Data’s rich seam of success for insurers

There’s no doubting that insurers have access to more data than they could possibly hope for. But volume doesn’t necessarily equal value. Some existing sources of data are less than reliable, while effective sources of information are unnecessarily discounted.

Insurers need to start thinking more inventively around data if they want to create truly effective risk assessments and price more accurately in an increasingly hard market.

Intelligent Insurer spoke to Monique Hesseling, managing director of insurance at Cloudera, who joins a panel of insurance experts at this year’s Commercial Lines Innovation USA Virtual Event. She will reveal how insurers can access new streams of data as well as deploy technology to overcome the limitations of partial or unreliable data from brokers and customers, establishing sound data validation practices that streamline your organization.

What new sources are most valuable?
A lot of people don’t realize that commercial lines insurers were at the forefront of using data. If we think about telematics and how that started from within fleet management, managing distance travelled or driver safety, a lot of new data sources stem from this.

New sources include geographical data, weather-based, sensor or behavioral data as well as any kind of unstructured data that used to be difficult to analyze, in addition to newer data sources such as video, picture or voice analytics.

How has the technology landscape evolved to allow this?
After the development of data lakes some years ago, the biggest development has been the shift to cloud. Reducing the costs of storing data—especially the separation of storage and computing power—and empowering government and industry to use unstructured data is all freeing insurers up to make the most of what’s available.

Our approach has always been to start by looking at strategy. What do you want to achieve? Do you want to focus on profitable growth? In that case, start by outlining the business strategy and then look at your data strategy, the operational analytics that allows everyone in the organization to execute.

Only after that can you look at the technology you need to support those strategic components. That instinctively leads to “should I go into the cloud” and “do I want best-in-breed in claims adjustment, or do I want to run on a platform?”, but these are really the last steps.

To what degree do we need to refine data?
The first question is: should we solve this problem? Data quality comes at a price. First you come up with a plan about what data and the quality you need. It doesn’t have to be 100 percent perfect in every case.

After we have set up quality requirements and guidelines, it’s vital to establish whether that data needs to be historically accurate, or whether it needs to be accurate to the new level going forward. The latter is a great deal cheaper.

Then, you set up a data quality program as part of your data governance strategy. Most data management solutions have the functionality to identify issues and even fix them. Most of our carriers automate a big part of that governance and quality process, then have data stewards—for whom it isn’t usually their day job—to cast a human eye over the process to fix escalations.

Which datasets will make pricing more accurate?
Anything that gives you more information on the behavior and status of a specific risk or exposure—driving behavior in your fleet, for example—will make your ratings more individualized and more accurate.

The questions include what is still acceptable to society and government? When do we stop looking at your specific risk and start treating you as part of a segment?

We could potentially run into questions of anti-selection and fairness. It could be more of an issue in personal lines, but you still run into this in areas such as workers’ compensation.

How should carriers go about refreshing their data strategy and getting leadership support?
Start with the business strategy then look to the three to five use cases that could support that strategy. Looking at profitable growth in marine insurance for example, the three to five use cases which could be subsets of the coverage include the type of cargo, the part of the world it’s travelling through, the voyage in real time, and so on.

Pick a tangible example where the business is likely to see immediate results, because if your case is just theoretical, it’s less likely to get support.

Monique Hesseling will be speaking at Intelligent Insurer’s  Commercial Lines Innovation USA Virtual Event (May 18 to 20). The event is free to attend for insurers and brokers/agents, but you must register in advance. Sign up to access the content live and on demand here.