Some items, like beach umbrellas, are more popular in summer than in winter. On the other hand, portable heaters enjoy much higher sales when the weather gets cold. These are seasonal items. But the weather is not the only factor that determines whether or not an item is seasonal. If a product’s usage is controlled by an event (such as Christmas or the start of school) or an annual activity (like yard clean up in the fall), the item is also considered to be seasonal. The usage of a seasonal product rises and falls throughout the year. Look at this seasonal item’s usage history:
| Jun 1999 |
May 1999 |
Apr 1999 |
Mar 1999 |
Feb 1999 |
Jan 1999 |
|
| Usage 1999 |
? | 300 | 150 | 80 | 50 | 30 |
| Dec 1998 |
Nov 1998 |
Oct 1998 |
Sep 1998 |
Aug 1998 |
Jul 1998 |
Jun 1998 |
|
| Usage 1998 |
50 | 100 | 150 | 300 | 520 | 460 | 400 |
Usage of the product is very low during the winter months. But in early spring, sales begin a gradual increase and peak during the summer months of June, July and August. If we forecast demand for June 1999 by using the formula for non-seasonal products with consistent usage (described in Part One), we get the following result:
| Month | Total Usage | Number of Business Days in Month |
Usage per Business Day |
| May | 300 | 19 | 15.8 |
| April | 150 | 18 | 8.3 |
| March | 80 | 22 | 3.6 |
| February | 50 | 20 | 2.5 |
| January | 30 | 22 | 1.4 |
| Month | Weight | Usage per Business Day |
Extension |
| May | 3.0 | 15.8 | 47.4 |
| April | 2.5 | 8.3 | 20.8 |
| March | 2.0 | 3.6 | 7.2 |
| February | 1.5 | 2.5 | 3.8 |
| January | 1.0 | 1.4 | 1.4 |
| Total | 10.0 | 80.6 |
The extension (80.6) is divided by the total weight (10.0) to determine our prediction of the demand per business day for June of 8.06 pieces. Because June has 20 business days, demand for the inventory period is 161.2 pieces (20 days x 8.06 pieces per day).
Remember that demand is defined as a prediction of the usage of a product during the upcoming inventory period. Is 161 pieces a good forecast of June’s usage? Probably not. After all, usage in June 1998 was nearly three times this amount (460 pieces). It is obvious that we need different formulas for calculating the demand for seasonal items.
We’ve found that one of the best indicators of what demand will be for a seasonal item next month is the usage recorded during the upcoming several months, last year. For example, one formula for forecasting demand for seasonal items considers the usage for the upcoming month and the following month last year, applying the following weights:
- Place weight of 2.0 on the usage recorded in the month being forecast, last year.
- Place weight of 1.0 on the usage recorded in the month following the month being forecast, last year.
| Month | Total Usage | Number of Business Days in Month |
Usage per Business Day |
| June 1998 |
400 | 19 | 21.1 |
| July 1998 |
460 | 18 | 25.6 |
| Month | Weight | Usage per Business Day |
Extension |
| June 1998 |
2.0 | 21.1 | 42.2 |
| July 1998 |
1.0 | 25.6 | 25.6 |
| Total | 3.0 | 67.8 |
The extension (67.8) is divided by the total weight (3.0) to determine our prediction of the demand per business day for June of 22.6 pieces. Because June, 1999 has 20 business days, demand for the inventory period is 453 pieces (20 days x 22.6 pieces per day).
But there is a problem with forecasting demand with history that is a year old. Business in the branch where the item is located, or in its particular line of products, may have increased or decreased during the past 12 months. For this reason, a “trend factor” can be applied to the results of the weighted average formula to reflect overall changes in your volume of business.
Many systems will allow you to manually maintain trend factors. Say, for example, you determine that the sales volume in our item’s product line increased 20% over the past year. To determine the actual demand forecast for the product, we’d increase the result of the seasonal weighted average formula by 20% to determine the actual demand forecast for June, 1999:
More advanced systems calculate a suggested trend factor by comparing the total usage in the last three completed months (before the forecast demand calculation) to the total usage in the same three months in the previous year:
Total Usage March 1998 – May, 1998 = 462 pieces
(530 – 462) ÷ 462 = 14.7%
Business in the past three months was 14.7% greater than the same period last year. This percentage is added to the results of the weighted average formula:
Whether specified manually or calculated automatically by the system, trend factors must be applied whenever seasonal forecast formulas are utilized to compensate for the change in business experienced over the past 12 months.
Next month, we’ll continue our examination of various methods of forecasting the future demand of products. Please check back with us as we continue this analysis. Remember, accurate forecasts will substantially contribute to the profitability of your company!