Empowering Borrowers in their Choice of Lenders: Decoding Service Quality from Customer Complaints

When shopping for lenders, most consumers choose a financial institution based on just a few key factors: the interest rate, the distance to the lender's nearest branch, an existing relationship with the lender, and the reputation of that lender. But most consumers fail to consider an important element that will be key to their long-term satisfaction: whether the customer service provided by the lender is commensurate with the price. Our underlying assumption in this paper is that a consumer's personality traits are associated with the issues they will face. We use state-of-the-art cross-domain word vector space mapping and representative trait vectors in this space to estimate ten personality traits corresponding to each text and use topic modeling for finding the topics in a complaint. We then use two modified collaborative topic regression methods to create two complaint topic trait spaces for each lender, and test our underlying assumption by using statistical tests for this unsupervised learning problem in three cases: mortgage loans, student loans, and payday loans. We propose that lenders could be recommended for a specific user by analyzing this space, recommending a lender with the fewest number of complaints per retail customer of that lender in the complaint space neighborhood of the customer. We suggest future work that may be undertaken for the three types of loans, including the possibility that lenders evaluate their service from a customer's perspective to track customer satisfaction over time, and extensions to other parts of the service economy.