A leading insurance provider in the US was experiencing high attrition rates, with (i) 4% churn on 14 million customers; and (ii) ~$157 million in annual loses due to attrition. The provider was looking to mitigate this problem by proactively identifying a higher proportion of policyholders at risk of attrition and understand the causes behind it.

Platform Solution:

DemystData identified the 100 most predictive attributes, and using those external variables, tested the same group of policyholders at a three-month interval prior to their six-month insurance policies coming to an end. DemystData also identified the categories of attributes and types of changes most likely to predict churn customers.


  • The insurance provider was able to identify an additional 32,000 customers likely to churn.
  • It was further estimated that up to US$13.9 million in annual benefit could be generated for this provider using this solution.