Around the globe, insurance claims fraud each year cost the industry an estimated $3.4 trillion—in the US alone, that is between $30 and $40 billion annually. The high cost of fraud, considered in the context of increasingly frequent natural disasters, the declining profitability of many insurers, and ballooning capital and other requirements imposed by regulators, only adds to the pressure on insurers to tighten their operations, and to plug every hole from which efficiency and profitability might leak.
An increasingly hostile environment
In the US, it is almost an item of faith that, after tax evasion, insurance fraud is the crime most frequently committed by Americans. While a sluggish economy, recessionary in many countries, makes fraud more tempting, the internet has made fraudulent claims both simpler and more rewarding for perpetrators, who range in size and sophistication from opportunistic individuals to organized “fraud rings.”
Insurers estimate that 10% to 25% of reported claims contain some element of fraud. That quotient includes everything from exaggerated injuries to inflated damage, staged accidents, and deliberately phony medical bills. Yet the industry has been slow to react to this growing attack on its bottom line. In a survey of several hundred European insurers, for example, while nearly 70% said that application fraud was a “key priority,” only 22% said they believed they had a program in place equipped to successfully fight this kind of fraudulent activity. Still worse, in an average year the typical insurer’s Special Investigation Unit (SIU) investigates only a paltry 1% of all annual claims.
For companies across the spectrum, the consequences are dire, and include eroding profits, limits to their ability to offer competitive premiums, declines in loss and combined ratios, and worse. Compounding these negatives is the constant pressure to control operating expenses, and, at the same time, the difficulty in identifying the actual costs of different types of fraud. Furthermore, the mandate to maintain excellent customer service for honest clients can in certain ways act as a counterweight to the need to identify fraudulent claims.
There are other embedded institutional challenges that insurers face as well. Analytic systems, for instance, may be out of date, or poorly integrated, so that neither SIU staff members nor executives can get an effective picture of key applicant and claims data related to fraud investigations specifically or the claims process generally. Legacy detection systems may reduce the effectiveness of SIUs by producing too many false positives. Too often, SIUs lack sufficient resources, such as skilled managers capable of leading well-trained investigators, and suffer high rates of staff turnover.
Addressing the problems
Insurers can beef up fraud detection, processing, and mitigation programs with an increasingly sophisticated arsenal of management and analytic tools. The primary focus of these tools is on increasing the technical and operational sophistication and efficiency of the SIU. Today’s state-of-the-art advanced data analytics offer companies operational and predictive capabilities that can deliver dramatic, measurable results. SIU case-tracker systems have become more accurate, easier to manage, and can handle a higher volume of cases while minimizing false positives; they can be used to prioritize SIU case loads more effectively. Not only can advanced predictive analytics be customized and integrated to break down institutionalized, data-based silos within a company, they can also be fine-tuned to distinguish between fraud that is opportunistic and individual, and that of the “fraud ring” variety. Moreover, effective analytics can optimize Guidewire systems, for example, thereby sharply reducing false positives in fraud scoring. Up-to-date data analytics can target a range of public records, such as ISO claim search and sanctioned doctors, to red-flag potential fraud in a timely manner. It is now possible as well, by identifying personal emotional reactions to basic questions, to use digital voice analysis as a tool to help identify fraud.
P&C insurers, most notably, have developed an impressive array of analytics, which today are spreading throughout the global insurance industry. Claims history analytics look at frequency, type, and overall claim amounts to determine the probable legitimacy of submitted claims. First notification of loss (FNOL) analytics leverage a catalogue of known suspicious loss indicators to predict and separate out likely fraudulent claims. Claims origination software, on the other hand, can scan sources of mishaps that trigger claims to determine how likely an incident is to be intentional or accidental, while enhanced billing analytics can help separate genuine claims from those that are inflated. Social network and social media analytics focus on claimants’ digital footprints to identify patterns of fraud, whereas geospatial analysis can usefully circumscribe a disaster’s impact area in order to isolate and understand claims filed from beyond that zone.
This is a lot of new technology, requiring too much management change for many companies to handle. Indeed, those that try will need an extremely dedicated chief executive, a fully committed C-Suite level team, and a carefully planned transformation program extending throughout the enterprise. Buy-in must be carefully planned and organized, with roles and responsibilities at all levels clearly communicated. Sufficient funding will be critical.
At the same time, many companies find that the stresses of multilevel institutional and operational enhancement of this kind are significantly smoothed by working with an experienced, proven partner. Such a partner can offer diagnostic and planning services to identify key inefficiencies or operational soft spots, and then recommend, plan, and implement solutions. This outsourcing allows a client to choose from a catalogue of customized service offerings, from end-to-end, enterprise-wide programs to carefully focused efforts that upgrade only one, or a handful, of operations. Outsourcing certain SIU functionality, for instance, often helps more objectively identify in-house problems and roadblocks, simultaneously strengthening the impact of the unit’s anti-fraud efforts and freeing up staff members to take on higher priority cases. Frequently for insurers the flexibility and efficiency that a knowledgeable partner offers translates to double the number of claims investigated annually, while also enabling them to more consistently and accurately identify those cases that are the most potentially damaging. Genpact’s own collaborations, for example, have cut operating costs by 25% or more, reduced false positives in fraud identifications by half, and dramatically improved fraud detection rates.
Results and benefits
A successful anti-fraud upgrade will show a range of meaningful, measurable results. An insurer should expect improved fraud detection rates, reduced false flagging, an increase in the SIU’s impact ratio and cycle time, and a significant improvement in the overall effectiveness of SIU operations. A successful program should lower operational expenses in a variety of ways, including cutting claim payment severity, making better use of investigative resources and customer service staff, improving employee retention rates, cutting down the time to review scored claims, and improving processing efficiencies.
These enhancements, which improve margins, enable highly competitive rate design, and upgrade customer service, also act as a deterrent (and surprise) to organized fraudsters. Indeed, one downstream benefit of implementing a highly effective anti-fraud program is that rip-off artists, sensing a fight they cannot win, shift their energies to easier marks: those vulnerable companies, with less effective, up-to-date, and well-designed fraud prevention programs.