Editor: So to begin, what is Telematics?
Steve McKay: Telematics refers to the integrated use of telecommunication and informatics; it is about capturing and storing information via telecommunication devices. Telematics in private passenger auto usually means gathering data electronically to gauge how many miles a vehicle is driven and the way those miles are driven. Usually, that data is transmitted wirelessly to a server where it can be stored, analyzed, and applied to rating.
Editor: What is the origin of Telematics usage in the property/casualty industry?
McKay: There has been academic discussion of telematics in the property/casualty industry since the early twentieth century. There was small-scale experimentation in the 1990’s, but the technology then was cost-prohibitive. Technology costs have dropped rapidly in the last decade, finally making telematics cost effective on a large scale.
Editor: What is the current stage of development, both internationally and in the US?
McKay: Most large insurers in the US and elsewhere in the world have some sort of pilot telematics program with hundreds or a few thousand cars enrolled. Progressive has the largest program in the US with its MyRate program, active in at least 17 states. One or two top 10 auto insurers are expected to launch similar programs within the next year. Some insurers in South Africa and Australia have true “pay as you drive” programs in which consumers pay for their insurance by the kilometer.
Editor: Why does the perception seem to exist that the US industry is lagging behind many other countries in terms of rolling out telematics based programs?
McKay: The US regulatory environment has long been more restrictive than that in the rest of the world, so international insurers generally have launched innovations sooner.
Editor: Can you describe the necessary hardware and software components of a functioning Telematics system for PPA?
McKay: You need an economical way to gather driving data, a means to send that data to a server, and a system for storing and managing driving data. You need the right driving data and enough of it to build credible models in order for the program to work efficiently. The most important thing about the hardware and software is that they work together with product pricing, customer experience, marketing, and distribution. For example, some insurers have tested programs that are expensive and require professional installation, limiting the market for them.
Editor: Are the any differences in the necessary hardware and software components for Commercial Auto?
McKay: Many commercial fleets have needs for telematics that go far beyond insurance, as there is great value in tracking long movements of expensive assets. Fleet telematics systems can justify expensive hardware that provides a broad range of functionality.
Editor: Please describe the necessary characteristics of the interface between the insurance program and the hardware/software components.
McKay: An insurer needs a way to (1) distribute the program to insureds, (2) integrate driving data into its product and, (3) make customers’ driving data available to them.
Customers need a way to sign up for the program, such as during a new business quote, and to get the hardware into their car, either by installing it themselves or visiting a professional. The challenge with marketing such programs is that they are so new; customers need an explanation that is clear but doesn’t take too long to convey.
The hardest part of product design may be deciding what the basic terms will be: Will customers have to pay a fee? Can they be kicked out of the program? What benefits can they get? When? How will their driving be evaluated? Can that evaluation be explained easily?
Making driving data available to customers means making it easy to get to, such as through a website, and easy to understand and respond to. If a customer wants to know how to become a safer driver, you want to provide simple feedback that is easy to act upon, not “you’ve exceeded your daily allotment of 3.4g’s.”
Editor: What are the greatest challenges to carriers wishing to develop a Telematics based program?
McKay: The greatest barrier to expansion of telematics programs is the cost required to get enough data to prove a program’s economic validity. Lacking institutional expertise with technology, many insurers are reluctant to make a sufficient investment to get a credible read on loss results and customer acceptance. They may believe in the theoretical power of driving data, but they would prefer to know that the technology will pay for itself.
Insurers also must face the challenge that it could be difficult for them to develop capabilities in technology that is specialized and changing rapidly.
Editor: What advantages does a Telematics program offer the insurer and insured?
McKay: Telematics programs can provide lower loss costs through positive selection, better driving behavior, and superior long-run segmentation; improved customer retention; and growth through greater conversion at point of sale and a message that cuts through the clutter.
Editor: What ancillary services are insurers making available to insureds through telematically based programs?
McKay: Most ancillary services use GPS data to geolocate a car on demand or provide notification of events like when a car leaves a certain radius or exceeds a certain speed. While these programs can be very profitable, a large market for them has yet to emerge.
Editor: How has the regulatory environment responded to Telematics?
McKay: The regulatory response so far has been very positive. Telematics data is free from concerns about fairness and has a clear link to accident risk. Rating on telematics data has the potential to reduce accidents and total driving, which in turn may have positive impacts on the environment. A variety of the public interest groups, notably the Environmental Defense Fund, have been vocal in their support of insurance telematics programs.
Editor: How are Telematics providers addressing insured’s privacy concerns? What about stricter regulatory environments such as California?
McKay: Insurers must take privacy concerns very seriously or risk a fatal backlash. Ways to protect consumer privacy include avoiding use of GPS data, which tends to be highly sensitive, and taking steps to prevent third parties, especially government entities, from being able to search your database. Most regulatory agencies, including state departments of insurance, have shown very positive support for usage based insurance.
California’s very different regulatory environment makes any changes, including the introduction of new rating variables, extremely challenging. Given its background, California’s recent allowance for using electronically verified mileage is actually a solid step in the right direction.
Editor: What are some examples of new telematically derived rating variables that might turn out to be predictive of driving behavior?
McKay: We have seen carriers use speed thresholds, miles driven, and deceleration as rating variables. There is good reason to believe that almost any movement of a car will show some predictive relationship to loss, particularly in combination with other movements. The speed at which a car turns, for example, seems intuitively powerful. But it will take a significant volume of driving data and serious analytic resources to find which variables are best, and explaining rating criteria to consumers introduces a very different set of challenges.
Editor: Is there a risk of adverse selection? Will only better drivers opt for telematically based programs? What other changes should an insurer make to their existing programs in order to accommodate a new Telematics initiative?
McKay: People know much more about their driving behavior than the information revealed in an insurance quote, so there is strong logic behind the idea that better, overpriced drivers will be most likely to enroll in telematics programs. That means that non-telematics pools will experience increasing adverse selection. The most difficult questions facing insurers contemplating telematics programs involve balancing marketing and pricing concerns. Should participants pay a fee for the program? What discounts can they qualify for? Will those discounts be different at new business and renewal? Can bad drivers be surcharged or kicked out of the program? Are there non-price incentives to change driving behavior? After making these decisions, insurers must make changes to many aspects of their customer experience: advertising, public relations, quoting, agent training, customer communications, and customer service.
Editor: Regarding ancillary programs, how are parents using Telematics to monitor their teenage drivers?
McKay: Parents use telematics programs (1) to provide specific feedback to their teens, along the lines of “I noticed you drive 90mph on your way home from school,” (2) to find the location of their teen’s car at any given moment, and (3) to receive alerts after certain events, like their teen’s car coming to an extremely sudden stop.
Editor: Taking a cue from other industries, can you speculate at what point telematics will become purely software based without the need for hardware? For instance, are there any plans to leverage the growing availability of ADAS (advanced driver assistance systems) and increasingly more advanced onboard computer systems?
McKay: Telematics surely will be software based within the next ten years, as cars gain internet connectivity and technology costs drop. Since there are very few such programs active today, it is likely to take two to three years before most car manufacturers make internet connectivity available in new models and many years after that before most cars on the road have it. ADAS data may be more a beneficiary of driving data and analysis than a supplier. Car manufacturers may be reluctant to share information on how their ADAS systems work, but they would benefit from analysis showing which movements of cars are most dangerous.
Editor: Where is there the greatest potential for innovation?
McKay: We have only just scraped the surface of how driving behavior can be changed with effective measurement and feedback. The better we understand what behaviors are predictive of risk, the better we can help insurance customers improve their safety. Determining the content and context of effective feedback will take considerable testing, but the potential gains are enormous and will benefit insurers’ profitability immediately.
Steve McKay is the CEO of DriveFactor. He spent eleven years as a product manager at Progressive Insurance, the last two responsible for marketing and customer experience for Progressive’s usage based program, called MyRate. Steve also has worked in consumer finance at Capital One and in strategy consulting. He has a JD/MBA from the University of Virginia and an A.B. in economics from Harvard College.