Configuration analysis is a procedure based on the Chi2-method. As far as we know, the German G. A. LIENERT orally reported on it two years ago for the first time. Since there are rather few presuppositions for its application, as experience has shown so far, it is especially well suited for marketing research considering the following types of problems: 1. Differentiating the connection of various alternative features; 2. Optimalizing a pure classification of test persons with regard to special features. As to the first problem, configuration analysis allows for an examination of non-linear interactions of higher orders which are known to be considered as non-existent in factor analysis, so that factors can come up which would not occur in a non-linear approach. In this regard, configuration analysis can be taken for an extension of factor analysis towards curvy-linearity.