Managing Inventory by Exception – Part #2

by | Sep 15, 2020 | Analysis, Best Practices, Forecasting, Purchasing, Replenishment Parameters, Stocking Decisions

Managing Inventory by Exception – Part #2

Interpreting Unusual Usage

In the last newsletter, we discussed the importance of evaluating possible unusual usage in accurately forecasting future demand of products. In the next several newsletters, we will look at some specific examples of unusual usage, properly evaluating them and adjusting actual usage to reflect what would have been sold or used under “normal circumstances”.

 

We will start by examining three instances of what might have caused an unusual “spike” in usage and some ideas of how to correct your usage history so that it reflects what might have been sold or used under “normal” circumstances.

 

An Unusually Large Sale.  In August, you forecast that you would sell 250 pieces of a product and actual usage was 800 pieces.  How do you investigate what caused the unexpected increase in sales:

  1. Examine the individual transactions that generated the sales of 800 pieces. Did one customer buy an unusually large quantity?  Perhaps the typical quantity sold in one transaction is somewhere between 10 and 25 pieces and one customer bought a total of 525 (in one or several transactions) throughout the month.
  2. Have the salesperson responsible for the account contact the customer to find out if the quantity purchased was for a one time need or if this quantity represents what the customer will continue to buy in the foreseeable future. Salespeople soon learn that this is a great opportunity to find out what is going on at customer sites and may discover other products you can supply to fill the customer’s changing needs.

 

The Effect of Events.  Did an event cause the unusual spike in usage?  This type of event affects usage, but does not occur at exactly the same time every year.  That is, they do not have predictable seasonal usage.  Events include:

 

  1. Holidays that do not occur on the same day every year – Easter is an example of a holiday that is also a short-term event.
  2. Sales and other programs designed to increase the sales of products.
  3. Atypical weather or a disaster. For example, does snow or rain significantly increase the sales of some products.
  4. Pandemics such as the COVID-19 virus.

 

When an event is identified you must adjust usage history to remove the effects of the event.  This is done with the equation:

 

Actual Usage ÷ (1 + Effect of the Event)

 

For example, we have determined that unusually high usage of 800 pieces was the result of a promotion that increased sales by 25%.  We will adjust usage history removing the 25% temporary increase in sales by dividing the actual usage of 800 pieces by 125% to determine the adjusted usage quantity of 640 pieces (800 ÷ 1.25 = 640). 

We would perform the same calculation if an event caused a temporary decrease in sales.  For example, a particularly rainy week caused demand for certain outdoor products to decrease by 50%.  Dividing the actual usage quantity of 130 pieces by 1 minus 50%  ( 130 ÷ 0.50 = 260)  determines the adjusted usage quantity of 260 pieces.  Keep in mind that adjusted usage should always be used in place of actual usage in all forecasts of future demand.

 

Please note that it is important to record the historic effect on usage of events.  For example, when we run a particular promotion, sales increase an average of 25%.  Or, when it rains all week sales of specific products decrease by 50%.  We can use this information to adjust future demand forecasts when those events occur again.

 

The COVID-19 virus is an example of this type of unusual usage.  But as we have seen, its effect on usage of specific products change over time.  You might have experienced a tremendous spike, or decrease in sales of an item last March, followed by a gradual return to “normal” usage.  It is important to closely monitor sales each month and adjust the percentage adjustment as necessary. 

 

Adjusting usage for spikes in activity is an important task in ensuring that the historical data you use to forecast future demand of products is as accurate as possible.  It has repeatedly been proven that accurate historical data leads to accurate forecasts.  And accurate forecasts are a key component in achieving the goal of effective inventory management.