Many computer systems utilize past usage history when forecasting future demand of products. Seasonal forecast formulas in these systems typically look at what was sold or used in this month (and possibly subsequent months) last year. For example, if you are forecasting sales of a seasonal item in April, 2014, you might utilize usage recorded in April and possibly May, 2013 in your forecast formula. The result may be adjusted by a trend percentage that compares usage in the previous several months this year to the same months last year. For example, if sales in January through March, 2014, are 12% greater than usage recorded in January through March, 2013, the results of the calculation using past usage history would be increased by this percentage.
However there are several problems with this approach to forecasting seasonal items. These challenges include:
- Many firms deal with seasonal that don’t start on a specific date. For example, the popular season may start at the first cold snap, heat wave or after the first frost. If the season started later than normal last year, there probably won’t be sufficient stock in inventory when the season starts.
- Seasonal products often have extended lead times. If it takes eight weeks to replenish the stock of an item and a season typically starts on October 1st, replenishment orders with the supplier must be placed no later than the first week in July.
- Customers are often “desperate” when they realize that they need these products. In colder climates hardware stores often run out of rock salt hours after an ice storm is predicted.
To deal with this situation we suggest you adopt the following forecasting and replenishment procedures:
- Define season start and end dates for all items whose popular season start date cannot be accurately predicted. This is a “best guess” by your management and salespeople and reflects the date that you want to be prepared for high sales activity triggered by something that causes customers to demand these products (e.g., heating and cooling parts, rock salt, snow tires, etc.)
- Calculate forecasts and safety stock quantities for each of the months in the popular season. Remember that safety stock is reserve inventory you maintain to avoid stock outs due to unusual demand or delays in receiving replenishment shipments.
- Scan the calculated safety stock quantities for all months in the season. Replace the safety stock quantity for the first month of the season with the largest calculated safety stock quantity for the season. This will help ensure that you have adequate stock on the date that you want to be prepared for the typical high demand that often occurs when an unpredictable environmental event causes customers to beat down your doors to get the products they desperately need.
Of course if a season has a late start in a particular year you will be overstocked for a couple of months while you work off the large “season-start” safety stock quantity. But isn’t ensuring that you have adequate stock when your customers are desperate worth this risk?