Ability to read and understand big data is critical
Vehicle insurers are facing new challenges to find the data allowing them to commence offering new telematic insurance policies.
According to a new research review of the consumer experience the pace and intensity of telematics development is clarifying the types of telematic policies and the information required to manage and operate a profitable line.
The importance of these issues to both consultants and their insurance clients was made clear when they were stressed by Towers Watson, when making the opening session of the recent international telematics insurance conference – a key note presentation which set the tone for much of the rest of the event.
It warned insurers that the move to usage based insurance (UBI) was
- a big project,
- influenced every part of the existing insurance enterprise,
- was expensive and required deep pockets and
- typically took three years to implement.
Consultancy view
The conference was shown how auto insurance policies were developing under the influence of telematics. The move was from:
- traditional risk policies – based on the driver, the driver’s estimated mileage, the location of the driver’s garage, and the driver’s prior claim history. Telematics enabled a simple move to:
- pay as you drive (PAYD) – based on use; the number of trips taken by the vehicle, the time of day those trips were taken, and the length of those trips (total mileage driven). Greater telematics sophistication allowed:
- pay how you drive (PHYD) – based on ow the vehicle is driven noting factors like rates of acceleration, braking, cornering, excessive speed and approximate location as recorded by GPS. A further development is:
- Full PHYD – all the above factors plus algorithms capable of calculating manoeuvring, anticipation, aggression and adaptability.
Dirty data
The ability to sort the useful data from the trash is an important exercise, the conference heard. This issue was also discussed at the October international convention in Japan where a session devoted to data heard that it is reliably estimated that 80% of all data is useless for immediate use and needs cleaning. Techniques were required to:
- place in the vehicle data in the correct driving context
- align this to the policy and claim information so that the consequent score was in the correct insurance context
- prevent repeated counting of the factors that were needed while ignoring those factors proved to be redundant
- uncover the single factors which acted as proxies for multiple factors.
Big project – deep pockets
The Towers Watson research emphasised the size of the telematics implementation project. Three years, they estimate and a huge cost requiring deep corporate pockets. It warned the project would cut across conventional business structures touching and fundamentally re-configuring
- systems (claims and billing)
- operations (support teams)
- processes (new business)
- training (workforce schemes and internal information promotion and evaluation).
