Table of Contents### Making Forecasts

Entire Site

Knowing the overall trend exhibited by a data set is useful because it tells you the broad direction that sales or costs or employee acquisitions is going, and it gives you a good idea of how related the dependent variable is on the independent variable. But a trend is also useful for making forecasts in which you extend the trendline into the future (what will sales be in the first quarter of next year?) or calculate the trend value given some new independent value (if we spend $25,000 on advertising, what will the corresponding sales be?).

How accurate is such a prediction? A projection based on historical data assumes that the factors influencing the data over the historical period will remain constant. If this is a reasonable assumption in your case, the projection will be a reasonable one. Of course, the longer you extend the line, the more likely it is that some of the factors will change or that new ones will arise. As a result, best-fit extensions should be used only for short-term projections.