Case Study – Churn Prediction Model

Industry

Banking industry (BFSI)

Function

Customer Analytics & Descriptive Analytics

Business Challenge

The client, one of the oldest and most established banks in the UAE, faced significant customer attrition in its credit card division. To address this issue, the bank aimed to develop a comprehensive BI solution and a churn prediction model to proactively manage and reduce customer churn.

Project Objective

BI Solution: Develop a bank-wide BI solution to monitor KPIs across various functions.
Churn Prediction Model: Build and deploy a churn prediction model for the credit card division to identify and mitigate customer attrition.

Approach & Solution

BI Solution Development: Utilized IBM Cognos Analytics to create a robust BI solution for monitoring key performance indicators (KPIs) across the bank.
Churn Prediction Model: Built a churn prediction model using Python, deployed on Watson Studio, with the scoring engine implemented in Watson ML.
Dynamic Insights: Integrated a customer feedback platform to generate real-time insights, with periodic analysis provided on a daily, weekly, and monthly basis.

Project Outcome

Predictive Scoring: Implemented a ‘predicted to churn’ score for new customers, enabling targeted marketing strategies.
Automated Process: Integrated the churn prediction model into the bank’s CRM and marketing tools, automating the churn management process.
Improved Strategy: Enhanced the bank’s ability to proactively address customer attrition through data-driven insights and targeted interventions.