External data minimized
Created new rating modifier
Changes to group policy pricing
Enhanced pricing accuracy
The challenge
Enhancing claims prediction and pricing accuracy
The health insurer sought to enhance its ability to predict claims experiences by integrating external data with its internal claims data. The primary goal was to improve pricing for new business by incorporating external factors, in addition to traditional variables like age, gender, and region, to create more accurate and informed pricing models.
Key challenges
Integrating external data with existing internal systems
Effectively leveraging external data to create accurate and reliable pricing models
The solution
Optimizing claims prediction with machine learning
External data integration
Integrated external data for predictions
Used behavior and socio-economic indicators
Created a rating modifier to refine risk
Enhanced claims prediction
Used ML to predict claims
Identified key drivers (e.g., TV usage)
Reduced loss prediction gap
Implementation approach
1
Discovery phase
Validated business case
Identified key variables
Mapped external data
2
Model development
Created 200+ features
Used ML for prediction
Rolled up data for pricing
3
Results
Developed rating modifier
Reduced prediction gap
Improved pricing accuracy
The impact
Enhancing risk prediction for smarter pricing
Improved risk prediction
Closed claims prediction gap
Improved loss accuracy
Strengthened pricing
Optimized pricing
Integrated rating modifier into pricing
Adjusted quotation process
Aligned premiums with risk
Data-driven decisions
Improved claims prediction with external data
Strengthened underwriting framework
Informed pricing adjustments
Looking ahead
Expand data sources
Incorporate additional external variables for more precise predictions
Refine pricing models
Continuously adjust group policy pricing for improved accuracy
Enhance underwriting process
Leverage predictive analytics for more effective risk management
