Making Sense of Big Data in Insurance – ACORD & MarkLogic
Article Synopsis :
Most insurers struggle to optimize the data they have, and along comes Big Data to complicate things. The research paper “Making Sense of Big Data in Insurance” by ACORD and MarkLogic, demystifies some of the hype around Big Data and provides a pragmatic review of the technologies available, their strengths and weaknesses, and some examples of Big Data technologies in use in the insurance industry today.
Relational databases are still the most efficient way to handle highly structured tabular data, especially in the context of Online Transaction Processing. Companies should look to benefit from new technologies associated with Big Data where they can provide value beyond the core transaction processing associated with policy administration and claims management.
The recommended adaption strategy when it comes to Big Data is a hybrid approach, utilizing both relational and non-relational technologies. The paper outlines hybrid use cases in the following core functional areas:
- Fraud Detection: Moving from a claim-centric to a person-centric approach, incorporating technologies beyond the firewall.
- Customer Experience and Insight: Incorporating unstructured data such as customer on-boarding documentation, customer-care call logs, clickstream data, auto telematics and geo-spatial data, to strengthen customer affinity.
- Claims Management: ACORD standard data messages provide structured data from first notification of loss onwards, a process which can be supplemented with unstructured text and multimedia information to accelerate decision-making and save time and (mainframe) processing costs for the carrier.
- Underwriting: In the Reinsurance and Large Commercial sectors, risk information is received in a variety of formats (e.g., Excel spreadsheets, Word documents, JPEGs, PDFs). Non-relational technologies enable insurers to store and access this data for analysis to highlight anomalies, patterns, and red flags.
The paper concludes with a useful primer, written more for the CEO than the CIO, on the technical aspects of Big Data, including the main benefits of Hadoop and NoSql databases.
Link to Full Article:: click here
Digital Insurer's CommentsGoogling ‘Big Data in Insurance’ returns 150-million hits. Yet various recent surveys from InformationWeek, Novarica, and Gartner tell us less than half of insurers are actually investing in Big Data technologies, and most of those are doing so on an experimental basis.
This report from ACORD and MarkLogic does an excellent job cutting through the hype to explore how insurers are actually using unstructured data and flat-file analytics to improve business performance. No doubt, the upside looks promising.
We agree with the authors that, in the near term at least, the value of new Big Data tools resides in their ability to help insurers extract increased value from existing sources of internal data. More evolution than revolution, it seems.
No doubt, over the long term, Big Data possesses revolutionary potential.
Link to Source:: click here