Fraud is a growing problem, particularly in the fields of motor and P&C, and fighting fraud has considerable cost and profit potential for insurers. Insurers should act now to capture this potential.
Insurers in Europe have experienced fraud on an increasing scale in their claims processing over recent years. Insurance Europe, the European (re)insurance federation, estimates that the total from all cases of fraud – both detected and undetected – amounts to 10 percent of overall claims expenditure in Europe. In the UK, the association ABI UK Insurance regularly publishes savings from fraud management, and cites the sum of GBP 1.1 billion in 2012 as the highest figure ever recorded (ABI UK Insurance Facts 2013). Italy’s politicians have already picked up on the topic and are demanding countermeasures from the insurance industry because insurance fraud – both unorganized and professional – is one of the reasons why Italy’s motor insurance premiums have risen to the highest levels in the EU. In Germany, too, insurance fraud has risen to a degree that causes great concern. The German Insurance Association GDV estimates that one in ten claims reported can be put down to insurance fraud, generating overall losses of EUR 4 billion. Insurance fraud is widely considered to be a peccadillo – many of the perpetrators do not even have a bad conscience.
In view of these developments, insurers need to act swiftly and decisively with specific counteractions to realize the still untapped potential from optimized fraud management as well as to reduce a potential competitive disadvantage compared to other insurers. This is not just about providing protection for honest insurance customers, but also to fend off the danger of anti-selection by fraudsters. If fraud is not fought more effectively across the board, fraudsters will turn their attention to companies that do not defend themselves so well. In some European countries such as the UK or Spain, joint initiatives of the entire sector against industry fraud are under way. An industry-wide database has been established in the UK, for instance, with comprehensive data sets on identified fraudsters and a dedicated department at Scotland Yard, financed by the insurance industry.
The European insurance industry has a way to go on fighting fraud
Many European insurers do have fraud management systems in place. Compared with the demands of professional fraud prevention and protection, however, the industry is largely still in its starting blocks. It definitely cannot match the sophistication of those it is fighting, some of whom are highly professional.
McKinsey conducted a benchmarking initiative at the end of 2013/beginning of 2014 in Germany and Austria to determine the status of fraud management in the insurance industry. McKinsey has used these results to identify the common pitfalls in the fraud management capabilities of European insurers, which fall into five broad categories as follows:
- Not in the focus of top management. Many insurers view fraud management more as a specialist than a top management topic. They also often assume that the share of potential fraudsters is below average due to careful underwriting and application checking processes at their own company. This often leads to high tolerance in the event of a claim, even if statements are clearly contradictory. In most countries, it is the exception rather than the rule for fraud management to be positioned as a top management topic, and for the level of tolerance to be defined as zero in the interests of customers who are honest.
- Limited importance of fraud in operational claims processing. Usually a clear focus on the topic of fraud is also lacking in individual claims organizations, and staff barely exchange notes. Alongside tools for the automatic recognition of fraud, systematic manual recognition via checklists, fraud manuals or the like has proven especially effective, and can be implemented fast. However, claims handlers responsible for manual fraud recognition are often under heavy pressure from the high efficiency demands they come under from their claims departments. The principle of having one claims file processed by several clerks is often the norm, too. Both techniques make it harder to effectively recognize and fight fraud. Even comparatively simple aids such as checklists or lists of guidelines can lead to considerably better results without increasing the workload.
- Insufficient specialization. Suspicious cases are sometimes processed by fraud specialists who largely work as normal claims handlers. Having dual roles as fraud investigator and claims handler leads to more limited specialization and insufficient focus on fraud management – particularly if the claims division is very busy and a backlog needs processing. It is international best practice to use dedicated fraud specialists with specific skills (such as former police detectives).
- Hardly any modern investigation methods. The success of fraud investigators does not just depend on specialization, but also on the choice of investigative methods used. Cognitive interviews are one example. These specialized interview techniques are already being used to determine the credibility of claims reports with great economic success in the UK and southern Europe, but are rarely used elsewhere in Europe. The honesty of the subject being questioned can be evaluated in cognitive interviews using sophisticated interrogation techniques and psychological principles.
- IT systems that are obsolete or have not been maintained. Most insurance companies use IT systems for recognizing cases where fraud might be suspected. But often they are antiquated and not regularly maintained or managed, delivering comparatively poor recognition rates and quality as a result. In contrast to the UK and US, an added issue in some European countries with strict data protection regulations is that detection systems in those countries generally do not use smart IT systems with network analyses that enable the identification of bands of fraudsters beyond individual cases.
Qualitative disadvantages as outlined above are typically reflected in quantitative results. Good practice standards indicate that the ambition level for savings from insurance fraud in Europe should typically be >=3 percent of claims expenditure. Optimized fraud management is thus one of the fastest and greatest levers in claims management.
To capture this potential, each insurer should begin with their own Claims Processing as well as related divisions, optimizing the detection and handling of fraud cases. In a second stage, it is also a good idea to make sure the company’s fraud management systematically includes protection from organized fraud and bands of fraudsters, and to make every effort to cooperate with other insurers to fight fraud.
First stage: Putting fraud management on a more professional footing
McKinsey has developed a framework enabling the systematic optimization of fraud management while conducting industry-wide cooperation with several major insurers. This framework is based on a clearly defined fraud strategy (zero tolerance of any proven fraud to protect honest customers, for example), and uses seven levers, both for core fraud management processes (detection, investigation, prevention) as well as the related support systems (performance management, IT systems, skill building and organization) (Exhibit 1).
The following examples highlight the actions open to an insurer for improving their fraud management:
- Optimize detection. Checklists and/or fraud manuals help claims handlers to recognize claims cases that are suspicious manually. Practical experience shows that this can significantly increase the detection rate without any IT investment. Claims handlers, even within one company, have differing skills and success in detecting cases where fraud is suspected. A structured checklist or fraud manual can greatly contribute to learning from each other and achieving a more homogenous and higher detection rate overall within the group.
- Introduce a triage function. The point here is to introduce a staged fraud management process using a triage (initial case assessment and decision on resource allocation) as the central decision-making point. The function uses the applicable fraud strategy and (usually) a cost-benefit analysis to decide whether an investigation should be made on a case where fraud is suspected. Alongside this, the triage specialist has three further tasks. One is to gather information on the cases and assign them to the fraud specialists ready for investigation, depending on their segmentation/specialization and current workload. Second, the triage specialist is responsible for the continuous optimization of fraud detection using information and coaching of the claims handlers. In the course of this he/she also ensures ongoing improvement of the guidelines and indicators of any existing IT-supported recognition. Third, he/she is responsible for KPI-based measurement of fraud management success using case figures and the savings achieved along the fraud funnel (Exhibit 2).
- Use extended investigation methods. These methods might for example be cognitive interviews to assess the honesty of the interviewee. This involves analyzing the consistency of replies to varying or repeated questions on identical points. The evaluation also includes memory of details and the liveliness of the description.
Depending on their size, available capacity and existing specialization in the company, insurers may compile a tailored program for optimizing fraud management from these and further levers. It is not just important to select suitable components and understand them in detail (asking questions such as which fraud indicators should be included in the manual checklist). What matters is to implement the program decisively and swiftly. Experience from related McKinsey projects has revealed a number of success factors:
- Fraud management needs to be on the top management agenda. Optimizing fraud management requires clear commitment of the board to tackle this topic and to initiate the corresponding cultural change at the company above and beyond the claims organization.
- Clearly define fraud and how to handle it. The fraud strategy should clearly determine how the company defines fraud, and the consequences it is prepared to bear. After all, the topic of fraud management extends beyond the claims organization to cover other divisions such as Sales and Operations. It is therefore important to agree the strategy with the departments affected. When should a case be categorized as fraud? From what point onwards should no further goodwill be offered? When should a contract be terminated, and when should a police report be filed? These criteria need to be communicated clearly and consistently, both internally and to customers.
- First build up processes, KPIs and roles, rather than immediately making major IT investments. IT generally provides great added value in fraud management, particularly in fraud detection. But it should build on carefully defined processes and KPIs, and definitely not be on a critical path. Waiting to start optimizing fraud management until IT solutions are ready is neither technically necessary nor economically advisable because many manual initiatives initially require barely any IT support. It has therefore often proven worthwhile to first establish or optimize the processes, KPIs and roles and then use these to derive the IT system requirements.
- Pilot early in parallel to design. The new approach should ideally be tested at an early stage in parallel to fine-tuning the design in a pilot project. This parallel testing of the new approaches can provide valuable company-specific experience for future fraud management in an iterative process that can be directly incorporated in the refinement of the concept.
- Invest in skills/specialization and the number of fraud specialists. Fraud management has great economic potential, but this can only be tapped if investments are made in building up skills and sufficient fraud specialists. The focus should particularly be on effectiveness and not unilaterally on efficiency. It is true that efficiency is a key topic especially in the back office, but the corresponding use of resources is more than justified in view of the economic value of effective fraud management.
- Pursue continuous improvement. Fraud management is a highly dynamic topic. The fraudsters – who are often highly professional – quickly adjust their approach to the behavior of insurers. Companies therefore need to counter this by improving their entire process to ensure that they are equally swift to react, from investigating and detecting fraud through to penalizing those involved. A key factor in this is the constant exchange between triage and fraud specialists as well as claims handlers, giving each other feedback and recognizing changes early.
- Optimize beyond the claims division. Optimizing fraud management is largely but not exclusively a matter for the claims division. This should be coordinated with Underwriting/Products, Operations and Sales to install rigorous fraud prevention and overarching consequence management (including the termination of contracts).
From the viewpoint of staff particularly in the claims division, fraud protection is a topic with positive connotations. They know from their day-to-day work how often fraud now occurs, are watching developments with concern, and would welcome better opportunities to fight it – especially if this does not increase their workload. After years of focus on efficiency in the claims division (which continues to be essential), fraud management is the opportunity to tackle a topic with positive connotations while leveraging the exchange within the claims organization that this inevitably involves to improve claims processes beyond fraud management.
Second stage: Rallying forces to fight organized fraud
Optimizing fraud management is not a one-time project but a continuous task requiring constant further development and improvement because it is essential to keep up with the methods and tricks of the culprits. The approaches already described relating to internal claims processes are the first step in terms of countermeasures. They largely refer to opportunistic fraud, i.e., individual customers attempting to make additional profit from an insurance case, and occasionally to professional fraud. It is however scarcely possible to detect bands of fraudsters and organized crime with instruments of this kind aimed at individual cases.
This is where the second stage for optimizing fraud management begins. This involves using other tools, such as automatic IT recognition based on network analyses, scrutinizing telephone numbers, bank accounts and addresses to identify any cause for suspicion. The methods deployed also have to change: a special team is needed to fight organized fraud, connected to the police and courts. Such solutions require more patience as well as significant investments in staff and IT support, while still remaining under the direct influence of the individual insurer.
Alongside the insurers themselves, the industry as a whole also needs to tackle fraudsters in the market across the entire country or even across Europe. Setting up a joint hotline for anonymous tipsters is one of several possibilities. It is ultimately in the interests of all market participants to take decisive action against fraud, protecting honest customers from having to fork out for fraud, or even becoming victims themselves.
The value proposition “safety” includes protecting the insurance community against the abuse of premium funds. This does not mean seeing every customer as a potential fraudster before being proven innocent. But the more insurance fraud increases, the more urgent countermeasures become for insurers. It is generally possible to achieve more effective fraud protection with limited additional resources. However, it is important to elevate the profile of fraud management from a functional as well as an economic perspective via targeted investments in staff and IT support, as well as cross-company cooperation.
This article was excerpted and translated from the German report Performance 2014, published in September 2014.