Mar 21 2023
Traditional risk models in the auto insurance space incorporate multiple data points, including vehicle details, the driver’s safety record, the vehicle’s geographic location, and that location’s associated weather patterns. While these factors may be weighed differently by each insurer’s model, the entire auto industry is now facing the reality that increased vehicle complexity will force a significant change to risk analysis.
PwC’s recent report on the trends transforming the automotive industry notes that a universal pivot is occurring toward cars becoming “E.A.C.S.Y.” - electrified, autonomous, shared, connected, and “yearly updated.” These long-term trends are already taking shape with the increased popularity of electric vehicles, rideshare services, investment in ADAS (Advanced Driver Assistance Systems), and software dependencies.
The “E.A.C.S.Y.” movement, PwC notes, “will have far-reaching consequences for the entire industry and its value chains. Elementary structures and attitudes will have to change fast to cope with the developments by 2030 and beyond.” ADAS safety features are already changing the vehicle’s importance in policy development, with a potential for higher levels of vehicle automation appearing in the near-term horizon. Insurance providers’ role in this industry-wide shift will be to weigh the vehicle itself more heavily than ever, but it’s a change far easier stated than accomplished.
At the manufacturer level, parts complexity is already increasing – Mayank Sikaria of the Forbes Tech Council notes that modern cars now have between 100-150 ECUs (electronic control units), many of which have unique dependencies on other parts of the car. Sikaria elaborates that beyond the underlying software, “the next layer of complexity is different vehicle trims. The feature codes associated with a premium trim will differ from a sport or base model. Then, there are the optional upgrades—adaptive cruise control, heated seats, lane assist, voice recognition, and more. There are also geographical and homologation requirements”.
These evolutions in vehicle construction and design explode the number of variables for insurance providers to consider, but there’s also the problem of how OEMs designate and package these features and their associated parts. What is the difference between, for example, “EyeSight Driver Assist” and “Swift Safety Packages”? Or even more concerning, what is the difference between “EyeSight Driver Assist” across model years of the same brand, or across different models or even different trims of the same model within the same model year? The more variables introduced, the more error is possible in risk modeling for policies and claims, and insurance providers may not be able to offer prospective customers the best possible policies, or they may needlessly inflate their loss ratios, with flawed models.
OEM build data is the manufacturer's record of each vehicle they produce and includes the description of all the options and packages that were included on the vehicle at the time it was produced. Direct access to OEM Build Data for insurance providers may seem to be an attractive option for gaining insight into the features of the vehicle. However, each OEM’s Build Data is usually mired in the marketing terminology we described, and can create an inefficient logjam for making judgments on policy creation and claims. This actually makes sense because the OEM Build Data exists to facilitate vehicle ordering and manufacturing, rather than to provide any clarity in understanding or comparing the features of vehicles across brands.
While discussing the possibility of OEM-sponsored insurance policies, Automotive News noted that the relationship between vehicle manufacturers and insurers is “transactional in nature and short-term in its outlook.” The author emphasizes that this relationship needs to change, because “OEMs are sitting on a valuable source of data that an insurer could use to differentiate and improve its product proposition for customers. But this isn't always made freely available.”
Given the accelerating complexity of the vehicles being manufactured, auto insurance providers working without clear and accurate vehicle data is a threat to continued operational efficiency. In a previous article, we emphasize that if you’re working with dissimilar vehicle data sources and inconsistent vehicle records, how can you be confident in your valuations? You can’t. A lack of clarity in the underlying vehicle data will hinder certainty in your portfolio’s value.
DataOne steps in to fulfill this gap in OEM/insurer relationships and the challenge of working with unaligned vehicle data. The data received directly from vehicle manufacturers is incorporated into DataOne’s VIN decoding toolsets. Beyond making this OEM Build Data available to auto insurance customers (from the OEM directly or through independently verified research), DataOne also uses a proprietary curation method to standardize and normalize build data into a unified format. All the OEM-specific language is parsed to identify the underlying equipment and feature sets, which are then assigned a unique DataOne ID, resulting in a quick, simple lookup for front-end and back-end insurance professionals. The resulting solution has the ability to act as a “skeleton key” for OEM build data, unlocking its potential to empower product managers, actuaries, and underwriters to have total confidence in their risk modeling processes.
If you would like to learn more about DataOne’s OEM build data management packages, you can visit our webpage, or speak to one of our product experts.
