How do you build linear cost functions?


Cost estimation techniques are used to estimate the parameters of linear cost equations, separating the variable components from the fixed ones. In the next pages, I will show how you can use four techniques: the engineering approach, the account analysis, the high-low and the least-square regression. A fifth technique is also often used in practice: the conference method. This method consists in gathering the opinion of several experts to identify cost drivers and build cost functions. However, since it relies on the input from several experts, this method is difficult to illustrate in the context of an introductory textbook. Moreover, the engineering approach and account analysis can be seen as a particular implementations of the conference method.

These estimation techniques are not mutually exclusive: they are often used in combination to improve, or in parallel to check, the reliability of any single method. Now the choice of the estimation technique to apply is often dictated by available data. Since these data are likely to vary a lot depending on the kind of cost to model, it is important to understand not only how each method works, but also in which cases you can or cannot use them.