Problem

The bank was struggling to accurately identify churning customers and lacked effective strategies for customer retention. This led to a decrease in revenue and customer satisfaction.

The main challenge was the imbalance in the data, with only a small percentage of customers churning. Additionally, there were over 25 variables to consider, making it difficult to pinpoint the key factors contributing to churn.

Solution

To address these challenges, we first cleaned and analyzed detailed data on over 170,000 customers. We used undersampling and feature selection techniques to balance the data and focus on churning customers.

We then applied various machine learning models such as Logistic Regression, Decision Trees, Ensemble Methods, and Neural Networks for customer churn prediction. After conducting a comprehensive evaluation using metrics like accuracy, precision, recall, and F1-score, we selected the best models for the bank's aggressive and conservative retention plans.

Impact

Increased customer retention rate by X%
Improved revenue by X%

"This project had an incredible impact on our business. We were able to accurately identify churning customers and implement effective retention strategies that improved our bottom line."

Team

Team Member #1 - Data Scientist
- Cleaned and analyzed detailed data of over 170,000 customers
- Used undersampling and feature selection techniques to balance the data
- Applied various machine learning models for customer churn prediction
Team Member #2 - Business Analyst
- Conducted comprehensive evaluation using metrics like accuracy, precision, recall, and F1-score
- Selected the best models for the bank's aggressive and conservative retention plans
- Based on cluster analysis, profiled users in layers and designed more valuable products for high-risk elderly customers.

Bank Boosts Retention and Revenue with AI