Strategic Revenue Analysis of Megaline's Mobile Plans
A data-driven analysis to determine the most profitable mobile plan and guide marketing strategy.
🎯 Project Objective
This project presents a comprehensive analysis of customer data for Megaline, a telecom provider, to determine which of its two prepaid plans—Surf or Ultimate—is more profitable.
The primary goal is to provide data-driven insights that can guide the company's future marketing budget allocation and maximize revenue streams.
🔑 Key Findings
Higher Average Revenue: The 'Ultimate' plan consistently generates a higher average revenue per user (ARPU).
Revenue Volatility: The 'Surf' plan's revenue is more unpredictable, driven by customers who frequently exceed their plan limits and incur overage charges.
No Regional Impact: Statistical tests showed no significant difference in revenue from users in the NY–NJ area compared to other regions.
📊 EDA - Exploratory Data Analysis Summary
The 'Ultimate' plan is more profitable
Our analysis of 500 Megaline customers shows that while the 'Surf' plan has the potential for high revenue from overage charges, the 'Ultimate' plan consistently generates higher and more predictable revenue per user.
We recommend focusing the marketing budget on attracting and retaining 'Ultimate' plan customers.
Average Call Duration per Month
Call duration is similar for both plans.
Users on both plans tend to have similar calling habits throughout the month.
Average Calls per Month
The number of calls per month is comparable between plans.
There is no significant difference in the number of calls made by users of either plan.
Calls Distribution
The distribution of call durations is similar for both plans.
Most calls are short, with a few outliers of longer duration.
Calls Distribution (Zoom)
Monthly Minutes Distribution
The distribution of call durations is similar for both plans.
Most calls are short, with a few outliers of longer duration.
Monthly Minutes Distribution (Zoom)
Messages Distribution
'Ultimate' plan users send more messages.
This is likely due to the larger message allowance in the 'Ultimate' plan.
Messages Distribution
Monthly Messages Distribution
'Ultimate' plan users send more messages.
This is likely due to the larger message allowance in the 'Ultimate' plan.
Monthly Messages Distribution (Zoom)
Average GB Usage per Month
'Surf' plan users frequently exceed their data allowance.
This indicates a potential for upselling 'Surf' users to the 'Ultimate' plan.
Monthly GB Usage Distribution
'Surf' plan users show a wider distribution of data usage.
Many 'Surf' users consume data close to or beyond their plan limit.
Monthly GB Usage Distribution
'Surf' plan users show a wider distribution of data usage.
Many 'Surf' users consume data close to or beyond their plan limit.
Monthly Revenue by Plan
The 'Ultimate' plan shows higher and more consistent monthly revenue.
While 'Surf' has occasional spikes from overage charges, 'Ultimate' provides a more stable revenue stream.
Total Revenue by Plan
'Ultimate' plan contributes more to the total revenue.
This chart highlights the overall profitability of the 'Ultimate' plan compared to 'Surf'.
User Mean Revenue Distribution
The average revenue per user is higher for the 'Ultimate' plan.
This confirms that 'Ultimate' users are, on average, more profitable.
Revenue vs Clients
The average revenue per user is higher for the 'Ultimate' plan.
This confirms that 'Ultimate' users are, on average, more profitable.
Monthly Average Revenue Distribution
'Ultimate' plan has a higher and more concentrated revenue distribution.
The revenue from 'Surf' plan is more spread out due to overage charges.
🔬 Methodology
Data Wrangling & Preprocessing: Cleaned and transformed raw usage data using Python, Pandas, and NumPy.
Feature Engineering: Created new variables, such as monthly revenue per user, by aggregating calls, messages, and data usage.
Exploratory Data Analysis (EDA): Generated visualizations with Matplotlib and Seaborn to uncover trends and revenue distributions.
Statistical Testing: Conducted formal hypothesis tests using the SciPy library to validate revenue differences between the plans.
This project was completed using proprietary datasets from the TripleTen curriculum. Therefore, the notebook is not directly runnable outside that environment. However, the repository file contains all code, outputs, and visualizations for a complete review.
🚀 Explore the Notebook
Explore the complete analysis, including all code and visualizations, in the project repository.
