Voice of Mobile Banking Users: A Text Analytics to Explore Customer Complaint Factors

The quality of mobile applications (Apps) is becoming an increasingly important issue. Users can usually download these apps from the app store and post reviews about them. These user reviews are becoming a rich source of data that can be used to identify complaints that users presented. Among these reviews, negative reviews are a good source of data for capturing user complaints. In order to analyze these reviews we extracted 3,359 app user reviews of Kakao Bank in Korea in 2019 from the Google Play Store, and found six types of user dissatisfaction in 574 negative reviews by means of sentiment analysis based on text mining and topic modeling. From topic modeling, we found that the most frequent service complaints were ‘app installation error’ (26%), ‘recognition error’ (17%), ‘connection failure’ (16%), ‘error after update’ (16%), ‘authentication failure’ (15%), and ‘unhelpfulness of customer- service staff ‘(11%). This study provides an alternative customer complaint analysis for mobile banking application developer and mobile banking operator to hear the voice of their customers by using a well-established text mining technique and by analyzing the reviews of dissatisfied customers.