Managing by Exception – Part One
By Jon Schreibfeder

We’ll begin by examining unusual sales or usage.  A forecast is a prediction of the quantity of a product that will be sold or used in an upcoming week or month.  Accurate forecasts are critical to achieving effective inventory management. When ever actual sales or usage differs significantly from the forecast used to plan replenishment, we have to determine: 1) why this difference occurred and 2) if it was unusual activity, or the start of a new trend in sales or usage.


  1. At the end of every month or week, you should calculate the forecast error (i.e., the percentage difference between the forecast and actual sales or usage) using the formula:


(Actual Usage – Forecast Demand) ÷ Forecast Demand


  1. Highlight situations where the result is unusually high (e.g., greater than 300% or sales/usage being more than three times the forecast) or unusually low (e.g., less than 50% or sales/usage being less than half the forecast).


  1. Buyers and salespeople should review transactions for each of the identified items. They might find:
  • That a stockout occurred which resulted in lower than anticipated sales/usage
  • A customer temporarily stopped using the product
  • An event (e.g., a promotion, unusual weather, the COVID-19 virus, etc.) caused a change in sales/usage that will have to be monitored for several weeks or months
  • A specific project caused a one-time spike in sales/usage
  • A new trend has begun and sales/usage will probably increase or decrease for the foreseeable future


  1. After determining the reason, the buyer should enter an adjustment to actual usage or sales to reflect what sales would have been under “normal” circumstances. Adjusted usage quantities should be used in place of actual usage when calculating future forecasts.  When deciding whether or not an override to the usage quantity should be entered ask: “what quantity of this item do we want to stock in this location to meet our customers’ future needs?” 


In our next newsletter, we will examine specific examples of adjusting sales/usage to obtain accurate future forecasts.