Conjoint analysis is one of the most important tools in market research, yet it neglects fundamental insights from 'Behavioural Economics' Hence, all conjoint results are subject to two inherent distortions: firstly, linear interpolation fails to depict the typically 'step-like' utility functions of quantitative attributes. Secondly, the effects of differing levels of product involvement are artificially evened out. An 'open source' correction algorithm that efficiently eliminates both problems and bridges the gap between the rationalistic conjoint analysis and the realistic insights from 'Behavioural Economics' has been developed. The algorithm sensitively captures these behavioural effects whilst simultaneously preserving the strengths of conjoint, namely its diverse simulation options.
Price optimisation is the most efficient lever to maximize profits but all classical price research tools fail to deliver optimal results as they still assume consumers to behave like a 'Homo Oeconomicus'. Using several examples, this paper will highlight the shortcomings of classical tools like conjoint analysis or PSM and develop an open source pricing research framework the 'Psychological Price Profile' that helps to broaden the research scope while still leveraging the advantages of the classical tools. Moreover, this paper will present the core results of a multi-national study involving 16 countries and 10 product categories.
The deduction of partial utilities within conjoint analyses is based on the assumption that people maximise their utility (homo oeconomicus). This assumption is contradicted by a broad swathe of empirical findings which have demonstrated that people systematically fail to make rational decisions. They nonetheless make predictable errors, so their decisions essentially remain calculable;. Consciously or unconsciously, they simply follow a decision rule that complements utility-maximising choice behaviour. In a first step, we were able to statistically model this sub-optimal choice behaviour, and in a second step we combined both algorithms the classic utility-maximising choice model and its sub-optimal counterpart ; within one Multi Rule Conjoint Analysis& (MRC) that adaptively models a respondent's behaviour on the basis of either of these choice models. Overcoming the narrow perspective of rational assumptions, MRC is able to improve predictions by more than 45% compared to standard CBC merely by more extensively exploiting an existing set of data.