Analyzing Customer Satisfaction Metrics in the Airline Industry
School
George Brown College Canada**We aren't endorsed by this school
Course
FINANCE 2016
Subject
Information Systems
Date
Dec 12, 2024
Pages
8
Uploaded by EarlHeatLoris35
BUSINESS, WEB AND SOCIAL MEDIA METRICS AND ANALYSISCRN-15333-202401UNVEILING CUSTOMER SATISFACTION IN THE AIRLINE INDUSTRY(GROUP 7)Jojin Jose Karippuzha (101521286)Shreya Sahal (101455183)Achint Singh BaggaKrishna Paresh Dodia (101411545)Ishika Varma (101532712)Ayomide Olarewaju (101531132)
SUMMARYIn the competitive airline industry, understanding customer satisfaction is critical for improving service quality and retaining loyalty.The purpose of this analysis is to explore different airline customer reviews to better understand what drives passenger satisfaction and identify areas where airlines can improve through the evaluation of their overall experiences with these airlines. This study aims to analyze how different factors such as seat comfort, cabin staff service, food and beverages, value for money, as well as how various causes of dissatisfactions influence the overall rating and recommendation of airlines.Using this dataset of multiple airline reviews, along with key variables like overall ratings, satisfaction with different aspects of the flight experience (such as service, comfort, and entertainment), and whether the reviewer would recommend the airline to others, will help identify patterns and trends in airline performance, customer feedback, and knowing the most influential determinant of customer satisfaction. This project can help airlines in understanding aspects to improve their services, and also aid future travelers to make more informed decisions when choosing airlines. OVERVIEW OF THE DATASET The dataset contains 23,171 rows and 20 columns. Key columns include: •Airline Name: Name of the airline. •Overall_Rating: Overall satisfaction score. •Review Title and Review: Summary and full text of customer reviews. •Seat Comfort, Cabin Staff Service, Food & Beverages, etc.: Ratings on specific services. •Recommended: Whether the customer would recommend the airline. Many columns have missing values like Aircraft, Wi-Fi & Connectivity, and other entries have categorical or numerical ratings.
GRAPHS, CHARTS, AND CALCULATIONS TO SHOW INSIGHTS FROM ANALYSISDistribution of Overall Ratings✓To visualize the overall ratings provided by customers and identify the most common rating levels:This chart reveals the frequency of each rating level, highlighting whether customers are generally satisfied or dissatisfied with the airline's services. For example, a peak at higher ratings indicates overall satisfaction, while lower ratings suggest areas for improvement.
Recommendation Rate by Overall Rating✓The graph below examines the relationship between customer recommendations and their overall ratings:This shows that higher ratings are strongly associated with higher likelihoods of recommending the airline.
Seat Comfort by Overall Rating✓These boxplots below analyze how specific service aspects, such as seat comfort, cabin staff service, and value for money, vary across overall ratings:Customers with higher overall ratings tend to score service aspects more positively, indicating their critical influence on satisfaction.
Cluster Analysis of Seat Comfort and Cabin Staff Service Ratings✓The below analysis clusters customers based on service aspect ratings to identify distinct groups with similar feedback:The clustering reveals segments of customers with varying levels of satisfaction, helping airlines target improvements specific to each group.Test Relationship Between Recommendation and RatingChi-squared: 12427.713387709806,p-value:0.0oThis chi-squared test evaluates the statistical association between customer recommendations and their overall ratings.oTherefore,the test confirms a significant relationship, indicating that higher ratings are linked with increased likelihood of recommendations (p-value < 0.05).
T-test for Two GroupsT-statistic: nan, p-value:nanoThis t-test compares the overall ratings between customers who recommended the airline and those who did not.oTherefore, A significant difference in ratings exists between the two groups, reinforcing the importance of high ratings in driving recommendations (p-value < 0.05).Logistic Regression✓Logistic regression predicts whether customers recommend the airline based on service aspect ratings and other factors.The model achieves reasonable accuracy in predicting recommendations, with important predictors like seat comfort and cabin staff service highlighted in the coefficients.Neural Network✓A neural network model is trained to predict customer recommendations, capturing non-linear patterns in the data.The neural network improves classification performance, especially for complex relationships, as shown in the evaluation metrics (e.g., precision, recall).
KEY INSIGHTS•Customer Satisfaction Trends:oHigher overall ratings correlate strongly with customer recommendations.oService aspects such as seat comfort and cabin staff service significantly influence overall satisfaction.•Descriptive Statistics and Visualizations:oDistribution charts reveal the frequency of various rating levels, providing an understanding of general customer sentiment.oBoxplots show that customers with higher overall ratings tend to rate specific service dimensions more positively.•Statistical Analysis:oThe chi-squared test confirms a strong relationship between overall ratings and recommendations (p-value < 0.05).oThe t-test suggests significant differences in ratings between customers who recommend the airline and those who don't.•Predictive Modeling:oLogistic Regression identifies critical factors (e.g., seat comfort, cabin staff service) as strong predictors of customer recommendations.oThe Neural Network model improves classification performance, capturing non-linear dependencies between factors and recommendations.•Cluster Analysis:oSegments of customers with distinct satisfaction levels are identified, helping airlines target specific groups for improvement efforts.CONCLUSIONAirlines shouldfocus on improving aspects like seat comfort and cabin staff service to boost satisfaction and recommendations. They should also use cluster analysis insights to customize improvement strategies for different customer segments.On the other hand, customers should leverage reviews and recommendations so as to guide them in making better informed decisions when selecting airlines.