The Brief
In highly saturated markets, customer retention is critical. Surfacing churn risk profiles helps providers take preemptive action before customer contract terms expire. This research project analyzed user records from four major Indian telecom companies—Airtel, BSNL, Vodafone, and Reliance Jio—to identify primary drivers behind consumer migrations.
Approach
I structured this analysis into the following phases:
- Data Cleansing: Imported messy user profiles, resolved null bill entries, and formatted variables such as contract duration and data usage metrics.
- Exploratory Data Analysis (EDA): Developed a Python analytics pipeline using Pandas, NumPy, Matplotlib, and Seaborn to search for correlations between customer complaints, bill averages, and churn events.
- Segmentation: Segmented customer bases into high-risk, moderate-risk, and low-risk churn brackets to help target retention campaigns.
The Stack
What I Surfaced
The analysis revealed key insights: customer churn is highly correlated to customer tenure lengths and plan contract commitments, rather than simple pricing tiers. Additionally, specific customer service complaint categories showed immediate spikes prior to service migration, signaling a critical window for intervention.
Outcome
Compiled a comprehensive written report and presented findings to faculty and peers, demonstrating proficiency in Python EDA pipelines, statistical visualizations, and business intelligence strategy.