Analysis-of-a-telecommunications-company

ConnectaTel – Customer Usage Analysis

Go back to the portfolio

An exploratory data analysis project focused on understanding mobile service usage patterns for ConnectaTel customers.

The project analyzes calls and SMS behavior, detects outliers, and builds customer segments based on usage patterns to support data-driven business decisions.

Identify usage patterns, detect atypical behaviors, and understand which customer segments show differentiated needs, in order to optimize the commercial offering and improve the user experience.

Python Pandas Seaborn Matplotlib Jupyter


Project Objective

The goal of this project is to analyze how customers actually use mobile services in order to:


Main notebook path

notebooks/Project-ConnectaTel_Gerardo_Olm.ipynb

Datasets

The project uses three datasets.

plans.csv

Plan catalog including:


users_latam.csv

Customer information.

Main variables:

Data quality issues detected:


usage.csv

User activity records.

Includes:

This dataset allows the analysis of actual service consumption.

Data quality issues detected:


Analysis Workflow

The project follows a structured data analysis process.

1️. Data Exploration


1.1. Working with copies


2. Data Cleaning


3. User Usage Profile

Usage metrics were aggregated per user:

This created a user-level usage profile.


4️. Exploratory Data Analysis (EDA)

Distribution analysis for:

edad_por_plan

mensajes_por_plan

llamadas_por_plan

minutos_llamada_por_plan

Distributions were compared between Basic and Premium plans.


5️. Outlier Detection

Outliers were identified using:

Variables analyzed:

boxplot_age

boxplot_cant_mensajes

boxplot_cant_llamadas

boxplot_cant_minutos_llamada

This helps detect:


6️. Customer Segmentation

Users were classified into two categories:

usuarios_por_grupo_uso

usuarios_por_grupo_edad

This allows identifying high-value customer groups.


How to Run the Project

Run in Google Colab

  1. Open Google Colab
  2. Upload the notebook

notebooks/Project-ConnectaTel_Gerardo_Olm.ipynb

  1. Upload datasets:

plans.csv

users_latam.csv

usage.csv

  1. Run the notebook cells sequentially.

Reproducibility

To reproduce the full analysis:

  1. Load datasets
  2. Execute data cleaning steps
  3. Generate user usage profile
  4. Perform exploratory data analysis
  5. Detect outliers
  6. Build customer segmentation
  7. Review insights

The notebook is organized sequentially for easy replication.


Key Skills Demonstrated


Author

Gerardo Olmedo – Data Analyst


Go back to the portfolio