Model outperformed all
Cuts reduced
Half premiums captured
Reduced loss ratio achieved
A home insurer aimed to grow its policy book by strategically targeting profitable business. It wanted to predict non-catastrophic losses per household using home attributes and estimate both unrestricted and ideal premiums to gauge profitability. It also wanted to identify policies for non-renewal based on profitability and catastrophic exposure, for a balanced and sustainable portfolio.
Key challenges
Estimating premiums for profit
Identifying unprofitable policies
Predicting household losses
Balancing growth and risk
The solution
Data-driven modeling
Integrated claims and premium
Identified loss drivers
Predicted losses with ML
Enhanced accuracy
Utilized home data
Built ensemble models
Compared methods
1
Peril-based models
Captured risk patterns
Enhanced accuracy
Applied ML techniques
2
Policy optimization
Flagged renewables
Assessed profitability
Identified key loss factors
3
Strategic targeting
Defined targets
Determined engagement
Set action timing
Improved model
Improved over prior models
Refined results
Reduced loss ratio
Loss reduction
Loss ratio
Oriented toward revenue growth
Improved renewal retention
Premium optimization
Captured premium
Targeted profits
Improved performance
Ongoing model refinement
Optimize predictive accuracy
Broader policy insights
Expand model application across new policy types
Sustained profitability
Drive sustained growth with data-driven strategies
