Here’s How Liberty Mutual Is Taking Claims Modelling into the Future

Being able to predict behavior is a core component of the insurance business. The ability to deduce a customer’s likelihood of making a claim, whether a claim is fraudulent, and dozens of other factors helps insurance brands offer them the most appropriate service for their needs.

Historically this ability has been more art than science and has relied heavily on insurance agents’ ability to assess a situation and engage in what is little more than a checklist of "symptoms” to produce options for a particular customer. However, thanks to the unprecedented volume of data now flowing through insurance companies, the proportion of science to art is finally becoming a lot more balanced.

Liberty Mutual is one brand trailblazing in the area of data driven predictive modelling and is making smart partnerships in the highest echelons of academia to ensure insurance offerings and claim decisions are as evidence based and fair as they can be.

Predictive Modelling

Liberty Mutual has access to 200 million data points from five million claims and is now leveraging insights from that information to provide its agents with a next-generation level of insight and guidance, which they can combine with their customer-centric skills to make claims management smarter, faster, more accurate, and personalized.

"By harvesting a world of data, enhancing predictive models, and leveraging human understanding and empathy, we’ve taken claims management to an entirely new level to improve every step of the claims journey,” writes Liberty in a white paper. "Liberty’s exceptional claims service depends primarily upon exceptionally talented and committed people. Analytical modelling augments their expertise, enabling them to focus wholeheartedly on the people involved in the claim.”

This increased focus on the science of predictive modelling for claims management is also being supported via a $25 million partnership with the Massachusetts Institute of Technology [MIT]. The five-year partnership aims to leverage the power of the latest developments in artificial intelligence to help the insurance brand support AI research in computer vision, computer language understanding, data privacy and security, and risk-aware decision making, among other topics.

"We are excited to embark on this project with MIT and look forward to leveraging their leading AI research to identify, develop, and ultimately operationalize several transformational AI-enabled solutions,” said Liberty Mutual Chairman and CEO, David Long. "Through this collaboration we intend to challenge the insurance industry status quo and be at the forefront of AI breakthroughs.”

Data-Driven Claims Management

The data-driven technology being deployed by Liberty Mutual will allow original sources of unstructured data such as images, notes, calls, records, and reports, to be scoured for pertinent information which can be compiled into meaningful information which can then be further broken down for extremely granular analysis.

This will empower claims managers to better spot patterns and build predictive models for specific claim scenarios and deliver a truly next-generation assessment capability.

And the results speak for themselves. Since this original approach was implemented, Liberty Mutual has seen:

• Identification of elevated risk [>$100k] claims by day 30 double.
• Four times increase in annual recoveries without increasing cycle time or expense.
• Development of an automated fraud model 20 times better at identifying fraudulent activity than random chance.
• Prediction of claims which may have compensability issues at a rate of 60%.
• A dual strategy model which encourages specialists to consider varied paths at claim outset, which allowed for a $20k reserve reduction is a single claim.

As you can see, by building significant in-house capability in data science and analytics, Liberty Mutual is able to use natural language processing and artificial intelligence to interpret records, assess risks, and devise industry-leading predictive models, which is already driving better and more accurate service and saving money throughout the operation.

"Progress is fueled by a dedication, passion, and commitment to never settle for less,” concludes Liberty Mutual. "Today, we’re building upon our legacy of innovation by using advanced tools to segment claims based on their complexity and criticality. Enhancing decision making, creating more accurate and useful predictive models, and delivering better risk management outcomes define our ongoing mission.”

Final Thoughts

With the vast quantities of data flowing through the industry these days, thanks to the increasingly digital-centric nature of the modern insurance business, any brand which is not making the most of these capabilities might as well have one arm tied behind their back. As we can see from the case of Liberty Mutual, data has the power to drive new methods of claims management and more effectively balance the science with the art.

Data-driven claims management is sure to be part of the conversation at Digital Insurance Summit 2023, being held in May at the Hyatt Regency Austin, TX.

Download the agenda today for more information and insights.