By Selim Cavanagh, Vice President of Telematics, Wunelli
For telematics to reach the next level and achieve mass market take up, there are two key challenges for the industry: data collection methods need to widen to support greater engagement amongst consumers and investment needs to put into the validation of the mass volume of driving data that will come from an increasing number of motorists.
UBI gives insurers a simple, digestible driver score that is demonstratively related to loss propensity. Driving scores are particularly powerful because they give incremental lift to traditional data sources, such as, credit scores or demographic data so that rather than relying on proxies, insurers can calculate risk based on actual driving behaviour.
Telematics products also provide continual data throughout the life of the policy, enabling insurers to monitor the ongoing risk associated with a specific driver and create engagement through feedback to help support better driving. We know, for example, from working with insurers that some have seen a 30% reduction in claims when drivers are given feedback on their driving behaviour. Self selection is a key factor in this (drivers choosing to be monitored through telematics). However, the transition to the mass market will come not only from self selection but also from better risk selection through driver behaviour data.
Clearly, data quality is crucial. If consumers do not believe an insurer can be trusted with their information, or if they doubt the accuracy of the data being shared, they will quickly become disengaged. In addition, when working with traditional data sources for underwriting and pricing, insurers expect the data to be validated and highly accurate. Data collected from telematics policies needs to be held to the same standard and as it includes personal information about consumers, it must be managed accordingly.
So how you collect driving data is the first issue. Fixed black box devices are used by a number of insurers and while the cost of acquiring and fitting these devices has fallen in recent years they are still costly and operationally challenging.
As a result, standalone smartphone apps have emerged as a more cost-efficient way to obtain driving data and this year further options will emerge to give insurers and consumers greater flexibility.
But if we look at telematics apps for smartphones, it’s easy to see why they have become popular. They are ubiquitous, and advances in app development mean that concerns over battery drain and data usage have been alleviated. Apps also offer greater engagement opportunities with some of the more successful products offering a fun, game-like experience. Other features to drive adoption include the ability for the phone to automatically start and stop collecting data—thus making it less intrusive for customers to use, as well as a driver ‘signature’ capability offering insurers the reassurance that the driver is actually the policyholder.
But while many of these telematics apps look similar, they differ dramatically on the back-end and accurate data can be difficult to achieve due to the noisy data that smartphones can produce. In a study of more than 20,000 drivers by Wunelli and LexisNexis Risk Solutions a significant variation in how different types of smartphones logged acceleration events - in some cases, up to 55% difference.
Given the hundreds of different handsets available on the market, this variation can significantly skew driving scores. You can imagine a scenario where a policyholder dramatically improves their driving score overnight because they have changed their smartphone. Consequently, data needs to normalised in order to obtain consistent scores across the entire spectrum of handsets.
In essence, to obtain reliable and accurate data, app-based telematics policies should take extensive measures to:
• Filter and validate driving data to ensure its accuracy and reliability
• Apply handset-specific rules to account for phone variances
• Augment data with maps and road speed limits to derive context and meaningful insight
• Accurately determine when a user is driving (and by extension, not driving)
In short, while smartphone solutions dramatically decrease the cost of running a telematics policy, only accurate driving scores can enable an insurer to improve its risk management capabilities and bottom line—and in the long run, to transform how the industry assesses and prices risk. With the right filtering and validation in place, data from a telematics app can be as accurate as data from other sources, such as black boxes.
With the UK Government announcing plans to work with the insurance industry to incentivise safer behaviours and to reward the uptake of those new technologies such as telematics to improve driving skills, we’re on the cusp of a significant change in the telematics insurance market and data validation will be absolutely key as we move forward.
|
