“Organizations need a clear understanding of current trends, challenges and available technology solutions to be able to deploy or improve an insurance claims fraud program.”
-Kent Lefner, 2012
Insurance claims fraud is big business — and it’s getting bigger. Fraud rates are increasing by 19 percent each year, according to the National Insurance Crime Bureau (NICB), and contributing roughly $336 billion per year globally to insurance claim costs, according to a 2010 Capgemini study. With more people resorting to fraudulent activities, companies are continually looking for ways to reduce the impact of fraud on their businesses.
Insurance companies’ action has been tepid but maturing with an evolution toward interconnecting operational systems and leveraging technology to automate workflow-supported analysis and prediction across lines of business. This movement will further enable insurance companies to rapidly detect and respond to fraudulent claims and save millions of dollars annually. With that in mind, organizations need a clear understanding of current trends, challenges, available technology solutions, and key areas to focus on to be able to deploy or improve an insurance claims fraud program.
The Fraud Landscape
U.S. consumers generally regard insurance fraud as acceptable. Twenty-five percent of Americans think it is “okay” to defraud insurance companies and 10 percent say they would do it if they wouldn’t get caught. Additionally, 30 percent of Americans fail to report fraud, according to the Coalition Against Insurance Fraud. Supplier attitudes are similar, with a third of physicians admitting, “It’s necessary to game the health care system to provide high-quality medical care.”
Parties that commit claims fraud are far more sophisticated than ever before. Some have involved organized crime players and even policyholders who actively share knowledge of how to succeed in insurance fraud. Perpetrators are producing a large number of attacks, more often over a longer period of time, making multiple claims aimed at hitting multiple channels and products simultaneously. Fraudsters are even leveraging knowledge of detection systems by recruiting insiders.
Current Claims Trends
Beyond implementing technologies directly designed to combat fraud, insurers are finding success in other ways. Insurance companies are now leveraging a single view of the customer with the ability to understand and analyze all of the interactions a firm has had with an individual or collection of related individuals in a household. Utilizing a single view of a customer can not only improve customer service and sales activity but also allow insurers to identify “family” related fraud matters as well. For instance, a single view of a customer means that the insurance company can, among other things, see if a customer bought more than one insurance product from the company.
If you buy car insurance from a company that is aware of all the policies you own from them, then they may recognize that you have also purchased life insurance from them or that a family member of yours did so. Without the advantage of creating a single view of the customer or the household, the ability is limited for insurance companies to quickly recognize fraud patterns associated with a single individual or family that holds multiple policies. A fraudster can easily submit fraudulent claims for multiple policies, and if insurers cannot recognize the individual across those policies, they will be slower to recognize the multiple fraudulent events.
Another marketplace trend that has been gaining traction is customers researching and buying products across multiple channels with the help of internet and mobile applications. In addition, insurers can use tools like FootPath Technology and Rapleaf to track shoppers through their cellphone to gain insight into customer habits and also spot possible fraud patterns more quickly by actively monitoring the various channels the customers use. These tracking tools allow insurance companies to understand all of the products a customer or (or a household) owns. As noted above, fraud related to one member of a household can serve as a flag for possible fraud related to the other members. These technologies can help speed up claims fraud detection.
The explosion of data is also likely to lead to important technology advances that may benefit insurance claim fraud management. The Big Data phenomenon is growing rapidly as social-networking sites, in addition to online applications, are creating an explosion of structured and unstructured data and content. Smart companies will not only invest in the intelligence gold mine found in that data but also leverage it to combat claims fraud. Investment returns can be dramatically affected by upgrades to data-management solutions; greater focus on integration with IT for the purposes of enabling “real time” information applications; and the use of text analysis, pattern recognition, and predictive analysis tools. While there are many companies focusing on creating Big Data analysis platforms, such as SAS, Oracle and LexisNexis, no market leaders have emerged to address insurance claim fraud concerns leveraging Big Data technologies. Effectively leveraging Big Data technologies is expected to yield important benefits for insurance claim fraud detection.
While the technology available to insurers is growing, implementation challenges remain a key barrier to realizing the full promise these technologies seek to fulfill. Typical challenges include:
- Disparate and disconnected policy and claims systems supporting the enterprise.
- Functional or data limitations within those systems.
- The effective use of large volumes of unstructured data.
- The inability to recognize a single view of the customer or agent and supplier across a multiline portfolio.
- Limited use of workflow automation to support detected fraud events.
These challenges often mean that insurers’ investment returns may be limited, fraud management effectiveness may be only marginally improved, and large volumes of fraudulent claims may still go undetected.
With more people committing fraudulent activities, the industry is faced with having to take more pronounced steps to combat insurance claim fraud. From a technology perspective, detection approaches and insurance systems today generally:
- Involve manual interaction within claims business processes.
- Focus on individual claims/policies instead of a large population of insureds and claimants.
- Involve simple rules using limited niche technology.
- Are siloed by line of business with no enterprise view or data sharing.
- Are primarily used to drive customer claims satisfaction first and fraud management second.
According to Gartner, claims fraud management technologies have the greatest potential for impact as a disruptive technology to the claims organization in the coming year. The adoption of Advanced Fraud Detection Solutions is still lagging, but the “agility and detective strength” they offer will likely persuade insurers to make them a higher priority. Many insurers are looking to combine fraud prediction and detection capabilities within a workflow-enabled environment. This enables insurance companies to rapidly and accurately detect fraudulent claims, prevent claims payout, and respond to fraudulent claims — saving millions annually.
The Value of Intelligence
Business Information (BI) system integrators are providing custom and system integration solutions for industry package implementation. Many of these packages focus on transaction-based profiling to track actual claim behavior, neural networks to simulate customer behavior and recognize fraudulent intent, and data mining to mine all data sources to identify relationships between known and possible fraud activities. Technology solution providers are working hard to catch up to the growing demand for integrated claims management solutions.
Providing technologies to integrate disparate systems such as multiple claims systems and deploying predictive analytics combined with automated workflow enable insurance companies to effectively detect possible fraud and to respond quickly. However, not all elements of the claim life cycle are currently addressed by focused technology packages. For instance, adjuster assignment when collusion flags are present is generally not automated in the adjuster-assignment phase.
Technology vendors such as SAS, Sybase, Actimize, FICO, and ACL have produced solutions to meet various claims fraud management needs within the claims life cycle:
- SAS applies an enterprise approach to fraud, waste and abuse detection and prevention by providing a hybrid software solution. It provides predictive analytics and neural networks techniques to detect fraud before claims are paid.
- SAP Sybase IQ aims to reduce financial cost of insurance fraud, increase speed of investigation, and improve accuracy of fraud detection.
- Actimize software is able to detect, report and prevent money laundering and fraud across multiple intermediaries, policy types and data sources.
- FICO™ Insurance Fraud Manager 3.2 can proactively detect fraudulent claims.
- ACL can highlight suspicious transactions to support fraud investigations and analyze claims history to identify suspicious transactions and verify compliance with policy limits.
- Patriot Manager from Aquilan monitors, detects, tracks/manages and reports insurance-related fraud activities.
- Guidewire and Pega software packages combine fraud-pattern recognition with workflow automation to enable response speed fast enough to prevent fraudulent claims from being paid.
Focusing on Key Areas
It may be obvious that investments made to improve the effectiveness of insurance claim fraud management often involve more than fraud technologies alone. Critical success factors to effective insurance claims fraud management technology implementation include focusing on data, expecting a cleanup investment, prevention not just detection, and accessing maturity across the claims life cycle.
Focus on data. It is critical that the data-processing environment, data models, and the data itself not only have integrity but also are integrated and usable. Effective data governance can improve and prevent data integrity problems from reducing investment value. The promise of striking claim fraud gold through the mining of unstructured data has not yet been realized, but companies can advance the topic by applying specific BI technologies such as predictive analytics, sentiment analysis, and others with an emphasis on large storage and latency minimization.
Expect cleanup investment. Very few insurers that have implemented solutions to drive a “single view of the customer” have established data consistency across disparate claim processing systems. Employing effective data governance and stewardship, effectively leveraging the mountains of structured and unstructured data across the organization, or having effectively integrated systems allow for the support of fraud management solutions. Insurers seeking to improve fraud management effectiveness should establish a clear business case that seeks to clean up and improve the organization in multiple ways, not just in fraud management or claims.
Focus on prevention, not just detection. Detecting a fraudulent claim after a payout is made may allow for recovery but may also increase litigation costs. Reducing such costs comes from prevention. Combining automated workflow with proactive detection modeling, like predictive analysis, has the ability to do that. Monitoring-based fraud detection software allows insurers to identify fraudulent claims in real time. This combined with the ability to drive action through automated workflow, email or flash reports to the special investigations unit can assist in establishing faster reaction times to fraudulent events.
Honestly assess maturity across the claim life cycle. Proactivity should start with assessing the organization. A focus on data and processing maturity should ensure that fraud is not just done at the claims application level but is embedded into the entire claims process. Improving the ability to detect trends and fraud patterns at each point in the life cycle can result in a more complete fraud management environment.
Insurance claim fraud is a big part of the insurance industry, but with the proper data, technology and automated processes investments can dramatically improve management effectiveness. Though many insurance companies are starting to look at ways technology can quell fraud, few have truly tapped into the benefits to be had by leveraging technology and a workflow-enabled environment. By first understanding key trends and challenges, and then examining important focus areas and technology-enabled solutions available to them, insurers can take the necessary steps to rapidly detect insurance claim fraud, respond to fraudulent claims, and save millions annually.