The error measurements are used in multiple ways, such as checking the quality of our forecast, adjusting the weights in the forecast to improve accuracy, and ensuring we have adequate accuracy in our forecast. QUALITY: First and foremost, we want to ensure that the weighted average approach is a suitable forecast model. We can plot the errors to see if there are any patterns in the errors (we do NOT want patterns as that means we failed to predict something). If the errors appear random, good. We would also do a quantitative check for autocorrelation. ACCURACY: We can calculate a summary metric for the errors (such as RMSE, MAE, or MAPE). We can adjust the weights used in the forecast to try to get a more accurate forecast. For example, we could try weights W1 = 35%, W2 = 30%, W3 = 20% and W4 = 15% to see if it gives a lower RMSE. The weights that provide the lower RMSE would be more accurate (have less error). ENOUGH ACCURACY: The summary metric (such as RMSE, MAE, MAPE) also give us an idea of the overall accuracy of our model. We can compare the accuracy from this model to the accuracy from other models to see which model is most accurate. Alternatively, if the accuracy is insufficient, we may choose to get more data or examine additional explanatory variables to see if we can improve the accuracy.