Context and objectives
An international insurer needed to transform its pricing practices to make them more data-driven.
The objective was to deploy algorithmic pricing practices that dynamically set product offers, prices and promotions.
Outcome
On average across the group, +10% on the top line and +2 percentage points on the bottom line.
Our approach
Step 1: identify and collect
- Identification and collection of all relevant data: internal data, customer data, vehicle data, geocoded data, other external data, etc.
- Separate projects for contract renewals and for new business acquisition
Step 2: build
- Machine learning scores to predict customer risk, price elasticity, customer behaviour, market competitiveness and lifetime value
- Definition of the business targets, the price optimisation framework and the solver
- In parallel, design of the IT architecture hosting the solution (dynamic price optimisation in production), with a make-or-buy decision at each step of the pricing process
Step 3: pilot
- Pilot of the first renewal and acquisition cases in one entity, then roll-out to four entities in four geographies
Our experts
Other case studies
All case studies- Unlocking value from alternative data Using non-traditional data sources for pricing in insurance Read the case study →
- Pricing strategy with competitor open data Competitive analysis from open data and machine learning models for a national insurer Read the case study →
- Open data strategy Building a custom open data strategy in insurance and banking Read the case study →