insurance
IKASI’s bespoke ensemble model allows insurance companies to identify which lost customers can be reactivated and which customers can upsell more policies.
Account for interaction effects
IKASI’s sophisticated experimental design algorithms take into account the interaction effects of holders of multiple policies.
best-in-class datasets
Our algorithms produced feature selection, objective function design, and hyperparameter turning produced a best-in-class dataset for modeling a very complicated set of data across a suite of over a dozen different insurance products.
Reactivate lost customers
IKASI’s bespoke ensemble model allows insurance companies to identify which lost customers can be reactivated and which customers can upsell more policies.
Case study
Top 25 Property/Casualty Insurance Company
Challenge
A large insurance company was looking to tactically apply promotions to improve customer retention across multiple insurance products.
Solution
IKASI’s AI model determined the optimal marketing spend at the individual customer level.
Results
■ A 4% Increase in overall net revenue across the product portfolio.
■ A 6% rise in customer retention rate.
■ Targeted groups saw an increase of 150% in the expected renewal rate of policies.
