By Stuart Rose, Global Insurance Marketing Director, SAS
Big Data is one of the hottest business and tech buzz terms these days. While the term is surrounded by much hype, and the phenomenon of ever increasing data is not new, the real question is how to extract value from the mountains of data available. This is where Big Data and high-performance analytics meet.
Big Data presents both a challenge and opportunity to businesses. IDC estimates that by 2020 the amount of data created and replicated will rise to 90 zettabytes. Social interactions, mobile devices, facilities, equipment, R&D, simulations and physical infrastructure are all part of this.
Great opportunity lies in the potential insights Big Data can provide. In a recent SAS survey conducted by IDC, all Canadian financial services respondents said the ability to process and act on data in real time is important. Insurance companies can’t wait days or weeks to look at different “what-if” scenarios before making a decision. For companies to be competitive and fully serve their clients, decisions need to be made in minutes or hours, not days or weeks. But the process of creating, testing, and evaluating each analytical model before it goes into production has traditionally be very time consuming and the result is that there is often only a single validated, production-ready model produced per day, per person. High-performance analytics can help.
Defining high-performance analytics
High-performance analytics (HPA) leverages Big Data to provide deeper insights in shorter reporting windows than previous technologies. HPA enables customers to prepare, explore and model multiple scenarios using data volumes never before possible so they can analyze all the data available not just a subset, and it provides much faster processing for complex analytical algorithms. This means faster results that provide a more complete view of the situation for better fact-based decision making.
For insurers, the adoption of HPA can mean cost savings, increased revenues, lower expenses/losses, improved forecasting and accurate decision making.
Before HPA many insurance companies would analyze samples of data, which meant not all the information available was included in the results. With HPA insurers can run the analysis on all their data, enabling them to create robust, precise analysis. Additionally, they can incorporate external data to supplement the results such as third party data sources like as Google maps, GPS, credit scoring, social media etc.
Furthermore, while insurers currently use a handful of variables to support segmentation and pricing, the increased processing speed of HPA enables a larger number of variables for “what if” analysis. This enables them to find the variable that will have the most positive impact on profitability.
What business issues in insurance can HPA support?
With earthquakes in Japan and New Zealand, floods in Thailand and Australia and tornadoes in America, 2011 was the costliest year on record for natural disasters, and with hurricane Sandy causing massive damages across its path, 2012 has also been costly for insurers. For insurers, these events can have a significant impact on their financial stability. Carriers need to evaluate their loss exposure and financial position to meet liquidity requirements, often in a real-time environment. But this can prove challenging or impossible due to restrictions within their existing IT and analytical environments.
Many insurance companies may be well equipped to manage the potential losses associated with claims from individual fires and automobile accidents as actuaries can predict future losses based on the wealth of historical data associated with such losses. On the other hand, as catastrophic events are relatively infrequent there is a scarcity of historical loss data, which make it virtually impossible to reliably estimate potential future catastrophe losses using standard actuarial techniques. High-performance analytics can help improve these crucial cat modeling systems by delivering near real-time analysis as data is received as well as through the ability to run robust ‘what if’ scenarios.
Fraudulent activities are on the rise. Unfortunately, if the fraudulent behavior is not discovered quickly it may never be detected by the insurer. High-performance analytics enables insurers to analyze the data both within their organization to detect unusual behavior along with external data such as social media to increase the likelihood of detecting fraudulent activities prior to the claim being settled.
A significant benefit of HPA is that it enables insurers to dramatically reduce the analytical lifecycle as rather than spending weeks or months developing models that take days to run, HPA enables users to run many iterations in a matter of hours or minutes. This enables insurers to react more quickly to ever-evolving fraudulent activities performed by organized crime syndicates. Ultimately, this is going to change the way analytics are typically performed as we move toward a more agile analytics environment.
Today’s customer has more ways than ever to interact with your company. As interactions move from in-person to digital channels, you not only have to react more quickly, you also have to be able to predict future behavior. With faster analytics you’ll be able to detect changes in customer behavior in near real-time during digital interactions. This means you’ll be able to deliver better customer experiences and make relevant, real-time offers with higher acceptance probabilities. This will also mean your predictive modeling results won’t just get delivered more quickly, but you will also be able to identify the best future action to take considering both financial and organizational constraints thanks to advanced optimization techniques. The result? You’ll have the best opportunity to grow revenue at the lowest cost, for a higher return on investment.
Ratemaking and Price optimization
Analytics is already widely employed by insurers in ratemaking and product pricing, but unfortunately, actuaries often use only a subset of historical data to run pricing models because using complete data sets to do so is too time-consuming. Insurers are now turning to high-performance analytics since the technology enables them to analyze the entire data set, and to do so in less time than previous technologies took to process incomplete samples.
Telematics is poised to become one of the most-used technologies in insurance as, according to ABI Research, the number of user will increase from less than 2 million in 2010, to nearly 90 million in 2017.
As many insurers are already struggling to analyze existing data, keeping up with the massive amount of data created by these in-car recorders may prove too difficult for some. However, with high-performance analytics insurers will be able to analyze billions of records of data in a fraction of the time required by traditional computing environments.
In short, high-performance analytics can speed up the iterative business process of data acquisition, data analysis, variable selection, modeling and model assessment. This results in the ability to run analytical models faster – in seconds or minutes rather than hours, days or weeks – and draw on all your data rather than just a sample. So ask yourself, “What would I do with more hours in my day?” The answer may just revolutionize how you do business.
About the Author
Stuart Rose is global insurance marketing director at Cary, N.C.-based SAS. Rose, a 25-year veteran of the insurance industry, began his career as an actuary. He has worked for a global insurance carrier in both its life and property divisions and has worked for several software vendors, where he was responsible for marketing, product management, and application development.
Stuart is a regular contributor to insurance publications and the Analytic Insurer blog, as well as co-author of the book Executive Guide to Solvency II. He frequently speaks at insurance conferences and will deliver a presentation on Big Data at the 2013 Insurance-Canada.ca Technology Conference.