Picture of Christopher A. Hopkins, CFA

Christopher A. Hopkins, CFA

Insurance companies and Big Data: the good, the bad, and the creepy

Following a devastating fire in Colonial Philadelphia, Ben Franklin co-founded the first American property insurance company in 1752. Then as now, the role of an insurer was to collect sufficient premiums to cover losses over time. Franklin established premiums based on estimated risk of loss assigned to each property, with riskier enterprises charged higher rates or refused coverage outright.

For 250 years the tools of the insurance trade remained essentially unchanged: actuaries, underwriters, adjusters, and sales agents evaluate available information, issue policies, and pay claims. What has changed is a supernova of data and the sophisticated computing power to process this mother lode. The industry is undergoing a revolution based on the ever-expanding digital profile of your life that portends greater profits for the insurers and better products and services for you, but at the cost of further erosion of personal privacy. Your insurance company probably knows more about you than your doctor.

The term Big Data was coined to describe the ballooning store of unstructured information that can be analyzed by powerful computers to detect patterns beyond the capability of humans to spot. There are currently 90 zettabytes of electronic information in the world (90 trillion gigabytes), approximately the number of grains of sand on all the world’s beaches. 90 percent of it was created in the past 2 years.

What you may not appreciate is the number of devices that Hoover up your information without your awareness. If you own a Roomba robot vacuum, the device is busily creating a detailed digital blueprint of your home which will the company’s new owner, Amazon, will add to its already burgeoning storehouse of information about you. It is estimated that there are 20 billion connected devices excluding computers and smartphones, all actively chronicling your lifestyle. Even your thermostat is spying on you.

Imagine how valuable this digital model of your activity would be to an insurance company. If Roomba says your home is cluttered, you pose a greater fall risk. If Ring tattles that you didn’t set the alarm, you’re a greater target for thieves. Buy scuba fins online? You get the idea.

Insurers are racing to embrace Big Data. While previously available information like prescription and medical records, credit reports, DMV reports and criminal record have been used in underwriting, the process depended upon sometimes unreliable information submitted by applicants and claimants, interpreted by human underwriters and adjusters in a slow and inefficient grind. That is changing.

Artificial Intelligence allows insurers to crunch massive amounts of newly available information about customers to offer more tailored products, shorten the sales cycle, expedite claims processing, and reduce fraud. The tradeoff for customers is the additional loss of privacy.

Insurance companies have always stalked social media to see if your lifestyle it at odds with the information you provided. Posting pictures of yourself base jumping will likely get your disability benefits cancelled. Artificial Intelligence can now dig deeper into your Facebook posts, assessing your state of mind, applying facial recognition to obtain information about you from posts of others, and examining what posts you “like”, a sort of digital Rorschach test to assess your risk. AI can even track your online review history to see if you are a pessimist (a trait correlated with higher claims).

About 15 percent of drivers already allow their auto insurance company to monitor their whereabouts and driving habits in exchange for lower rates. Currently, this level of surveillance requires consent, but one could image such permission being a requirement for coverage in the future. And don’t be surprised if Google shows you a coupon for the pizza place you pass each day at 5:00.

The next frontier for AI-based underwriting is wearable technology like smart watches and trackers. The plethora of data about your daily physical movements is a gold mine for insurance companies and potentially beneficial to consumers who are health conscious and seeking better rates. Already, life insurance policies are available that require 24/7 monitoring of your activity, and send you alerts to nag you to move or warn of unusual metrics. The technology has proven too intrusive for most customers but is gaining acceptance slowly as Americans become more inured to ubiquitous digital tracking.

Data privacy advocates remain cautious. Facial recognition software has so far proven to be less accurate with people of color, raising concerns of discrimination. People of lower income who cannot afford to be as connected can also be at a disadvantage. Insurance companies have so far been less than fully transparent about how they use or share your data with others or internally to cross-market. And of course, the more your data is collected and disseminated, the greater the chance it will escape into the wild. 

Big Data holds the potential for improved customer experience. Start-up firms are using technology to bind coverage in 10 minutes or less from your smartphone and no human interaction. Traditional firms can issue some life policies with no physical exam based on AI. Chatbots apply a subset of AI called Natural Language Processing to mimic human interaction and continually learn to better interact with customers. Reduced costs and increased competition have benefitted many consumers with lower premiums and better service, and higher margins are welcome to shareholders. Our widening trail of digital breadcrumbs ensures a continuing tradeoff between privacy and convenience that has yet to find its equilibrium.

Share this post