Personalization in P&C: Fixing Adverse Selection Before It Fixes You

In this episode of The Tx Show, Rakesh engages with Arun to challenge one of North America insurance’s biggest buzzwords: personalization. While carriers invest heavily in AI, journeys, and next-best-action, margins remain under pressure from severity inflation, social inflation, and rising expense ratios. Arun argues that most personalization today is cosmetic, not economic, and explains what actually moves the P&L: risk mix, price adequacy, claims severity/leakage, and cost-to-serve. The conversation gets practical with real examples from auto and property, a claims pathway model that avoids exception overload, and the 4 steps a COO can run without waiting for core system change. 

Key highlights include: 

• Why personalization fails when it improves experience but not underwriting and claims economics. 

• How coarse segmentation drives adverse selection in auto and property portfolios. 

• A pragmatic approach to “granular pricing” using workflow controls when filings and cores lag. 

• Claims personalization that works: early FNOL triage, routing, and leakage control without inflating exceptions. 

• A 4 step execution plan focused on combined ratio outcomes, not model accuracy metrics. 

• Governance essentials for fairness, regulatory scrutiny, and defensible data usage.