Analytics continues to be a hot topic, but as with other successful advances, people are asking what's next. Consider the iPad: after being released to great applause, an outcry about what should be included in the next version soon followed. Analytics is no different and to those who are asking what's next, the answer is high-performance analytics -- analyzing greater amount of data, faster.
The amount of data we have to analyze is expanding exponentially. Social media, sentiment data, blogs, sensor data, transactional data, third-party data, other big data sources ...the list goes on and on. But it's not just about the big data. Insurance companies can't wait days or weeks to look at different what-if scenarios before making a decision. Decisions need to be made in minutes or hours not days or weeks. Add to that the very time-consuming process of creating, testing and evaluating each analytical model before it goes into production often results in the delivery of only a single validated, production-ready model per day per person. High-performance analytics addresses all of these challenges.
Let's begin by answering the question "What is high-performance analytics?" It includes a set of analytical capabilities that can be executed in a highly scalable, in-memory distributed architecture. It allows customers to prepare, explore and model multiple scenarios using data volumes never before possible, and it provides much faster processing for complex analytical algorithms -- both of which deliver better answers faster to those who need them for decision making.
That's great but often the question asked is so what? What can I do with faster, better answers? Here are some ideas:
Telematics The adoption rate of analytics is dramatically increasing. According to a study by ABI Research, the number of telematics user will increase from less than 2 million in 2010, to nearly 90 million in 2017.
Insurance companies are going to be inundated with data from these in-car recorders. Many insurers are already struggling to analyze existing data, how are they going to analyze this data explosion? The answer is high-performance analytics which will enable insurers to analyze billions of records of data in a fraction of the time required by traditional computing environments.
Ratemaking and Price Optimization Today many insurers are using advanced analytical techniques such as generalized linear modeling for ratemaking and product pricing. A recent survey by Towers Watson showed that 70 percent of US Insurers are using predictive modeling for personal auto insurance. However actuaries have often relied on using a subset of historical data to run pricing models since it is too time-consuming to prepare the data and run the models. To combat these problems, insurers are now turning to high-performance analytics to provide faster processing on the growing volumes of available data.
Customer Intelligence As customer interactions in insurance moves from in-person to digital channels, you not only have to react faster; you must also be able to predict future behavior. Faster analytics means you'll be able to detect changes in customer behavior in real time during digital interactions. In turn, you'll be able to improve customer experiences and make relevant, real-time offers with higher acceptance probabilities. Faster analytics also means your predictive modeling results won't just get delivered more quickly -- because with optimization techniques, you'll be able to identify the best future action to take considering both financial and organizational constraints. This results in the best opportunity to grow revenue at the lowest cost, leading to increased ROI.
CAT Modeling Industry observers have suggested that 2011 may end up being a record year for catastrophe losses. These events can have a significant impact on the financial stability of insurers. Carriers need to evaluate their loss exposure and financial position to meet liquidity requirements, often in a real-time environment. However many are restricted from achieving this because of the limitatiosn of their existing IT and analytical environments. Last year proved how crucial CAT modeling systems can be and high-performance analytics can help.
High-performance analytics is not the answer to all insurance problems, but it can speed up the iterative business process of data acquisition, data analysis, variable selection, modeling and model assessment. As a result, reports can run faster from a matter of seconds or minutes as compared to hours and even weeks. The question becomes "How can I use those extra 92 hours to benefit my business?"
About the Author: Stuart Rose is global insurance marketing manager at Cary, N.C.-based SAS. Rose, a 20-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. He has driven successful development and implementation of enterprise systems with insurance companies in the U.S., the U.K., South Africa and Continental Europe.