Number crunching on a huge scale could spread UBI

If usage-based insurance (UBI) hasn’t yet become the Next Big Thing, as many of its proponents have been predicting for years, it’s not because it is a bad idea.

To link premiums to actual driving behaviour, based on real-time data, and use the promise of lower premiums to improve the driving skills of the insured always seemed like a win-win solution. Not only did consumers save money but they had fewer accidents, making motorways safer for everyone. And insurers benefited because they would pay out less in claims and, in case of accident, would have better data to handle them.

If the solution hasn’t fulfilled its promise, the reason was not conceptual; it was technological. Collecting real-time driving data at first required hardware, a black box or an OBD-2 device, which was costly to insurers and inconvenient to users. The use of smartphones as data sources has eliminated most of these barriers, practically eliminating expenses for insurers and greatly reducing friction for users. As a result, major auto insurers in the US and Europe have jumped wholeheartedly into the UBI market.

Now UBI is about to undergo another technological transition that may actually take the solution into the mass market: the use of ambient data to assess risk. Not only will the passive collection of collateral location, GPS and other driving-related data make UBI practically seamless for the user, it also offers insurers a new way to offer and distribute the product. And it will bring in major stakeholders from other sectors, companies with immense data and subscription assets.

“While some insurers have had success with UBI, others have struggled with technology cost and the friction associated with getting a consumer to install a sensor in their car or an app on their phone,” says Geoff Werner, UBI global product leader at Willis Towers Watson.“Leveraging existing ambient data about the consumer removes these obstacles, thus making telematics-based insurance more applicable for the mass market.”

His firm is currently in talks with a range of potential partners, including telecommunications providers, consumer electronic companies, car manufacturers, providers of mobile operating systems and providers of apps that monitor location. “Our focus is on companies with large customer bases and a few start-ups with significant growth potential,” Werner notes.

Many companies from these sectors already are or have been active in UBI. For example, Verizon, Telefonica and its subsidiary O2, Vodafone and Deutsche Telekom via its IT subsidiary T-Systems have offered or are offering UBI products, often in partnership with major insurers. This is true as well of a number of carmakers, such as GM and Renault. In addition, global navigation leader TomTom is active in both commercial and private-user UBI markets. And Google made a brief foray into car insurance but the solution, which offered consumers a means to compare car insurance offerings, was not successful.

The ambient data would be taken from connected-car sensors and systems, mobile phone mast triangulation data and location data taken from suitable apps and location histories of mobile operating systems. “This data is really powerful,” says Werner.

The solution may even be able to leverage location data captured in the background by ubiquitous social-network apps, which are almost always open, to be used in conjunction with traditional risk-assessment strategies to provide a partly monitored, mixed UBI solution that would still be a significant improvement over wholly proxy-based auto insurance.

According to Dominique Bonte, managing director and vice-president, B2B, at ABI Research,Facebook could play a role in this model. “Everybody has Facebook on their phones. And users could then compare their driving scores on the site, which would add a social element. That could create a viral effect.”

Werner says companies such as his own have built up an enormous databases comprising very granular per- second telematics data, as well as associated insurance claims data. “So, we’re well positioned to analytically develop scoring algorithms for these emerging data sets, which are often less granular,” such as data taken from a connected car, for example, which only include a limited number of driving events.

Bonte agrees, saying: “You don’t need to capture every single data point to be able, as an insurer, to assess if this is a risky driver or not. Maybe you don’t even need to use data from the sensors on the phone. A carrier also tracks where a consumer is; the carrier knows exactly which way station the consumer is connected to while in the car.”

This possibility has opened opportunities that could make UBI practically friction-free. At the beginning, driving data will be transmitted to a company via an app or a vehicle-embedded device and then scored and sent back. “But I suspect that driving data will be scored in the vehicle or on the phone in the not-too-distant future to further reduce costs and facilitate instantaneous feedback,” Werner predicts.

The data suppliers would be able to monetise the solution in several ways, he explains. One way is to fill requests from insurance companies for the data, as for example Renault has been doing with its UBI offering. They could also provide leads to a panel of insurance companies with whom they have created a relationship. “This could take the form of a price comparison service, similar to aggregators in the UK,” Werner says.  Or they could provide the insurance themselves.

Bonte has some reservations regarding the aggregator business model.  “On paper it looks obvious and logical, and what we should be doing,” he says. “But the reality is that they have struggled to build a position for themselves in the market. I think users want to interface with the brand directly.”

One challenge for the insurers will be to assess risk consistently using non-uniform data stemming from different data sources. Werner says a massive granular database will be able to overcome this barrier. Another challenge will be the need to reassure consumers that their data will be used for purposes they value.

“We think it’s imperative that there’s transparency for consumers,” he declares. “Ideally, consumers should receive information about their driving and be given driving improvement tips.” Werner goes on to say, “Once the consumer is fully aware of his or her driving habits, then it’s appropriate to ask them whether they’d like to receive any insurance offers. This approach gives the consumer the knowledge and control necessary for an attractive consumer proposition.”

The solution has generated enthusiasm among a number of UBI players, who are eager to go to market. “Insurance companies and data suppliers are very interested in progressing the use of ambient data for insurance purposes,” Werner says. “It’s very likely that we will see several products go to market next year.”

Leave a comment

Your email address will not be published. Required fields are marked *