A technique used in forecasting models that assigns different weights to past demand periods, instead of considering each one equally. A simple moving average technique takes the total demand for the last 'X' number of demand periods and divides by 'X' to get the average period demand- each period is treated the same. By contrast, in exponential smoothing a smoothing (alpha) factor (ex.- 0.1) is multiplied by the demand from the last period, and 0.9 is multiplied by the calculated average period demand for periods prior to the last, thus assigning a different weight to the last period. Raising the alpha factor gives more weight or emphasis to demand from the most recent period.