Case Study 13—A Simple Approach to Forecasting Sales
In the introduction to this book, we briefly described a forecasting model under the heading of “More Complex Arithmetic Models.” Well, they are that, but when compared to regression models that rely on statistical theory for their efficaciousness, these models offer a simpler approach to forecasting sales. In this case study, we delve deeper into the workings of one of these models and show how it can provide an excellent forecast solution when faced with declining sales during a recession.
Month Length Adjustment
Our subject company is a retail store located in a small city on the Gulf of Mexico. The store alleges that its sales were affected by the BP oil spill that started on April 20, 2010, and that those effects began on May 1 and ran through August 31, 2010. In column C in Table 13.1 we show 52 months of sales from January 2006 through April 2010. We then account for the length of month effect with the following formula found in column F: (365.25/12/number of days in a month). For example, cell F6 contains the following formula: (365.25/12/31) = .981855. Taking this adjustment factor and multiplying it by the actual sales for January 2006 of $34,565 gives us adjusted sales for that month of $33,938. At the end of our forecasting exercise we will have to reconvert forecasted adjusted monthly sales back to their natural state by dividing by each month's adjustment factor. A line chart of the 52 months of adjusted ...