Big Data, Analytics, and Machine Learning: Changing Insurance
The evolution of analytics and machine learning are shaping the industry’s future, mitigating risk, improving business, and helping develop types of policies that individuals and businesses need, even before they know they do.
Success, failure, and change in the insurance business have always been largely data-determined. But today is different. The proliferation of big data and the advent of advanced analytics and machine learning are changing the game entirely. The winners are those that can access the most relevant data, analyze it in new and unique ways, and apply it at the right time and place, all at extraordinary speed.
These new technologies are not only improving company performance, they are also helping insurance carriers identify new areas of risk and opportunities for growth.
The evolution of auto policy rating is a ready example of how more and better data, used well, leads to improved risk analysis and pricing. Historically, age, marital status, and driving history were used to price a policy. Over time, other data such as credit scores and good student discounts were included. Carriers now have the opportunity to include driver behavior data, captured directly from vehicles through telematics devices, to further improve pricing accuracy.
Carriers are also using new sources of data to proactively mitigate risk, helping to minimize loss, or even preventing it altogether. Use of information from water detection sensors in the home to stave off significant damage or flooding, or from devices worn by truck drivers, miners and other employees to monitor alertness, are just early indicators of the extraordinary preventative value inherent in the combination of data, devices, and analytics.
As access to data from sensors and connected devices expands, property and casualty (P&C) carriers also have increasing visibility into consumers’ lifestyles, patterns, and preferences. Layering analytic solutions and machine learning technology on top of that, carriers can quickly mine that information for new areas of risk, and new coverage opportunities.
It’s clear, for example, that there is an increasing need for more non-traditional coverages-as-needed, those linked to personal or time-and-place circumstances, and personal/commercial hybrids, to name a few. I call this sort of policy personalization “insure me” coverage.
As a simple example, though the days of buying travel insurance at an airport kiosk or from a brick-and-mortar travel agent are all but gone, the proliferation of online data and transactions has brought with it travel insurance 2.0.
