2024

Jozsef Mezei; Laleh Davoodi; Shahrokh Nikou

<p>In order to understand the antecedents of customer satisfaction, businesses can analytically utilise the growing amount of customer information. Unstructured text data can be used to uncover important information owing to developments in Natural Language Processing and text analytics approaches. In this paper, we focus on customer reviews posted on e-commerce shopping platforms. We perform manual data annotation to determine the sentiment of the review with respect to the most important aspects of the customer journey. The 14 extracted aspects are grouped into three categories that correspond to the stages of the customer's interaction with the e-commerce platform. We make use of a configurational approach, Fuzzy-set Qualitative Comparative Analysis, to understand how the sentiment with respect to the three stages combines to achieve positive customer satisfaction. The outcomes of the analysis show that all three stages of the customer journey play important roles in determining the final evaluation of a customer, leading to a positive or negative sentiment. The theoretical and practical implications are discussed.</p>