A genetic algorithm-based learning approach to understand customer satisfaction with OTA websites

In an extremely competitive marketplace, it is increasingly important for online travel agencies (OTAs) to understand customer satisfaction of different segments. The survey method has been widely used to gain such understanding. However, few previous studies on the tourism and hospitality business have proposed intelligent solutions to analyze such survey data to understand customer preferences on different criteria for different segments, and to determine how customers obtain overall satisfaction across different criteria. In this study, we follow a design-science research paradigm to develop a genetic algorithm-based learning approach to understand customer satisfaction and their psychometric reasons. We further validate this approach through an empirical study for evaluating OTA websites. The results show that different customer segments have different opinions on the importance of various evaluation criteria. The results also reveal that customers tend to judge OTA websites in terms of certain important criteria, instead of by the weighted average of every factor concerned. The proposed approach and the findings of this study can provide constructive suggestions to practitioners and researchers for developing customized marketing campaigns and improving the services of OTA websites.