A longitudinal examination of customer commitment and loyalty

Purpose – This study aims to provide the first longitudinal examination of the relationship between affective, calculative, normative commitment and customer loyalty by using longitudinal panel survey data. Design/methodology/approach – Repeated measures for 269 customers of a large financial services provider are employed. Two types of segmentation methods are compared: predefined classes and latent class models and predictive power of different models contrasted. Findings – The results reveal that the impact that different dimensions of commitment have on share development varies across segments. A two-segment latent class model and a managerially relevant predefined two-segment customer model are identified. In addition, the results demonstrate the benefits of using panel survey data in models that are designed to study how loyalty develops over time. Practical implications – This study illustrates the benefits of including both baseline level information and changes in the dimensions of commitment in models that try to understand how loyalty unfolds over time. It also demonstrates how managers can be misled by assuming that everyone will react to commitment improvement efforts similarly. This study also shows how different segmentation schemes can be employed and reveals that the most sophisticated ones are not necessarily the best. Originality/value – This research provides the first examination of models for change in customer loyalty by employing survey panel data on the three-component model of customer commitment (affective, calculative, and normative) and considers alternative segmentation methods.