3. Modeling customer preferences

From the preference ratings collected on the different configurations (on a scale of 1 to 10, for example), you need to estimate the partial utilities of the different attribute modalities. Remember that conjoint analysis is based on the assumption of an additive model, i.e. that the overall utility of a product is the sum of the partial utilities of its component attribute modalities.

The aim is to solve the following problem: from the preference rating of a set of products, determine the values of the partial utilities of the attribute modalities, so that, for a given product, when the partial utilities of the attributes are added together, we obtain a total utility which is equal to the preference rating (to within one additive constant). It is in this sense that conjoint analysis is a decompositional method: the overall preference rating must be decomposed onto...