Effective Inventory Management (EIM) in 1996, our goal was to help organizations throughout the world achieve the goal of effective inventory management – that is, to meet or exceed customers’ expectations of product availability while maximizing the organization’s net profits or minimizing its inventory related costs.

One of our primary areas of concentration has been the improvement of future predictions of product sales or usage, otherwise known as “demand forecasts.” In previous newsletters, we have discussed how to develop demand forecasts and replenishment plans for:

  • Items with sporadic usage
  • Items with recurring usage
  • New inventory items
  • Products maintained specifically for one customer

But as we have discussed before, the accuracy of any demand forecast is dependent on the “goodness” of your sales or usage history. Does this data truly reflect what would have been sold or used under “normal” circumstances?

A lot of software packages identify possible unusual usage at the end of each week, month, or other inventory period. The simplest of these systems compares usage in the month just completed to the total usage recorded over the previous “x” months, or it informs you if you were out of stock for more than “y” days. More advanced systems bring to the attention of a buyer significant differences between the demand forecast and actual sales or usage recorded during the inventory period just completed. A common trait of both methods is a belief that usage or sales history has been correctly recorded.

But this is not always a safe assumption. Here are some things to consider when determining how your system should update usage history:

Record Usage for the Product Your Customer Wants. Usage should normally be recorded for the product ordered or requested, not the product actually shipped. If you are out of the product your customer orders, you’ve probably disappointed the customer. If you provide another item in its place and record usage for the substituted product instead of the product originally requested, you are setting your company up to disappoint the customer again.

Record Usage When a Customer Wanted the Product. Be sure your system records usage based on the promise date (i.e., when the customer wanted the product) not when you eventually delivered it.

Record Usage in the Proper Location. Imagine a situation where one of your good customers usually obtains material from branch “A” of your company. But today branch “A” has a stockout of a very popular item and you have to supply the customer’s needs from branch “B.” Usage for this sale should be recorded in branch “A” because this is the branch you want to resupply to meet the customer’s future needs. If you post usage in the shipping location it is unlikely that branch “A” will ever have enough stock of the product to meet this premier customer’s needs. To ensure that usage is recorded in the proper location, you can make a note of a customer’s “home” branch in their customer record. This warehouse will post usage for all shipments to the customer regardless of the location that actually supplies the material.

Ensuring that usage is properly recorded may seem like an unattainable goal. But it is not. First, understand how your computer system deals with each of these situations. Then meet with your software support team to determine how your system can be modified or enhanced to ensure that usage is properly recorded.

There is an old expression, “What we shipped in the past is a good indication of what we will sell in the future.” As we have seen in this article, this attitude can easily result in many stocking problems. In order to help develop the most accurate forecast possible, replace the “old” expression with a commitment to be sure that usage is recorded for the product, in the time period, and in the location to best meet your customers’ ongoing needs.