Feature Complementarity and Assortment in Choice

This research argues that the purchase probability from a given choice set is contingent on the complementarity of the features differentiating its options. In particular, two types of features are distinguished: complementary features, which are characterized by the additivity of their utilities, and noncomplementary features, which are characterized by nonadditive utilities. In this context, it is argued that assortments in which options are differentiated by noncomplementary features are likely to be associated with a greater probability of purchase than assortments with options differentiated by complementary features. This prediction is supported by data from three experimental studies. The article concludes with a discussion of the theoretical implications and offers directions for further research.