Big Data Disruption in the Insurance Industry

By Revolution Analytics senior program manager James Peruvankal

Big Data is driving immediate changes in the insurance industry that will have long term effects well beyond its industry impact. Reacting to pressure from competitors and shareholders, insurance companies around the world are working to improve their analytical capabilities to deliver innovative and personalized products to both acquire new customers while ensuring the profitability of all their customers remains high.

Big Data allows companies like Tesco to supply retail loyalty card data from their stores to their insurance business. This data helps to make modeling for pricing data more accurate across demographics and geographies. Large scale data mining and machine learning developed for managing their loyalty card program, deliver insights that their insurance arm can use to improve their price and risk modeling to improve profitability while maintaining appropriate loss reserves. Big Data operations like this help organizations fight off the industry newcomers and other competition who fail to innovate with their data sources. Traditional insurers need to respond to the challenges posed by many innovative newcomers or they risk going the way of the small local supermarket or independent book store (remember them?).

In the US, home insurers, especially in the wake of the 2008 financial crisis, face significant challenges to maintain profitability. Pricing models are many times inadequate and there are significant regulatory constraints. For one insurer a combination of data visualization and multiple machine learning models were used to develop a new pricing model to identify seriously underpriced policies. They found that 5% of the poorly priced policies contributed 14% of the loss ratio. The new pricing model helped to restore their profitability. 

To learn more about how Big Data is impacting the use of advanced analytics in the insurance industry, you may want to subscribe to our online course, delivered by Allan Englehardt, this September 23rd – 27^th. You can enroll through the link below.

Revolution Analytics Training: Using Advanced Analytics in the Insurance Industry

Comments

[Glenn Strycker]

My experience as an analyst in the insurance industry is that there is a managerial/business motivation to have overly complicated models (gee-wiz, look at my fancy model that uses triple-interactions to segment the population into 1MM groups), and these models overfit. Unless the upper management really understand statistics, the larger companies are at risk of having poorer pricing models than less sophisticated competitors.

On the other hand, the industry is so highly regulated that the full use of big data is limited -- I would have been more interested in staying in that industry had I had the freedom to implement my findings.

[Rovi]

Great wonderful Post!