Insurance - a Risky Business

How the insurance industry is using big data to change the way we evaluate risk


The convergence of data analytics and the insurance industry seemed meant to be. The insurance industry uses multitudes of information to assess risk and formulate policy premiums. Data analytics utilizes information from enormous data sets to find business insights and streamline operations. So it’s no surprise that analytics has seamlessly integrated into the world of insurance, helping companies assess risk in new, more accurate ways than ever before.

Insurance companies always needed data. But the effect of analytics into the process has been twofold: a broader comparative scope of data, allowing for a clearer picture of risk; and also, a focus on behavior-based information. Previously, insurance policies would be formulated by comparing a customer’s history with those of other customers, providing a simple, if less than accurate assessment of risk. Nowadays, companies are taking advantage of continuous, real-time data, fed through complex algorithms that paint a much more complex, and much more accurate picture. Here’s a look at the massive strides being taken in different areas of the insurance industry, garnered by the integration of data analytics.

Auto Insurance

Auto insurance companies are some of the most competitive in the industry. Customers are used to shopping around, using price comparison services and changing companies at will. This competition pushes insurers to assess risk as accurately as possible, to offer qualified drivers low-priced premiums that also make a profit.

A highly accurate, expansive and immediate way of assessing drivers’ risk is in real-time behavioral data. Many companies have turned to telemetry-based packages, which use a smartphone app, or a specially designed device installed in their car to measure drivers’ behaviors, providing more information than ever before. Using predictive modelling, and comparing this data with thousands of others in a database – an algorithm reveals the likelihood of the driver being involved in an accident or having their car stolen.

However, many drivers feel that it is an invasion of privacy to release personal data to their insurance company. In a recent study by Insurance Nexus, an insurance analytical group, customers were asked if they would share personal driving data, if offered an app that located free parking spaces. 46% of respondents reported they would not feel comfortable releasing information about their driving habits. One customer in particular cited their poor driving as a reason, thinking their rates would go up if the insurance company actually knew how they drove.

Property Insurance

As home buying rates have gradually recovered from the housing market crash of 2006/7, the need for Homeowners insurance has increased considerably. And with this increase, the rate of fraud has risen correspondingly. The Coalition Against Insurance Fraud estimates that fraudulent activity in Homeowners insurance costs approximately $80 billion a year, which makes an optimized analytical solution a top priority for insurers.  

Carriers have been utilizing point-of-sale predictive analytics to try and single out potentially problematic customers. Data scientists build complex algorithms encompassing client’s prior claim histories, insurance credit scores, public protection classifications, catastrophe models, even aerial and weather imagery of the property to create a much clearer forecast of the policy needed and risk involved. The more variables included, the more accurate a picture the insurer can create.

Find outhow Eccella helped a global reinsurance company provide portfolio risk assessment solution to property insurers.

Health & Life Insurance

As discussed in a previous blog, health and life insurance are also utilizing big data analytics to more accurately assess risk. Due to the ever-increasing popularity of health apps and wearable trackers, insurance companies have an opportunity to obtain precise information on an individual’s behavior throughout the day. These devices can count calories, track steps taken, show how intense the user exercises, as well as measure heart rate and other valuable information. Although this data is private, insurance companies can offer services based on the use of these tools, giving customers incentive by offering lower premiums to healthier customers. The Insurance Nexus study also revealed that the average customer would request a 37.5% discount on their premium, in exchange for sharing personal data retrieved from a wearable health tracking device2.

Fraudulent Claims

Another area where data analytics is taking the forefront of the insurance industry is in fraud. According to an investigation by the FBI, each year fraudulent insurance claims cost the average US family between $400-$700 in increased premiums. Insurance companies are beginning to use analytics in the form of predictive modeling to limit fraudulent claims. Specific algorithms select variables from past fraudulent claims, and compare them to current claims. When there are matches, the likeliness of fraudulent behavior rises. These variables could be behavioral patterns, credit scores, and partners involved in the claim, such as an auto repair business.


The effect of big data analytics in the insurance industry, as it progresses in the next few years, will be staggering. The breadth and depth of accurate information will result in more transparency, as our behavior will begin to be collected at a greater rate. The upside to this is a better value being transferred to customers. And the take home message from insurers goes something like this: Your insurance rate will be cheaper, if you deserve it….and we’ll know if you do.