“The goal is to turn data into information, and information into insight.” ~ Carly Fiorina
Businesses built on a strategy of risk management should investigate the potential benefits of big data principles, technologies, and analytic methods (collectively, big data). Data analysis, including applied predictive models, is only as good as the data analyzed. A long-standing challenge, however, has been the acquisition and analysis of useful data. There are many reasons this challenge exists ranging from a lack of system integration to a constrained reliance on structured data stored in relational databases.
Traditionally, organizations have been served by disparate systems that separate data by function to support operational processes. When data sits in such silos, its value is diminished, accessible only to departmental applications. Analysis in this context paints only part of the picture. Separate analyses performed against disparate data sets must be assembled and summarized in an attempt to develop informed inferences. But, can disparate data that results in equally compartmentalized analyses render a whole that is more valuable than the sum of its parts? Not likely. And, to imply a level of precision that can’t be known is not only unproductive, in the world of risk management it is downright dangerous.
Certainly, departmental systems are often integrated as part of a data warehouse effort. Big data extracts otherwise hidden value; it offers an opportunity to infuse valuable external data (e.g., census data, social media data, behavioral data) into existing data stores and technology capable of processing massive amounts of structured and unstructured data in near real-time. The prospects for enhanced risk management are boundless.
Injecting new types of data into traditional data sets will uncover myriad and previously hidden correlations, many suggestive of behavioral tendencies that span a population. Such enhanced analysis adds depth to any risk management strategy by offering greater insight into why people do what they do and what they are likely to do in various scenarios.
According to IBM, 90% of all the data that exists today was created in the past two years. It stands to reason that data aggregation, access, and analysis demands a flexible, evolutionary approach to remain relevant.
Actuarial and other types of risk analysis are nothing new, but it’s limited to traditionally accessible data. Big data principles, technologies, and analysis provide great opportunity. First, useful, previously inaccessible data can add value to any risk analysis strategy, offering enhanced perspectives on a population or portfolio risk profile. Second, big data is predicated, in large part, on the addition and processing of new structured and unstructured data. The result is a solution that will evolve and support the infusion of new data into existing data sets as it becomes available. Embrace that we don’t know what we don’t know and we therefore must implement solutions that expand, evolve, and flex to remain relevant.