A Structural Model of Customer Learning on Mobile Payment

As one of the prominent areas of FinTech, mobile payment is increasingly accepted by more and more consumers. To investigate consumer mobile payment adoption behavior, we develop a Bayesian learning structural model using a rich data set from JD.com, the largest direct selling online retailer in China. We find that there exist two significant learning processes on consumer mobile payment adoption, i.e., mobile channel learning and online payment learning. On average, mobile channel learning plays a more important role in consumer mobile payment adoption than online payment learning. However, the speed of online payment learning is faster than the mobile channel learning. Second, we find a significant inertial effect on both learning processes. High membership level customers at JD.com are less likely to migrate from PC channel to the mobile channel, and also from offline cash-on-delivery payment to online payment than low membership level customers. Similarly, customers with longer shopping experience at JD.com are less likely to migrate from PC channel to the mobile channel. However, such inertial effect on online payment learning mainly exists for PC customers who did not migrate to the mobile channel. For those mobile customers who migrated to the mobile channel, the longer their shopping experience at JD.com, the more likely they switch from the offline cash-on-delivery method to the online payment. Third, we find that firm promotion policy on mobile channel is more effective to drive consumers to adopt mobile payment than the promotion policy on online payment, and interestingly the promotion policy on mobile payment directly is least effective. Finally, our policy simulations further show that the promotion frequency and duration are more important to drive consumer mobile payment adoption than promotion amount since the former two policies are more effective to facilitate consumer learning than the latter policy