Forecasting is widely used in business, to predict future sales, use of raw materials, determine the amount of production, estimate financial needs, and human resource needs. Basically, forecasting is done to minimize business risks. The time series data is most widely used in business. There are many forecasting methods, however, not all forecasting methods are suitable for particular data. In this paper, we present how to choose a forecasting method that is fit for a particular data. The principle of selecting a fit model is to compare standard deviations between one method and another.