What metrics are essential for evaluating the accuracy of past sales forecasts?

Essential metrics for evaluating the accuracy of past sales forecasts include forecast error, mean absolute deviation (MAD), and mean absolute percentage error (MAPE).

Forecast error is a fundamental metric used to evaluate the accuracy of past sales forecasts. It is calculated by subtracting the actual sales from the forecasted sales. A positive forecast error indicates that the actual sales were higher than forecasted, while a negative forecast error suggests that the actual sales were lower than forecasted. This metric provides a straightforward way to measure the accuracy of a sales forecast, but it does not account for the scale of the forecast or the sales.

Mean Absolute Deviation (MAD) is another important metric for evaluating the accuracy of past sales forecasts. MAD is calculated by taking the average of the absolute differences between the forecasted and actual sales. This metric provides a measure of the average error in the sales forecast, regardless of the direction of the error. MAD is particularly useful for comparing the accuracy of different forecasting methods or models, as it provides a single, easy-to-understand measure of forecast accuracy.

Mean Absolute Percentage Error (MAPE) is a similar metric to MAD, but it expresses the forecast error as a percentage of the actual sales. This makes it easier to compare the accuracy of forecasts for different products or time periods, as it accounts for the scale of the sales. MAPE is calculated by taking the average of the absolute percentage differences between the forecasted and actual sales.

In addition to these metrics, it can also be useful to look at the distribution of forecast errors. This can provide insights into whether the errors are random or whether there are systematic biases in the forecasts. For example, if the forecast errors are consistently positive, this could indicate that the forecasts are systematically underestimating sales. Conversely, if the forecast errors are consistently negative, this could suggest that the forecasts are systematically overestimating sales.