The exponential smoothing models extrapolate historical data patterns. Simple exponential smoothing is a short-range forecasting tool that assumes a reasonably stable mean in the data with no trend (consistent growth or decline).

More than 25% of U.S. corporations use some form of exponential smoothing as a forecasting model. Smoothing models are relatively simple, easy to understand, and easy to implement, especially in spreadsheet form. Smoothing models also compare quite favorably in accuracy to complex forecasting models. One of the surprising things scientists have learned about forecasting in recent years is that complex models are not necessarily more accurate than simple models.

The simplest form of exponential smoothing is called, appropriately enough, simple smoothing. Simple smoothing is used for short-range forecasting, usually just one month into the future. The model assumes that the data fluctuate around a reasonably stable mean (no trend or consistent pattern of growth). If the data contain a trend, use the trend-adjusted smoothing model (TRENDSMOOTH).