View this article online: https://www.claimsjournal.com/news/national/2014/11/18/257881.htm
Insurers are inundated with varying estimates of the magnitude of fraud perpetrated against the industry, and are aware at a high level of the likely scale of effects on their businesses and therefore the bottom line.
Such estimates can be useful, but also can lead to frustration for insurers for two very different reasons. First, success achieved by counter-fraud initiatives in the industry may be unjustly perceived as diminished due to the release of new estimates which often cite a continued rise in fraud. Second, despite the overall financial magnitude of the estimates, a clear path for an insurer to fund and embark on significant new anti-fraud initiatives is still by no means guaranteed.
Insurers share common pain points initiating new counter-fraud projects, and it’s up to risk, fraud and operations leaders to break down the barriers that are restricting modernization of the processes. For some areas, change will involve awareness and a culture shift; in others it will take buy in from the C-Suite.
This article provides guidance for smoothing the path of selecting and deploying fraud detection technology for insurers who are seeking to make a change in their counter-fraud capabilities.
Today’s obstacle to senior buy in is multi-pronged. Priorities of senior stakeholders at insurers are never limited to tackling just fraud. Therefore, the coordination of multiple internal departments will be required. CFOs are concerned with impacting the bottom line, improving profitability and reducing unnecessary claims spend. Whereas, CCOs are focused on processing claims quickly, reducing management costs and ensuring minimal operational disruption, while treating customers fairly to ensure a positive claims experience. Concurrently, CEOs are invested, on a higher level, in all of the aforementioned operational processes with the additional objective of growing market share.
While it’s a multi-pronged obstacle, there is an answer– the deployment of a fraud detection and prevention solution that enhances the business as a whole and benefits senior stakeholders. When investigating best solution vendors for this, check to make sure that their existing clients are all achieving increased: profitability, customer satisfaction, accuracy of claims adjudication, and decreased claims handling time or effort.
These benefits are realized by deploying a fraud detection system that risk-assesses all claims and all associated claim parties with a high degree of accuracy. This capability not only generates reliable red flags and suspect alerts, but also green flags the lower risk claims quickly. With a high degree of confidence in the alert quality, insurers can utilize this capability for fast tracking or straight-through processing claims. The capability fits seamlessly and enhances rather than disrupts critical business processes. The decrease in claims processing time, from weeks to days or days to hours drives customer satisfaction and the CCO’s agenda. The increase in fraud detection and prevention increases fraud loss avoidance and profitability links to the CFO’s objectives. Comparing benchmark studies from vendors will provide insight into which vendors are addressing these areas the best.
An attractive Cost Benefit Analysis (CBA) is essential amid intense competition for budget. However, it is notoriously difficult to obtain accurate fraud loss figures for your individual organization. This means quantifying future benefits is difficult at best. The lack of precise future monetary benefit estimates can easily stall a new project initiation.
Raising awareness of this problem with those assessing the CBA is key. In contrast to fraud detection initiatives within banking where some forms of losses can be accurately quantified, obtaining an accurate benefit figure for insurance fraud may not be possible ahead of a project initiation. A budget holder insisting on accurate figures is in essence a conscious or unconscious decision not to go ahead with a new fraud detection initiative.
An essential step for progress may actually be budget holders taking on the risk of relaxing their dependency on obtaining accurate benefit figures, and instead accepting estimates with broad ranges and some assumptions.
To support budget holders in taking this difficult leap of faith, utilize the prior experience of other insurers. Speak with those who already benefit from the use of the system that you are considering to obtain broad estimates, and check these against costs to see if the resulting figures are viable for your organization. Expect the solution vendor to be ready and experienced in supporting you in running paper based or operational evaluation exercises to help establish broad monetary benefit estimates for you.
With broad potential benefit figures now accepted and available, the challenge of calculating total cost of ownership of the solution needs overcoming to finalize the CBA.
Expect the solution vendor to provide exact details of what is and is not being provided.
Focus on responsibilities to understand what needs purchasing from the vendor versus what internal assets will be required for successful deployment and operation. Ensure this is verified with reference clients.
Once responsibilities are established, total costs can be further understood by requesting total pricing over a multi-year period, say five years, for the software and services. To obtain further total cost certainty, consider managed service options, such as offsite hosting or first line support, where the vendor provides both price certainty and takes on what would otherwise be your responsibilities and your risks. Managed services provide you with clearer costs estimates up front and the potential to spend from operational as opposed to capital budgets.
The internal cost of obtaining data from internal and external systems may be the largest spend on the project and can be seen as cost prohibitive. Therefore, the data acquisition phase requires considerable experience and scope clarity to run successfully and without unnecessary effort or rework. Expect the solution vendor to provide substantial expertise and documentation to aid the planning, execution and troubleshooting of this phase based on their experience with other insurers. For example, data required should be almost entirely known upfront, right down to the field and enumeration level. Detailed explanations of each field’s priority, purpose and impact on the vendor’s solution should be clear. Also, simulations should be available to work out the optimal balance of effort versus reward from the data acquisition. Starting with only a loosely defined set of data requirements or non-fraud specifics, i.e., generic insurance data requirements, is fraught with risk of either data overload or rework of data acquisition at a later stage.
Internal IT costs of owning the architecture can be high, or there may be a lack of IT expertise in-house. Expect the vendor to provide detailed architectural guidance and again consider managed services including managed support and/or hosting from the vendor.
Verify that the vendor has the expertise and infrastructure readiness for you to “lease” off the shelf solutions at a price point that is attractive to what your internal costs and risks would otherwise have been.
Over the last five years the number of vendors with insurance fraud detection offerings has increased while differentiation between vendors appears to have decreased. Offshore providers are also looking to break into this market.
A method to accelerate the analysis phase is to focus on what has already been successfully provided by the vendor to other insurers, not on what could be provided by a vendor. This approach can quickly reveal the primary advantages and features that benefit insurers from each vendor, rather than the unlimited art of the possible. Seeking a solution that is largely or entirely pre-packaged, i.e., pre-configured and used successfully by other insurers significantly decreases the analysis paralysis. Eliminating the many elements of the unknown or yet-to-be-defined can rapidly clear the table to identify what is tangible today.
Confirmation of the suitable solution for your organization is outlined in the following three steps. First, ask the vendor which of the modules available are being utilized today by other insurers. Second, obtain feedback from industry analysts who have experience regarding what insurers are utilizing vendors for. Third, meet with current clients to confirm exactly what features and benefits they feel are key to their success. If competition with other insurers becomes a block to dialogue, then look to find an insurer in a non-competitive geographical market or other line of business who is using the solution.
The power of analytics based fraud detection systems is not fully realized by many insurers today. This is in part due to the aforementioned obstacles, but also due to a lack of awareness of capabilities. For example, recognition of just how effectively some of these systems operate today and how a deployment can streamline processes and enhance rather than replace human talent is missing from the industry.
Well-designed analytics based fraud detection systems identify unusual and suspicious behaviors hidden in the vast quantities of data insurance companies collect. They augment detection with advanced, automatic entity and social network analysis. Social network analysis in this context is not about tracking people and businesses on social media sites.
It’s about automatically detecting links and associations within the data held by insurers in their policy, claim systems and other data sources.
Analytics based systems are not meant to replace boots on the ground investigators. Instead they enable these investigators to target fraud and develop cases they couldn’t see before. The increased view into entity and network relationships advances organized insurance fraud detection and reduces false positives on opportunistic and premeditated fraud.
These systems help insurers automate what can be automated. This allows handlers to interact with customers and colleagues to ensure day-to-operations of claims adjustment and payment runs smoothly. Additionally, it brings investigators in at the right point, with the right intelligence to maximize operational efficiency in the SIU. It also means, machines can do what machines do well, collect and aggregate data so that the patterns of suspect behaviors can easily and automatically come to light. Detecting suspicious claims quickly through the use of automation before payments are issued is crucial to breaking the costly pay-and-chase cycle.
Successful analytics based systems are refined over time. This dynamic process means they are able to identify previously missed or invisible fraud at go live and then ongoing. These systems can drive continuous improvement efforts to detect complex and changing fraud. Today’s fraudsters are savvy. Organized fraud is business run by agile, profit-focused criminal groups operating for financial gain. A system able to keep pace with them is paramount to mitigating the impact of fraud.
On the flip side of the automated analytics coin, is a common perceived pain point – finding and retaining fraud analytics experts to configure and maintain the automated system. Having such a team in house is a fantastic position to be in. However, it may not be your reality today. The key for leaders seeking to implement an anti-fraud product who do not have an abundance of such expertise or this talent secured for the long term, is to identify solutions that will work effectively without the dependency of an expert analytics team. Should these resources become available then the system should certainly offer them the functionality to further tune and enhance the analytics. If not available then the vendor you select must provide pre-configured alerting that is proven to work, plus include the expertise to perform tuning before and after implementation.
While insurance fraud is an industry wide issue, how it’s best mitigated and managed is something each insurance company must, at least partly, identify for themselves. Each company needs to conduct an internal analysis to uncover specific needs and compare that to what is available in the market. Following the suggested routes for senior buy in and cost benefit analyses can help support your risk, fraud and operations leaders in uncovering the best counter-fraud solution that fits your particular sector and financial objectives.
Peter Cates specializes in BAE Systems Applied Intelligence’s NetReveal counter-fraud solution. He works closely with insurers to understand individual fraud detection priorities and those of the wider industry in order to support effective deployments.