Identifying Key Predictors of Airline Passenger Satisfaction: A Comparative Model Evaluation Study

Customer satisfaction is an important criterion for passenger airlines striving to offer better services to passengers and to survive the intense competition in the airline industry. The current study compares various predictive modelling techniques to understand the factors contributing to customer satisfaction, such as legroom, food and beverage quality, seat comfort, and check-in services. The dataset used in the current study is obtained from an open-source repository and comprises passengers’ feedback about the above touchpoints. After initial data preprocessing and exploratory data analysis, three prediction models—Discriminant Analysis, Logistic Regression, and Random Forest—are employed to predict passenger satisfaction. Performance indicators are employed to assess each model and determine the most effective predictor of customer satisfaction. The findings provide significant insights into customer preferences, enabling airlines to customize their services and enhance passenger satisfaction and overall experience.