How should you interpret a difference between the prediction or target and the actual performance? For instance, what if the cost of raw materials were higher than what you would have predicted for the observed volume of activity? We will dive deeper into this question in Chapter 6 about variance analysis. For now, I just want to insist on the fundamental ambiguity of these discrepancies between expectations and realizations, and how you would proceed to reduce this ambiguity.
A first interpretation is that the performance is atypical: the production process was either highly inefficient or exceptionally efficient; the waste of raw materials was abnormally high or low. Then you should look into why performance was exceptionally bad or good to learn from this experience and either reduce risks of bad performance and increase chances of good performance.
Another interpretation is that the benchmark or target is wrong: the estimates of the cost function producing this standard are biased or unrealistic. You should therefore check again the validity of your model. In that regard, the quality of available data (i.e. recorded usages, prices or proportions) and the stability of the underlying technology should be ascertained. In any case, keep in mind that variances (i.e. gaps between expectations and performance) raise questions not only about your performance, but also about your models.
Please indicate how clear and understandable this page was for you: