Insurers have been battling against rising insurance fraud - the FBI estimates fraudulent claims total more than $40 billion per year in the US alone, and cost the average US family an additional $400 to $700 per year in the form of increased premiums. However, the cost of fraud is not only the direct cost of the fraud itself, but also the indirect cost of customers lost– discouraged by slow responses and frustrating processes. In fact, according to a survey conducted by the accountancy firm EY, 87 per cent of policy holders have stated their claims experience significantly impacts their decision whether or not to remain with their current insurance provider.
For smaller insurers the cost of fraud is compounded. Budget and resourcing constraints, data challenges and accelerated growth combine to leave smaller insurers particularly exposed to fraudsters. The solution lies in fraud prevention; however, when it comes to investing, how can insurers gauge the ROI of a fraud detection solution?

Miss the data, miss the money

For small to medium size insurers that process circa 150,000 claims annually, a random spot check to manually review the processes and decisions for each claim can provide an indicative view of the cost of fraud. Using a manual review process, it is estimated that 5 per cent to 10 per cent of claims are, in fact, fraudulent. If the average cost per insurance claim is around $10,000 USD, just the 5 per cent of claims that are fraudulent could cost the said insurer $75 million USD. While this approach requires minimal investment, it begs the question as to how many fraudulent claims are missed, not to mention factoring in costs incurred throughout the claims process.
Additionally, with in-person checks unlikely to make a comeback, the industry is becoming ever more reliant on data to approve or deny claims. If insurers continue to rely on manual processes in the new data epoch, they risk potentially huge increases in missed fraud as well as missed opportunities for improving customer retention through smoother on-boarding and claims processing. To grow and thrive, insurers need to harness their data through intelligent analytics and machine learning – understanding their customers, good and bad, and measuring fraud for an intelligent view on risk, opportunity and ROI.

Calculating the ROI of insurance fraud prevention

Identifying the volume of fraud impacting a single business can be tough when using manual based methods. We know that the answer lies in data, machine learning and analytics, but for any solution to make sense, it needs to show significant ROI.
Through our work with large and small insurers across the world, we’ve identified a formula for making this assessment and calculating the savings per converted claim in a complimentary report - Calculate the Return on Investment (ROI) of Insurance fraud solutions.
The industry average for cost savings is around £6:£1, meaning that for every £6 saved, an insurer would expect to spend £1, and so when evaluating the ROI of an insurance fraud solution, the key is looking at how to increase that ratio.

Key considerations for insurance fraud detection models

There are two key considerations to keep in mind when transforming fraud investigation processes to improve ROI - balancing checks with the customer journey, and the role of intelligent technologies in improving speed and accuracy.
As we look to the future, the digital shift presents both a large opportunity and extensive potential risk. For smaller insurers the impact of the digital era can be defining, and successfully navigating promises around data solutions, cost and benefit requires a considered approach. The question of ROI is both more complex and more straightforward as our ability to measure, track and assess moves from manual obligation to competitive edge. The answer lies in knowing the right questions to ask.