(Image credit: Adobe Stock.)
Analytics is a hot topic; everyone is talking about it. In a recent survey by Capgemini, 78 percent of insurance executives interviewed cited big data analytics as the disruptive force that will have the biggest impact on the insurance industry. But analytics is not new to the insurance industry. In fact, one could argued that the first mortality tables were a form of analytics, since the actuaries were using historical data to forecast the survival rates of their policyholders and insured. But compared to many other industries, insurance companies are seen as laggards when it comes to analytics. Why?
One reason is that until now, analytics has predominately been seen as a back-office function, used by silo departments, such as claims, actuarial and marketing. To become analytically efficient, insurance companies need to create a strategic culture where data and analytics is part of the corporate DNA.
Bigger is often better, and today’s big data environment is no exception. Many associate big data with the Four V’s: volume, velocity, variety and veracity of data. However, the ways big data is changing the insurance industry can be broken down into three categories:
- Unstructured data. The insurance industry has always been heavily paper-based, whether it is adjuster notes, police reports, medical records and underwriter information. Until recently it has been difficult for insurance companies to analyze that data. With new technologies like text mining and sentiment analytics, insurers are beginning to gain useful information from this previously inaccessible data.
- External data. In the past, insurance companies have focused on their own internal data. The data created and stored in their transactional systems such as policy administration, claims management, billing and agency management. Today insurance companies are supplementing that data with external third-party data, including credit scoring, government demographic data, and even geospatial data such as weather information and Google maps.
- New data. This is data that wasn’t even imagined 10 years ago. Social media data, telematics data—information from in-car data recorders to monitor driving behavior and now the potential from the Internet of Things.
But the real challenge begins when companies begin extracting meaningful insights from this explosion of data. Determining how to take advantage of all this data to price better, expand markets and improve the business of underwriting risk and handling claims. Fortunately, the science of extracting insight from data is constantly evolving. Below are several analytical techniques that can help insurers get value from data to make strategic decisions faster:
Data visualization, where information is presented in a pictorial or graphical format, is helping insurance professionals see things that were not obvious to them before. Insurance companies analyze historical data—which includes information from policy administration solutions, underwriting applications and billing systems—to forecast and predict future losses.
A picture is worth a thousand words, especially when you are trying to understand and gain insights from data. It is particularly relevant when you are trying to find relationships among thousands or even millions of variables and determine their relative importance.
Data visualization is an art and a science unto itself, and there are many graphical techniques that can help insurance executives better understand their data. However, one of the biggest challenges for non-technical and business users is deciding which visual should be used to represent the data accurately. Auto-charting determines the most appropriate visualization by understanding the data and its composition, what information insurers are trying to convey visually and how viewers process visual information.
Imagine if power insurers could harness the insights hidden within that vast sea of structured and unstructured data.
Predictive modeling or data mining goes beyond reporting on what has happened to discovering why it has happened and what is likely to happen next. Data mining consists of multiple modeling techniques such as decision trees, neural networks and clustering combined with advanced regression and statistical routines to deliver to more accurate models. Typically, predictive modeling is performed offline as a back-office function by actuaries and other statisticians. More than ever, insurance companies need to integrate analytics into their transactional systems to improve operational efficiency
From a technology aspect, link analysis converts data into a set of interconnected linked objects that can be visualized as a network of effects. From a business perspective, it enables marketers to analyze social networks and identify relationships among customers, and then use that information for more accurate profiling and segmentation. The explosive growth of social media has greatly strengthened the power that the opinions of friends and family have on a customer’s buying decisions. Link analysis visualization tools enable insurers to see a complete picture of individual customers, their products, transactions and networks at the click of a button.
Imagine if insurers could harness the insights hidden within that vast sea of words. Medical records, emails, underwriter notes, call center logs, video, even Twitter and blogs are all examples of unstructured data. By most estimates, at least 80 percent of the information available in an organization is actually unstructured data, and that percentage is likely to increase with the growth of social media. But for insurance companies, it might just as well be called invisible data. For an industry that is driven by data, text analytics is still new to most insurance companies.
Text analytics is the use of computer software to annotate and extract information from electronic text sources and analyze that information for business purposes. Text analytics includes sentiment analysis, which automatically locates and extracts sentiment from online materials; and text mining, which provides powerful ways to explore unstructured data to discover previously unknown concepts and patterns.
Data has always played a critical role in insurance. The insurance business is built on analyzing data to understand and evaluate risk. There is no doubt that big data will affect the insurance industry. But data on its own will not transform the industry. Analytics provides insurers with operational insights into their business, but this is still not enough. Only when insurers raise analytics to the enterprise strategic level, making decisions based on the insights, will they break through to the real potential of big data.