PROPERTY/CASUALTY INSURERS, ALREADY GRAPPLING with a massive influx of third-party and usage-based data, are about to see their data stores explode with the emergence of increasingly granular driver-behavior feeds, time-series and mapping data, collision avoidance and telematics systems, and the emergence of self-driving cars, which are aware of and record not only their own every move, but those of other vehicles. This creates a seismic shift that promises to dramatically alter the auto insurance industry, the systems it operates and the markets it serves.
The insurance industry always has been driven by data and data mastery. But today P&C insurers are embracing data-driven decision making as a means of increasing profitability and differentiation in an increasingly commoditized market. And market competition is likely to soon be defined by how well carriers leverage data-based decision making across every aspect of their business, including rating, product design, underwriting, claims, acquisition, retention and cross-selling.
Consider:
Telematics, collision avoidance and self-driving technologies are lowering accident frequencies, which will drive down premiums, IT cost structures and prompt the reengineering of internal processes and core systems.
The volume and velocity of the data created by these advanced automotive technologies will accelerate the importance of data mastery and analytics to auto insurers.
The rate of consumer adoption will require insurers to accelerate speed to market, or suffer from adverse selection.
"The combination of the pace of technological change and consumer acceptance is important," says Steve Sorenson, EVP product operations at Allstate. "It's causing us to spend more time outside the four walls of Allstate. Dramatic change and disruption is starting to be the norm, so you can't sit still and watch from the sidelines. It's anyone's guess how quickly self-driving cars will come to market and expand. We have to anticipate that the timeframe for change is no longer years down the road. It's already begun."
The changes brought about by this rush of technological advancements will present CIOs with three key challenges: develop a technology road map suited to these emerging business challenges; prepare their systems and staff to react ever-more quickly to increased speed to market based on technologically-enabled consumer demand; and reduce IT cost structures in anticipation of safer roads, fewer claims and potentially declining premiums.
CHALLENGE 1: THE ROADMAP
Auto insurance business and IT executives need to prepare for the age of big data as created by telematics devices, mobile phones, traffic cameras and other devices. These new technologies will generate and wirelessly transmit incredible volumes of data into insurers' data stores, where it will be analyzed and incorporated into core systems to drive decision-making.
This proliferation of data sources and volumes of super-granular, scientific and time-series data, including date, time, speed, GPS, acceleration or deceleration (Gforce), cumulative mileage and fuel consumption, for example, simply cannot be accommodated by today's claims, policy administration, rating and underwriting systems. Insurance core systems, and particularly legacy systems, never were designed to collect, store, analyze or transmit that kind and volume of data. Instead, these new databases and analytics tools likely will be cloud-based and outsourced, and then integrated with those core systems.
For the past 30 years, the insurance industry has been driven by an essentially a query-based paradigm. Large datasets were constructed in SQL-based technologies and required someone familiar with the business, processes and the data to write queries and discover relationships in the data. That approach simply would be inadequate as insurers attempt to use advanced-automotive data to develop new rates and products.
Virtually every department and function, including actuarial, product design, billing, claims, marketing, technology and analytics teams, as well as the officers in the C-suite, soon will want or need access and the ability to manage, model and understand it through analytics and visualization. Analytics capabilities, in fact, are quickly becoming critical differentiators between industry leaders and laggards.
"Progressive's strength has always been about what you do with the information as opposed to whether we're the only ones to have it, or the only ones to be able to extract it from the vehicle," said Glenn Renwick, Progressive Insurance's president, CEO and chairman in an April 2014 quarterly earnings call. "The winners will be the companies that have adapted the technology and have the talent to be able to use that kind of data and convert it into truly marketable products."
But the quality of the answers has been only as good as the quality of the query. Insurers that have perceived these activities as lower-level tasks or that have not deployed adequate funding, talent and other resources to make best use of the available data through advanced analytics are placing their companies at a serious disadvantage. And, because these advanced-automotive technologies reduce accidents in the long term, they could reduce premium and change the way that it is paid.
CHALLENGE 2: SPEED TO MARKET
The second CIO challenge is to prepare for increased speed to market in an environment of escalating consumer expectations. As these advanced automotive technologies begin to reach consumers, consumers will rightfully expect insurers to support them with new insurance rates, products and services, which will need to be developed and modified much more quickly than previous generations of insurance products.
In addition to planning for a messaging backbone capable of handling millions of simultaneous events and a repository for storing and analyzing the data, CIOs would have to offer analytics tools, training and access to more of their business partners.
"Speed to market and pace of change are so much different than they used to be," Sorenson says. "Traditionally, IT got involved in the middle or toward the back end of products and services as they were being designed. Now we are partners on the front end. The tools and the environment we need to create are very different from what we have had historically. And it's not just the technology. It's the people who are able to manipulate that data and do things that are positive for customers."
The ability to do dynamic rating and risk assessment, based on actual behavior and near real-time data, is a primary and emerging benefit of these technologies for insurers, and data management skills will be crucial to the success of these efforts.
The CIO of a commercial P&C carrier, who spoke only on the condition of anonymity, explains that to move to dynamic rating, insurers must have the right information at the right time and must ensure there is no selective provisioning of the data coming in. The details of data transmission, whether it is streamed or stored and forwarded for example, and the ability to integrate with third parties are important and vary from one technology partner to the next, he says. To whom the data is transmitted, and whether it is detailed or aggregated/average data, also are important considerations.
"Data could be transmitted to the technology partner, which then could forward either complete or processed data," the CIO says. Other models could emerge, too, he says. The data could be absorbed by a bureau, such as the Insurance Services Office, and insurers could subscribe to access the rates. Data also could be collected by other operators, as are motor vehicle reports, and made available on an aggregated/average basis during rating times, and not on a detailed basis. "As an insurance company, we would want all the detailed data," the CIO says.
The ability to collect and analyze detailed vs. aggregate data also is an important factor in creating new products. For example, on the same earnings call, Renwick acknowledged Progressive's ability to review customers' trips and add overlays to deepen the company's understanding of driving behavior and the company's intention to adjust premiums accordingly.
"We're doing it as opposed to talking about it," Renwick said. "We overlay the speed of the car and the delta to the speed limit. We overlay the frequency of accidents in all of the United States on that very thing. We are scratching the surface of what its potential can be. We'll try to build it into the product at a pace that doesn't absorb all of the pure premium advantages and also at a pace that we believe that consumers are willing to accept."
Bringing new rates and products to market quicker will pressure all upstream activities and require increased technology and training investments across the organization. New actuarial and predictive models likely would be needed to incorporate and make use of the new data, which could have a domino effect, requiring the replacement of many P&C legacy administrative platforms. These new systems would need to be flexible and rapidly configurable to allow for the design and roll out of new products and value-added services, such as coverage priced by the trip, by the mile, or based on micro-level consumer behavior, that would anticipate and ride along with the rise of these new data sources.
The changes that could occur in underwriting, which soon could be based on actual individual behavioral data, gathered by connected cars, over time, rather than generalities and decades-old assumptions and commercially available scoring, could be huge and most certainly would require the application of automation, advanced analytics and the ability to access frequently refreshed external data. Business models could change rapidly and new ones could emerge. Commercial underwriting could move toward a pay-by-the trip or real-time view of behavior to change how pricing is done. The downstream effects could significantly affect the relationship with customers and business buyers of fleet insurance.
For self-driving vehicles, the changes could be even more extreme. "Underwriting and rating factors such as driver records and even age of the driver would be meaningless," says Louie Bode, senior solution architect at Great American Insurance. "Without human intervention, vehicles would not speed and would not drive recklessly. A 10-year old would have no risk difference when operating the vehicle than would a 35-year old. Insurance underwriting and current rating algorithms would get tossed out the window."
These advancements could drive a host of changes to the customer experience and greatly expand the relationship opportunities between customers and insurers through a variety of value-added safety and convenience services.
Consider again Renwick's thoughts about overlaying speed limit data over customers' trips. Knowing customers' locations in real time opens virtually unlimited opportunities to communicate with customers and influence their driving and spending habits. Real-time alerts for aggressive driving; rerouting based on real-time traffic and safety alerts; geo fencing, which could be used to limit a car's range, routes and risk exposures, are just some of the opportunities available.
"What we are really trying to peg is just how fast the world is changing," Sorenson says. "We are all consumers and we have certain expectations about how companies should interact with us, and that is very different from the way most insurance companies are approaching it. We don't compare ourselves against other insurers. We look to other industries, because that is where consumers live and breathe every day."
There are trillions of dollars at stake at the intersection of driving, insurance, customer behavior, customer segmentation and loyalty. If they do not invest now in analytics and flexible configurable systems, smaller insurers could be driven out of the market or find themselves saddled with less desirable customers, including unsafe drivers that opt out of offering behavioral data, and those that operate vehicles with antiquated safety equipment. And, because these advanced-automotive technologies reduce accidents in the long term, they could reduce premium and change the way that it is paid.
CHALLENGE 3: MANAGING COST
The third challenge for CIOs then is to reduce their cost structures in anticipation of these changes. This could be particularly painful for insurers that have deferred or delayed technology investments, because the revenues soon may not be available to support old infrastructures and processes.
Most insurers would argue that their businesses are sound; that the vast majority of cars are still traditional and human-driven, and that the total premium base is relatively unaffected by telematics, collision avoidance technologies and self-driving cars. However, even in the early days of consumer adoption, the cumulative impact of these advanced-automotive technologies could have a material impact far earlier than the adoption cycles of previous car technologies would indicate. Early data shows that both the incidence and severity of accidents involving self-driving cars are significantly better than human-driven cars. After more than 700,000 miles, the Google self-driving car had not yet had a reported collision while driving itself as of late May.
Collision avoidance technologies, including forward collision warning (FCW) and lane departure warning (LDW) systems form the conceptual foundation of self-driving vehicle functionality and increasingly are available in lower-priced traditional vehicles. Those and other more advanced technologies, such as automatic and camera-assisted braking and parking technologies are maturing quickly, and it's reasonable to assume that the technologies will work their way into the mainstream, as have air bags, antilock brakes and seatbelts; first as product differentiators, and later perhaps as federally mandated safety features.
In fact, FCW systems, which offer a warning to drivers seconds before a crash, and LDW systems, which offer a warning to drivers when they depart from their lane without signaling, were found to reduce the frequency of claims by 45 percent, according to "Actuarial Research on the Effectiveness of Collision Avoidance Systems, FCW & LDW," a white paper created by Ron Actuarial Intelligence LTD. Ron advised a discount of 15 percent on rate selection for vehicles with a similar safety systems. The impact of FCW and LDW systems on claim severity were outside the parameters of the study.
"Severity is a tough one to gauge as there could be higher repair costs associated with new equipment and technology in vehicles," Allstate's Sorenson says. "These new technologies will make a difference. How much difference, and how soon, is hard to know. There may be fewer accidents due to things like crash-avoidance systems, but the full benefit won't be seen till all vehicles on the road have the technology and that's going to take a while."
In the early days of telematics, collision avoidance technologies and self-driving cars, insurers likely would see profits rise across the board, at least temporarily, as the declining frequency of accidents accelerates and underwriting processes become more accurate. Some insurers may elect to use those profits to fund efforts for market share. Early actuarial data likely would support decreased premiums, but those insurers that have been slow to adapt could face severe dislocation and pricing disadvantages as more self-driving and advanced vehicles enter the roadways.
PLANNING
It's safe to say the insurance industry is facing a defining moment. Technology is now advancing so rapidly that it could soon dramatically change every facet of the industry.
Those who lead must prepare for a technologically-driven disruption, or risk being saddled with bad risks and withering into irrelevancy. CIOs and other senior leaders should create robust simulations of the business under different assumptions about adoption and the impact of the technology (see To-Do: Future History), or the industry could find itself insuring automobile components, such as batteries or light detection and ranging systems, as Progressive's Renwick recently told a group of investors and analysts.
Consider exercises in product design. Even in a scenario 100-percent populated by self-driving cars, there may be opportunities to partner with OEMs to fashion new products at the intersection of personal and product liability or to leverage data and analytical skills to use auto insurance as a platform for other financial services, business models and value-added services. The companies that take this responsibility seriously can build a new future, not just try to adapt to the one built by the leaders. As Dwight D. Eisenhower said, "Plans are nothing; planning is everything."
Chunka Mui is managing director of the Devil's Advocate Group, a consultancy that helps organizations stress test innovation strategies, and is the author of "The New Killer Apps: How Large Companies Can Out-Innovate Start-Ups."
Mike Boyle is CEO at Perseus Technical Strategies, which offers technology, product, business and executive development advice and counsel to clients in the financial services, insurance and software industries.
Chris McMahon is senior editor of Insurance Networking News.
