Looding...

Identify High-Risk Customer Groups

Insurance
case study

Business Impacts

2.7X

Improvement of previous model

Customer Key Facts

  • Location : North America
  • Industry : Insurance

Problem Context

The customer is one of the largest providers of supplemental insurance in the United States, providing financial protection to more than 50 million people worldwide. They wanted to identify the customer groups that have a higher chance to churn and better understand the reasons behind this higher risk of churn while also identifying levers they can pull to mitigate the risk of defection.

Challenges

 

  • Only two years of data made the train and test split difficult
  • Very few independent metrics available to capture trends or take a top down approach
  • Dataset available was highly imbalanced

Technologies Used

RStudio

RStudio

Amazon EC2

Amazon EC2

Amazon RDS

Amazon RDS

Apache Zeppelin

Apache Zeppelin

Identifying Customer Groups That Are Highly Likely to Churn and Potential Reasons Behind the Churn

Solution

Quantiphi created a risk scoring model that outperformed the customer's existing model in identifying the customer groups that are at a high risk of churning out and also suggested interpretable reasons behind the same.

Result

  • Enabled developing an action list for the business to prevent loss of customer groups and potential acceleration of business growth by providing intuitive recommendations

Thank you for reaching out to us!

Our experts will be in touch with you shortly.

In the meantime, explore our insightful blogs and case studies.

Something went wrong!

Please try it again.

Share