Customers Churn Prediction
1. Understanding the business domain
Customer churn is among the most important factors for banks and other financial institutions.
The main reason is that retaining existing customers is more cost-effective and profitable. This
report outlines the analysis and predictions of customer churn. The dataset has different attributes
related to customers' demographics such as gender, and age, behavioral attributes such as tenure,
and churn, and transactional attributes such as monthly charges, contract type, and total charges.
The main purpose of this project is to build a model that can predict whether a customer will
churn or not based on the input features. Since the output variable is binary and represents
whether a particular customer will churn or not, the classification model will be used. The
classification model and useful insights from data analysis and data visualization will provide
recommendations to banks to understand the reasons for customer churn and take necessary
actions to reduce it. The insights found related to different attributes that have a significant
impact on customer churn can help banks' administration take the necessary actions. The
prediction of customer churn with high accuracy can help banks to reduce the churn rate.
2. Understanding the Dataset
The dataset has several behavioral and transactional attributes. The key attributes include:
Age: Customer age in years.
Gender: Customer gender, (Male/Female)
Tenure: Duration (in months) the customer has been with the service provider.
Monthly Charges: The monthly fee charged to the customer.
Contract Type: The type of contract the customer is not (Month-to-Month, One-Year,
Two-Year)
Total Charges: Total amount charged to the customer (calculated as Monthly Charges *
Tenure)
Churn: Target variable indicating whether the customer has churned (Yes, No)
3. Purpose of the Project
The main purpose of this project is to build a classification model that can predict whether a
particular customer will churn or not with high accuracy. The analysis, hypothesis, and
visualizations performed in this project will provide useful insights and information related to the
key factors that have a significant impact on customer churn.
The key factor that influences customer churn: Find out the key attributes that have a
significant relationship with customer churn.
Classification Model for Customer Churn Prediction: Build a classification model
that can predict customer churn with high accuracy.
Useful Insights from Data Analysis, Visualization, and Hypothesis Testing: Find
useful insights by using data analysis, visualizations, and hypothesis testing, that can help
banks make informed decisions.
4. Formulating the Hypothesis
The following hypotheses are formulated for this project:
Hypothesis 1: Age has a significant effect on customer churn, with older customers
being less likely to churn as compared to younger customers.
Hypothesis 2: Customers with longer tenure are less likely to churn as compared to
customers with shorter tenure.
Hypothesis 3: Customers with higher monthly charges are more likely to churn as
compared to customers with lower monthly charges.
Hypothesis 4: Customers with longer contracts (1-year, 2-year) are less likely to churn as
compared to customers with month-to-month contracts.
Hypothesis 5: Customers with no tech support are more likely to churn.
