2. AIRLINE CUSTOMER SATISFACTION
Interactive evaluation of global satisfaction, in-flight vs ground services, delay impact, distance, class, and customer type.
The report isn’t mobile responsive yet, sorry.
NAVIGATE THE PROJECT YOURSELF
Context & Objectives
Dataset & scope: Link to Kaggle; scope is focused on EDA + BI storytelling (no ML classifier)
Business questions:
Global: how does average satisfaction vary across all service areas?
In-flight vs ground: which factors (comfort, cleanliness, baggage, online processes, etc.) most affect satisfaction?
Delays: how do departure/arrival delays impact ratings?
Segments: do class, age, loyalty, and distance explain differences in satisfaction?
Process & Data Model
Cleaning (Power Query): fixed locale/decimal issue; treated 0 as null in ratings; unpivoted 14 service columns into Area + Score.
Derived fields: Total Delay (arrival+departure), Age/Distance/Delay buckets, Final Satisfaction (Low/Med/High), composite ID to count unique passengers.
Key measures (DAX): Avg Global Satisfaction, Avg In-Flight vs Ground Experience, % Survey Completion, % Flights with Delay, % High Satisfaction, % Passengers Filtered (ALL), % Passengers Satisfied (REMOVEFILTERS), % Factor Weight vs global mean.
Why ALL vs REMOVEFILTERS: used ALL for “share vs global” and REMOVEFILTERS for “rate within group.”
Report Structure (5 pages)
Report Structure (5 pages)
Global Services Evaluation: global gauge, avg satisfaction by area, survey completion treemap.
In-Flight Services: factor weight treemap, avg by area with segment tabs, representativeness cards.
Pre/Post-Flight Services: same framework for check-in, baggage, booking/boarding/support.
Cross-Area Correlation Analysis: narrative tiles + scatterplots (comfort vs food, cleanliness vs baggage, etc.).
Heatmap: full Python correlation matrix (Pearson) for all 14 service factors.