By embedding and automating predictive analytics across the enterprise, insurance companies acquire the ability to better understand and predict -- in real time -- their customers' future behaviors by analyzing, modeling, and scoring demographic and transactional data from operational systems, as well as attitudinal data gathered through customer feedback and surveys.

The convergence of predictive analytics, business processes, and IT architecture signals a transformation in the way insurance companies serve their customers. It marks a fundamental shift in the way value is measured, from an emphasis on products and solutions to actual experiences.

TowerGroup, a leading research and advisory services firm focused exclusively on the global financial services industry, believes predictive analytics must play an integral role in the way insurance companies now do business. In a recent report on the insurance industry, TowerGroup revealed that "data management and predictive analytics are no longer merely 'nice-to-have' technology initiatives. Carriers that fail to recognize this fact will see significant deterioration in their results, as well as plummeting loss of competitive position."

Predictive analytics empowers claim handlers, underwriters, insurance brokers, customer service representatives, and other key stakeholders across the enterprise with the predictive insight to create value from underutilized data from every channel to improve customer interactions.

With a new focus on data about people and added importance being placed on achieving customer intimacy, insurance firms put themselves on a track to becoming a predictive enterprise in which business objectives are interconnected.

With the ability to capture this vast supply of customer data -- both structured and unstructured -- insurance companies are able to truly know who their customers are and dynamically improve interactions from underwriting to claims so that they never lose sight of the customer's wants and needs.

Industry analyst firm Gartner, Inc., also recommends that insurers seriously assess their current IT investments. In her report entitled, "Hype Cycle for P&C Insurance, 2008," vice president and analyst Kimberly Harris-Ferrante said, "For success, property and casualty insurers must continue to invest in technologies to stay competitive and to differentiate themselves -- especially in claims and customer service.

"Technologies such as claim management solutions, predictive modeling solutions, GIS, and wireless claim applications will help insurers to improve these tasks and, thus, reduce costs and improve customer service quality and retention. Quality claim processes that are efficient, rapid, and better managed, including real-time fraud identification and streamlined processes, will be a critical success factor by 2010."

Taking a Bite

Fraud remains the number one concern of insurance carriers. A study released in November 2008 by the Insurance Research Council estimated that fraudulent claims and buildup added between $4.8 billion and $6.8 billion in excess payments to auto injury insurance claims that closed with payment in 2007.

That's only one example of how prevalent fraud has become. Common forms of this illegal activity can also be seen in health care, property and life insurance claims, among others, and include "padding" or inflating actual claims, misrepresenting facts on an insurance application, submitting claims for injuries or damage that never occurred, and staging accidents.

Insurance carriers are doing their part to not only uncover existing instances of fraud, waste, and abuse, but also prevent future occurrences. They are integrating predictive analytics into their business rules based on industry best practices to capture key insights that can then be used to detect and minimize fraudulent claims -- in real time -- at each stage of the claim lifecycle.

A claim handler can not only extract value from standard yes/no answers with a customer, but also receive detailed information and sentiments behind those answers to instantly determine those claims that qualify for immediate approval and those that are flagged as suspicious and require follow up.

Alabama-based Infinity Property & Casualty Corporation (IPACC), a provider of personal automobile insurance with an emphasis on non-standard auto insurance, deployed predictive analytics software and solutions to reduce its payments on fraudulent claims, while improving its ability to collect payments from other insurance companies.

Until predictive analytics was deployed, the identification of potentially fraudulent claims had previously been the responsibility of claim adjusters who had varying degrees of training and used inconsistent practices. As a result, data related to suspicious claims was typically not gathered rapidly or fully detailed. That was a major obstacle, as speed of investigation and early gathering of key data are extremely important for claim investigators.

Predictive analytics significantly improved effectiveness of that process by automating the workflows and data gathering related to fraudulent and subrogated claims. With the technology now driving this critical step for the adjusters, IPAAC is able to quickly spot suspicious claims and forward them to investigators, who can then begin their investigations within days of the original claim being filed. They also have more accurate data with which to work.

Optimizing Claim Processing

It's estimated that claim management costs can be as high as 20 percent of an insurance firm's operational expenses.

Without risk-aware customer profiles that include predicted loss ratios and fraud-risk scores, insurance carriers not only expose themselves to such losses from current customers, but also the potential for acquiring new high-risk customers.

Predictive analytics addresses those concerns by combining data from acquisition campaigns with proactive claim risk optimization to filter out prospects that match predetermined claim risk profiles.

When new claims enter the system -- from any channel -- they are immediately analyzed against risk profiles and either approved for processing or flagged for investigation. With this additional step, the productivity and accuracy of the entire claim-handling process is improved, from first notification of loss to settlement.

In essence, the claim process incorporates the same predictive logic and automatic procedures used to identify possible fraudulent claims. Predictive analytics pinpoints those claims that are most likely legitimate and should be processed quickly. This process is automatic, so claim handlers don't even realize that they are using predictive analytic technology.

ALKA, one of the top five insurance companies in Denmark, relies on predictive analytics to fight fraud and enhance customer service. The technology determines during the claim notification process whether an incoming claim qualifies for immediate approval while providing the call center agent with the appropriate action to take.

This ability increases customer satisfaction levels through faster processing of payments and improved subrogation results by enabling an insurance provider to focus on claims that are more likely to be paid back. It also ensures that honest and loyal customers are consistently presented with the best prices.

Becoming a Predictive Enterprise

Just how long the economy will remain in recessionary mode is anyone's guess. What is important is that the insurance industry acknowledges the impact it is having and continues to seek out ways to make the best use of current resources. That means embracing the convergence of predictive analytics, business processes, and IT architecture. Doing so means that insurance organizations have begun their journeys to becoming a predictive enterprise where processes are intertwined and personalized relationships with millions of customers are created.