Insurance databases with very large amounts of data are slowly becoming diverse but still cannot be characterized as highly volatile. These databases are ‘big’ indeed but cannot be classified as Big Data. What’s next?
The 4V model
Big Data, in principle, is not a technology but an idea that reflects some new directions in the approach to data management, storage and analysis.
In 2001 META Group (now Gartner) published a report titled “3D Data Management: Controlling Data Volume, Velocity and Variety” where a model for Big Data was described as 3V, then supplemented with another component to form the 4V model:
- large amount of data (Volume)
- variability in the data (Velocity)
- large variety of data (Variety)
- important data value (Value)
The insurance business has always been based on information collection, processing and evaluation. Along with the use of information technology, we were given new possibilities to collect even more information. Even though insurance databases containing very large amounts of data are slowly becoming diverse, they cannot be characterized as highly volatile. We can process such data using traditional analytical tools. Although these databases are ‘big’ indeed, they cannot be classified as Big Data. Yet.
Big Data in insurance will have the greatest impact on actuarial calculations, claim handling, underwriting, analysis of customer needs and prevention of insurance frauds.
Reaching for new data sources
Undoubtedly, Big Data will be used in insurance one day. This will occur when insurance companies reach for completely new data sources, such as social media, telematics and Internet of Things. Then, the existing analytical tools will no longer be able to cope with growing data volumes. This data, of new quality and dynamics, will be the Big Data – and some new specialized tools will be needed to process it. New infrastructure capable of meeting the requirements of the 4V model shall come in handy too.
Insurers got interested in Big Data about two years ago. According to analyst companies, Big Data in insurance will have the greatest impact on actuarial calculations, claim handling, underwriting, analysis of customer needs and prevention of insurance frauds.
Financial sector specialists surveyed in “Big Data+” by Computerworld and IBM also pointed at improving marketing campaigns and better profiling of product portfolios as potential benefits of the Big Data use.
Mariusz Janczewski, Senior Insurance Consultant, Comarch