Twitter as a customer service management platform: A study on Indian banks

Abstract

In this study customer behavior on Twitter for the five private commercial Banks in India has been explored through the analysis of historical user Tweets. Lexicon-based sentiment analysis has been done to explore frequently used terms on the Tweets and Time-series plots have been generated to determine the flow of user sentiments. T-Test for Two-Sample Assuming Unequal Variance has been used to identify the significance of the sentiment type for each Bank. Results have indicated unique insights about the customer behavior like "Significance of Sentiments on Engagement", "Pattern of Tweeting and Timings", "Impact of Bank's activities on sentiments". This study suggests ways to handle negative Tweets and emphasizes on the need for media monitoring tools for the timely discovery of Tweets. The study also fills multiple gaps by exploring the behavior of customer's negative sentiments, the impact of grievances in creating and sharing eWOM, and Twitter use behavior.