Insurance Premium Pricing

A large insurance company wanted to retain its market share in the increasingly digitized and online marketplace for car insurance without undertaking excessive risk.

Challenge

Attract & Retain Car Insurance Buyers with Competitively Priced Insurance Policies

In order to prevent existing buyers from switching insurance providers and attract more customer, the sales and pricing division wanted to  understand

  • How to determine customers with higher probabilities of claiming damages?
  • How to craft a commercial policy for customers in different risk classes?
  • How to attract customers with long lifetime potentials (candidates for cross-selling)?

Approach

Risk profiling of customers to compute premiums

Information such as customer age, location, history, wealthy score, vehicle features, and ownership status was used to build a model that assed the likelihood of a customer crashing his/her car and each customer got a real-time risk profile.

This was then used to compute an insurance premium that was co-related to the customer profile. For low-risk customers, customer lifetime value was being predicted to offer discounts if they were likely to navigate away from the offer (based on web-activity).

Methodologies

  • Probability Models
  • Clustering Techniques
  • MRD Pricing AI

Results

  • Retained Market share on the online market
  • Improvement of customer portfolio with strategic (loyal, price-insensitive & SoW Inc. potential ) customers