This project simulates a comprehensive data analytics pipeline for a telecom provider operating in the Wolaita Sodo district, Ethiopia. The goal is to track, analyze, and forecast subscriber growth across different kebeles (localities) and key customer segments (Individual Users, Businesses, Telebirr Adopters).
The analysis transforms raw acquisition and churn data into actionable business intelligence to guide strategic investment and marketing focus.
The company needs a data-driven solution to understand the spatial and behavioral dynamics of its subscriber base. The analysis aims to answer critical questions:
- Market Saturation: Which kebeles are nearing saturation, and where is the greatest opportunity for new subscriber acquisition?
- Growth Drivers: Which customer segments and geographic areas are driving the highest Net Growth and MoM/YoY rates?
- Forward Planning: What is the reliable 12-month forecast for total subscriber growth to inform budget and infrastructure decisions?
This project demonstrates expertise across the entire data science lifecycle, from simulation to visualization.
| Category | Tools & Libraries | Skills Demonstrated |
|---|---|---|
| Data Simulation | Python, pandas, numpy, random |
Generating realistic time-series data with trend, seasonality, and geographic bias. |
| Data Wrangling | Python, pandas |
Feature Engineering: Calculating Cumulative Subscribers, Market Penetration Rate, and Month-over-Month (MoM) Growth. |
| Time Series & Forecast | Python, Prophet (Meta) |
Applying a robust model to decompose trends and forecast the next 12 months of cumulative subscribers. |
| Visualization | Python, matplotlib, seaborn |
Creating a multi-panel, high-fidelity Dashboard for executive reporting and geographic comparison. |
Deliverable: A comparison of Market Penetration Rate across 15 kebeles.
- Insight: Urban areas like Fana Womba show > 80% penetration, indicating saturation, while Periphery kebeles like Humbo-Aba offer high untapped potential.
- Action: Recommend shifting acquisition resources from saturated Urban centers to high-potential Periphery kebeles where the Penetration Rate is lowest.
Deliverable: Time series analysis of segment growth.
- Insight: The Telebirr Adopter segment saw an 20% spike in growth in Year 2, confirming the success of the recent digital push."
- Action: Continue investment in digital financial services infrastructure to maintain the momentum of the fastest-growing segment.
Deliverable: Predicted cumulative subscriber count for the next year.
- Insight: The model forecasts a total subscriber count of 3500 by December 2025, providing a clear target for capacity planning.
- Action: Use the predicted figures and confidence intervals for budgeting and network capacity expansion.
Developed by: Aklilu Abera