Many experts agree that identifying fraud indicators early in the process is a key to successfully resolving cases. "The sooner you can begin the process of weeding out and collecting information, the better. That's true of any kind of investigation," says Doug Ashbridge, director of special investigations at Farmers Insurance (Los Angeles). "There's a fair percentage of claims that people identify as suspicious up front that aren't [actually fraudulent] at all. The faster you can get those issues resolved and get the claim taken care of on behalf of the customer, the better."
Dennis Parker, insurance marketing manager at SPSS, SPSS, a Chicago-based predictive analytics provider, says it is "absolutely vital" to identify a claim as potentially fraudulent within the first one or two weeks after it is filed. That's because most fraudulent claims that are defeated are not denied by the insurer, they're dropped by the insured. Fraudulent claims are more likely to be dropped earlier in the process because the perpetrator has less invested in the outcome.
"It's more psychological than anything else," Parker says. "If I can identify [fraud] and get the claimant into the office for an interview within the first seven days, then the perpetrator and his cohorts don't have a lot of 'skin in the game,' so to speak," Parker explains.
On the flip side, with claims that aren't identified until two to three months into the process, perpetrators may have already put a lot of time and effort into the fabrication process. They also may have a lot more to lose by dropping the claim.
"Whoever is involved in that claim is likely to have acquired bills and bought things with the expectation that this money is coming in. So they're a lot less likely to walk away from the claim," Parker says.
In his previous profession at an insurer as a fraud manager, Parker says he would measure his investigators by how well they reduced the number of days between the date of loss and the date of referral to the SIU. Dollar savings doubled, he estimates, by reducing the average number of days by two weeks.
"It was remarkable how much more money [was saved] and how many more claims were solved -- and found to be ... fraudulent or not fraudulent -- by using an objectives approach and forcing the investigators to go out and help adjusters find more fraud," Parker says.
Data Mining Improves Fraud Mitigation Efforts
Data mining technologies can help insurers like Farmers and Nationwide access and leverage the institutional knowledge vital to fraud mitigation efforts that is locked inside their current and historical claims data.