Taylor Fry



GI Technical Pricing & Optimisation

In today’s highly competitive insurance market, there are two fundamental questions for any insurer: what is the true risk cost of each customer? And, given the fierce competition, how price sensitive is each customer?

Combining technical expertise with broad industry knowledge, Taylor Fry offers high performance technical pricing models that allow you to accurately calculate the true risk cost of each customer. Our optimising ensures that you’re setting the right prices to achieve optimum profitability in a competitive market.

Good modelling = improved business performance

A good pricing model brings direct practical benefits to your organisation. Taylor Fry models:

  • identify profitable and unprofitable clusters of customers
  • help determine the correct pricing for policy options
  • permit fine-grained monitoring of the performance of claims costs.

Our price optimisation carefully considers the elasticity of demand and the competitiveness of the industry at a customer level, enabling you to maximise profits across your insurance portfolio. Optimisation ensures you maximise your profits by setting prices that recognise the trade-offs between market share and profit margins in a rigorous and quantitative way.

We combine our technical expertise in modelling with a solid understanding of general insurance to bring:

  • detailed knowledge of the current market prices for home and car insurance
  • extensive experience in implementing pricing systems from beginning (data extraction, cleaning) to end (monitoring and regular scoring). This ensures our work is easily integrated into your systems.
  • pricing models that deliver enhanced geographical and socio-economic insights into your customer base.

Case study: portfolio optimisation

A client asked us to help them identify where they could increase the value of their portfolio. We were able to demonstrate that a significant proportion of the client’s customer base was being undercharged when a new sophisticated pricing model was compared to the existing one, and then to recommend a new pricing structure to increase profits.

Predictive power leads to increased profits

Built on Greg Taylor’s visionary use of generalised linear models for pricing purposes, we have developed world-class skills and tools so that our models extract maximum insight. We have the capacity to work with the massive datasets often seen today, and have tried-and-tested techniques for further augmenting our models. The result is what we consider to be the best pricing models in use today anywhere in the world, delivering maximum profits for our clients.

Below is a typical output of our pricing models. It demonstrates the difference in predictive accuracy between the old model and our new, far more sophisticated model that reveals some significant, previously undetected, mispricing.

The new predictions for average claim size are closer to the actual than the old prediction, demonstrating the significant improvement in the model. For example, the far left bars show a group of customers with an actual average claim size ($4200) substantially lower than that predicted by the old model ($6600). Our new model was far more accurate in its prediction ($4400).

When such differences between the old and new predictions are multiplied across the whole portfolio, our new model generates a significant reduction in mispricing and a significant increase in profits.


Richard Brookes


(02) 9249 2911

Jonathan Cohen


(02) 9249 2930