IN THIS CHAPTER
Knowing when you need to pay attention to order — and when you can ignore it
Recognizing the importance of time periods
Making your time periods equal
In most cases, you’ll get the most out of your historical baseline of sales data if you put it in chronological order. And because you’re going to forecast into the future, the order should be ascending chronological order. This chapter shows you the easiest way to arrange that order for your baseline.
Your forecast will be for a particular period of time. To some extent, your baseline’s time periods determine the length of time your forecast will cover. For example, if your baseline’s time periods are years, forecasting revenues for the next month is tough. On the other hand, if you need to forecast a year, you may have to jiggle your baseline some. In either case, the length of each time period in your baseline is important.
When you forecast, you’re trying to separate the signal (the regular, dependable component of your baseline) from the noise (the irregularities that come from unpredictable events, like sales reps being out sick, random changes in your customers’ buying patterns, and so on). To help with this separation, you want to impose some order on the chaos. One of the ways you do this is to use equally spaced time periods of nearly equal length.
One of the characteristics of a useful baseline is that ...