1) Business Understanding
- Goal: Predict which customers of a telecommunications company are likely to churn (
Churn = Yes) based on demographic, service subscription, and billing data. - Expected Impact: Reduce churn through more targeted retention strategies.
1.1. Business Context
Customer retention is a top priority for telecommunications companies due to the high cost of acquiring new customers and the intense market competition. Churn (customer attrition) directly impacts recurring revenue and the long-term health of the business.
This project focuses on a fictional telecom company operating in California, providing home phone and internet services. The strategic goal of the organization is to develop targeted customer retention programs by leveraging predictive analytics based on historical customer behavior data.
1.2. Project Objective
Build a predictive model capable of identifying customers with a high likelihood of churn, allowing the marketing and customer relationship teams to proactively design personalized retention strategies.
The main data science goals are to:
- Detect key drivers of churn;
- Estimate individual churn probabilities (scoring);
- Prioritize retention actions based on churn risk and Customer Lifetime Value (CLTV).
1.3. Business Problem
How can the company anticipate customer churn using demographic profiles, subscribed services, and billing data in order to optimize retention strategies and reduce quarterly churn?
1.4. Project Goals
Primary Goal:
- Develop a binary classification model to predict whether a customer will cancel their service in the next billing cycle.
Secondary Goal:
Generate business insights through descriptive and inferential analytics, including dashboards that highlight:
- Impact of features such as contract type, tech support, billing profile, etc.;
- Demographic and geographic patterns among high-churn customers;
- Actionable recommendations for targeted customer segments.
1.5. Stakeholders
- Retention and marketing executives (campaign strategy)
- BI team (dashboard monitoring and KPI tracking)
- Data scientists (model development and maintenance)
- CRM team (automation of personalized offers and actions)