Insurers need to tread carefully during telematics gold rush has been saved
Insurers need to tread carefully during telematics gold rush
A gold rush of data is taking place in the insurance industry, despite its long-term data focus. As insurers hurry to capitalize on the gold rush by mining and refining new data sources, particularly those provided by telematics and the Internet of Things (IoT), they will also need to be mindful of the risks.
July 11, 2018
A blog post by Sam Friedman, insurance research leader, Deloitte Services LP
While data has always been the lifeblood of insurance, the property-casualty industry of late has benefitted from infusions of new types of information, mostly thanks to the Internet of Things (IoT) and telematics programs. Yet as insurers race to capitalize on this gold rush, the risks involved in mining and refining emerging data sources for underwriting, pricing, claims, and marketing may create some significant speed bumps.
That was one of the key takeaways at the recent symposium, “Next Gen Analytics and IoT: Powering InsurTech,” organized by the Center for Executive Education at St. John’s University in New York City. Opportunities to monitor and analyze the actual behavior and preferences of policyholders is rapidly expanding, thanks to the proliferation of sensors in vehicles, homes, and businesses, as well as on individuals (via wearables).
Data mining or minefield?
The use (and potential for misuse) of personal information is very much in the news these days. Insurers may assume they are largely immune to such concerns, given that in the past consumers have readily provided a plethora of data to carriers voluntarily in exchange for coverage or claims payments, with the most recent example being real-time telematics for usage-based auto policies. What may prompt concern, however, is how the industry goes about translating all the granular data it is now collecting to justify pricing and claims decisions with the help of increasingly sophisticated predictive models.
During the symposium, insurers were urged to beware of a number of regulatory challenges that could arise as the industry scoops up the digital bread crumbs consumers are leaving in their wake. Among the questions raised during the event for insurers to ponder:
Regulators are tip-toeing through the same data minefields as insurers, making sure the new IoT tools and techniques being deployed don’t end up doing policyholders more harm than good.
- Are policyholders fully aware of the kinds of personal data being collected telematically, and do they understand how that data is being utilized by insurers?
- Is this new data being leveraged appropriately to underwrite and price policies, as well as settle claims?
- How might insurers explain and justify the results of their new predictive models and telematic-data-driven decisions to regulators and policyholders?
Who owns the data generated by policyholders in usage-based insurance policies?
Regulators at the symposium made clear they don’t have all the answers, but expect to start asking more probing questions. They are tip-toeing through the same data minefields as insurers, making sure the new IoT tools and techniques being deployed don’t end up doing policyholders more harm than good.
Telematics in insurance
Greater disclosure was cited during the symposium as a possible antidote for many of the potential concerns raised about insurer use of telematic data. But the industry was also warned that disclosure alone is not necessarily a panacea. Is the disclosure comprehensive? More importantly, is it comprehensible to policyholders and regulators? During breaks, a number of attendees speculated that insurers may not be able to simply disclose their way around this problem. Instead, they may need to be more transparent in explaining exactly how IoT data is being used, and point out tangible benefits for consumers.
Predictive models based on mysterious algorithms analyzing alternative data may appear to be an impenetrable black box. Consumer advocates and regulators are therefore likely to want a closer look inside, as well as see certifiable research demonstrating why real-time data is superior in terms of risk assessment than the more static information that’s been provided up to now. Exactly how do these new telematics correlations play out? For example, are drivers found to be stopping short more often than average actually at risk of greater accidents? While this might sound intuitive, is the connection still just a hypothesis, or is there hard evidence to prove that assumption?
On the other hand, a number of industry speakers pointed out that insurance underwriting and pricing have always been something of an enigma to policyholders, but thanks to telematics, consumers are now being judged on logical factors they can control. Some suggested that telematics data generated by real time monitoring seems inherently more defensible, transparent, and fair than depending on some of the proxy data, such as credit history, that have historically been cited by many insurers for their decisions. At least with telematics, pricing is set according to how a person actually drives.
There were also many side benefits to collecting telematics data cited during the symposium. In auto insurance, the biggest benefit might be improved public safety, since policyholders are expected to drive more carefully if they know they are being monitored (and that their premium is on the line), while receiving valuable feedback on their driving habits and warnings about hazards in real time.
Indeed, there’s likely much to be gained by policyholders having their insurer as a ‘back seat driver’ if it helps them avoid accidents and save lives as well as money. The same might be said for having an insurer monitor one’s home, business, or very person (thanks to wearables). ‘Big Brother’ may not be so bad after all, as long as both parties—those being monitored as well as those doing the monitoring—are clear about what’s being watched and why it’s mutually beneficial in a world where ‘privacy’ is becoming relative and personal data is an increasingly fungible asset.
QuickLook is a weekly blog from the Deloitte Center for Financial Services about technology, innovation, growth, regulation, and other challenges facing the industry. The views expressed in this blog are those of the blogger and not official statements by Deloitte or any of its affiliates or member firms.
Strategies to stay competitive in the insurance market
Refine traditional marketing for digital success