Voice of the customer in the insurance industry: a data-driven approach to understanding and enhancing experience

The purpose of this dissertation is to develop and apply a systematic, data-driven framework for the analysis of Voice of the Customer (VoC) in the insurance industry, with a particular focus on the claims settlement process. The study aims to demonstrate how customer experience data can be jointly analyzed with operational performance indicators in order to improve customer experience and enhance operational efficiency. To this end, the research analyzes post-interaction customer satisfaction survey data, including quantitative evaluations on a 0–5 rating scale, Overall Satisfaction (OSAT) and Net Promoter Score (NPS) metrics, as well as free-text customer comments. The proposed methodology integrates statistical analysis of ordinal data (descriptive statistics and Spearman correlation analysis) with Natural Language Processing (NLP) techniques, such as topic modeling and sentiment analysis, to interpret unstructured textual feedback. The application of the proposed framework to a real-world dataset from insurance claims handling highlights the key drivers of customer satisfaction, including service speed, clarity and quality of communication, and perceived fairness of treatment. In addition, the analysis reveals strong associations between customer satisfaction indicators and operational metrics, such as claims resolution time. The proposed approach can serve as a practical guide for implementing Voice of the Customer analytics in insurance organizations, as well as in other complex service environments. Its generalizability lies in the formulation of a comprehensive analytical pipeline, spanning from data collection and preprocessing to the translation of analytical findings into targeted, evidence-based customer experience improvement initiatives.