Uncovering unobserved heterogeneity bias: Measuring mobile banking system success

Abstract Mobile banking (MB) involving the use of mobile devices to access bank accounts for conducting financial transactions has proliferated in recent years but inconsistently among banking customers. This diversity of use increases the complexity of uncovering unobserved heterogeneity bias in the success of MB systems. This research, first, uses a study of objective measures to cluster 4478 MB users into three homogeneous segments based on the system utilization behaviors captured in the log data. The users were, then, surveyed using a field study of subjective measures based on the information systems (IS) success model. A priori sample segmentation used for facilitating the discovery of unobserved heterogeneity bias in the full sample. The analysis of the subset with 445 users who responded indicated that the path coefficient and explained variances (R2) of the IS model were higher in the segments compared to the full sample; and the influence of success factors on satisfaction and intention to use varied significantly among the segments. Our study, consisting of objective and subjective measures, has theoretical and practical implications for MB usage. It contributes to the IS success model by confirming the existence of unobserved bias in full sample and inconsistent effects of the quality factors on satisfaction and continued usage in the segmented samples. It also assists the banks in identifying MB features that are more appealing to the varying user groups, which could help them with customer retention.