Another parallel between insurance and tractors, he said, is that tractors are remarkably slow and nobody wants to be stuck behind one while out on the road. So, while the importance of tractors is well understood, it doesn’t mean people want their journeys delayed by them. This is much the same with fraud detection. MGAs, brokers, insurers and ultimately consumers, rightly expect robust fraud detection without the claims journey being unnecessarily delayed.
“Like a slow road journey, a slow or cumbersome claims journey is a real turnoff,” he said, “and it’s a barrier to [firms] doing business. But, of course, [the claims journey] is incredibly important and it’s a regulatory requirement. Therefore, none of us should forget about it or ignore it in the pursuit of a great customer experience. It’s certainly not something we should forget about in terms of profitability, or ultimately the endgame of keeping insurance premiums competitive for all of our benefits.”
Read more: Sedgwick boosts claims fraud savings
So, how can firms tackle the problem of catching the vast majority of the bad guys, without end customers being stuck behind that proverbial tractor of a delayed journey? During the MGAA webinar, Carman touched on a variety of conversational touchpoints – including the role that counterfraud technology can play in the financial success of a firm. The insurance industry often thinks about fraud as a claims problem, he said, but enriching underwriting data and investing time in a fraud strategy actually directly translates into underwriting profitability.
Digging deeper into what a strong fraud strategy looks like and why it is so important, Carman highlighted the tougher market conditions currently being faced across the broader insurance piece. The most effective way of navigating a hard market is by demonstrating strong and effective controls, he said, and through partnering with a TPA that takes responsibility for driving that fraud performance. That TPA must place equal importance on people, data and processes – it’s a three-legged stool and if you remove one of those metrics from good fraud detection then it falls over.
“Some TPAs talk a good game about fraud but when you scratch beneath the surface, in terms of what actually exists in terms of counterfraud controls, they may have a counterfraud team or access to a fraud specialist to refer fraud back to the carrier for handling. But they’re at the end of the process,” he said. “And sometimes we as an industry [can be] a bit guilty of this – we think about fraud as the fraud investigators.
“Actually, it’s two steps behind that. It’s about the handlers, it’s about the adjusters and good identification. And that’s the role that a TPA plays. And when we examine TPAs closely, most are still using traditional key fraud indicators and handler training on fraud awareness as a primary means of identifying fraud. And in 2021, there’s got to be a better way of doing things.”
At the foundation of strong fraud performance is meaningful action, ownership and accountability, Carman said – when Sedgwick engages with clients, what comes through loud and clear is a simple message: “we don’t want a fraud tractor, don’t slow down the process and don’t involve our customers in the process.” He noted that he agrees wholeheartedly with this but emphasised the role technology plays in achieving that demand.
In 2020, Sedgwick set about challenging its own thinking around detection performance and whether its detection strategy represented more a “fraud tractor or a fraud Ferrari”. The firm had already recognised its strong counterfraud function, Carman said, which is great but it is at the end of the process so the real question was rather how strong it was at detecting fraud in its TPA or commercial home or motor operations.
“I think we recognised that we could be better at identifying fraud risk for our clients and we saw an opportunity to take our performance from good to great in the identification piece,” he said. “Technology plays a pivotal role in countering organised and opportunistic fraudsters. In simple terms, the very best technology can compute vast quantities of data and spot the anomalies in a way that humans can’t.”
An individual claims handler who is dealing with maybe 10 cases a day is not going to be able to spot the connections or anomalies between the case they’re dealing with and another they handled 18 months ago from a different policyholder, with a different address on a different claim. It’s simply not possible as humans to trawl through the 1,000s of bits of information we receive every day, he said, and derive meaningful insights from that. So, that is where technology steps in.
“So the claims handler can spot that from a behavioural point of view, there’s an issue here,” Carman said. “But that Big Data perspective and technology [that has] the ability to compute vast quantities of structured and unstructured data, to apply rules, machine learning principles and fraud detection algorithms are crucial weapons in 2021 for detection. Without that, you’re sort of scrambling around in the dark for the bits that you can’t see, and will never get to.”