When you have past data to call on, use it. Always compare your forecast to past results. Look to the past as a reality check. Understand what's changing and why, and what may remain the same. While mathematical and statistical analyses are great, a simple forecast-to-past comparison is quick, practical, and very powerful. Every professional analyst knows to look first for real past results before projecting into the future.
The simple illustrated example below shows furniture unit sales for 1998 and 1999, as actual past results.
With that data as actual real past data, consider the two hypothetical next-year forecasts below. Each of them shows projected future unit sales compared to monthly unit sales from the recent past. Ask yourself which is a better forecast.
This first forecast for unit sales in Year 2000 seems unrealistic. Why are sales so high early in the year, when they haven't been like that in the past? Why would sales go down toward the end of the year? You would want to ask the forecaster what causes these radical changes.
This second Year 2000 forecast, on the other hand, seems immediately more logical. Notice how closely it follows the previous years' results. This is an obvious application of common sense in forecasting.
