For the better part of the last decade, digitization and data have transformed insurance, remaking and streamlining everything in the value chain from risk assessment to claims handling to advisory services. But even the myriad ways the sector has been using its valuable data assets up until now represents just the tip of the iceberg when it comes to big data’s overall potential to remake the industry (we have been collecting plenty, but monetizing with far less frequency). Moreover, as the GDPR in Europe continues to strengthen its data protection and privacy regulations, insurers will need to work thoughtfully to adhere to legal guidelines while utilizing their substantial data assets in novel ways.
Unstructured data: treasure trove or hornet’s nest?
To date, structured data represents the foundation of the industry’s most actionable insights -- claims history, age, dates, geolocation, and so on. Structured data has historically been relatively easy to collect, organize and analyze, while unstructured and semi-structured data (i.e. texts, voicemails, social media posts, video content, and even wearable technology) have been more challenging. While unstructured and semi-structured data represent a high-value area of analytics that’s still largely unmined by the insurers, there remain questions about how to ensure its ethical application.
In health insurance, creating fair policies remains the backbone of risk sharing and the definition of solidarity — where the healthy support those who need care. But in other areas of insurance that are not compulsory, risk assessment can be quite different.Sibylle Fischer, Strategic Venturing & Startup Scouting
Data-driven personalization versus privacy?
Customer-centricity remains a business imperative today and data-driven personalization is a core element to those products and user experiences that put the customer first. At the heart of personalization are customer analytics -- data that allows for enhanced service based on analyzing past user preference, purchasing and other behavior patterns. That data is then used in a variety of ways: to assess risk, to predict future behaviors, to recommend relevant products and services, as a few examples. These kinds of customer insights are incredibly valuable, both in terms of their ability to generate sales and as a way to build brand loyalty.
However the relationship between data-driven personalization and privacy is being increasingly questioned, especially as the data available becomes more multi-dimensional, drawn from many different areas of an individual’s life. In health insurance, creating fair policies remains the backbone of risk sharing and the definition of solidarity — where the healthy support those who need care. But in other areas of insurance that are not compulsory, risk assessment can be quite different.
Big data plus big oversight equals incremental change
Data-driven personalization is powerful technology, but if used incorrectly it could call into question a range of ethical issues, from privacy concerns to outright discrimination. That means even as technology and innovation move at warp speed, in the name of risk-fairness and equity insurers will need to take a measured approach to introducing new big data initiatives, with programs and products carefully vetted, mapped and implemented incrementally. Part of this strategy will also likely mean developing new AI, tools and methods for analyzing data in a way that can identify bias. In many ways, the discussion about ethics and data is only getting started, yet it is likely to remain a central issue in the evolution of the industry. Moving forward, insurers will continue to look at how to positively balance equity issues with making the most of their tremendous data resources -- and for now there’s still plenty of room to grow.