As fraud schemes continue to mutate and grow in complexity, theypresent irksome puzzles for even the most talented andknowledgeable SIUs and law enforcement officials. In order toidentify sophisticated, large-scale rings and isolated incidentsthat siphon precious resources from P&C insurers and consumersalike, the average claims professional or investigator must employa combination of technology and analytical savvy, pairingcutting-edge software tools with imagination and tenacity.

Today, Michelle Bergeron, SIU Analytics Program Manager atEsurance and 2009 recipient of IASIU's “Analyst of the Year” award,advised attendees at the IASIU 26th annual seminar about how tofine-tune their approaches by properly analyzing policy data andsales system information in the workshop titled, “Policy Data:Analyze That.”

The presentation incorporated a meld of lecture and livedemonstrations via multiple software tools readily available toanalysts. Bergeron kicked off the discussion by sharing techniquesused to unravel real pervasive fraud schemes, supplying anecdotalevidence of what works and what may be necessary beyond your modusoperandi to aggressively target fraud.

“I've developed a system of taking characteristics of all newpolicies and analyzing them on a monthly basis,” she explained.“For instance, I gather data for all policies purchased and thencheck back historically to spot similarities or trends. Let's sayyou notice that a car just added to policy was on two otherpolicies before (and not registered to the same person). That isquestionable. It begs further probing.”

Bergeron added that organized activity has increased insophistication and that analysts must be attuned to even minutedetails while ramping up common strategies. “Those committinginsurance fraud are using three email addresses now, burner phones,mail drop addresses at UPS stores, all of which can changefrequently,” she said, emphasizing the importance of looking forpatterns, however subtle they may be.

“We had a group operating in a New York city burrough that wouldobtain policies mostly with email addresses such as Juan 123 atyahoo; maria 123 at yahoo. Although the emails varied, theypatterned the same. So we set up a filter for policies establishedin the area incorporating the '123' pattern. We could then look tosee if policies fit other criteria, such as liability only coverageon an older vehicle. Then we checked the policy for claims, say, 20days out later on and worked from there.”

A savvy analyst must look beyond ISO and examine the wealth ofdata gathered internally at one's claims organization orunderwriting department. That may very well involve shattering silos between the two and even merging disparatesystems to present a unified front against fraudsters. If thatsounds intimidating, then Bergeron assures you the extra effort andproactivity will be justified and very likely rewarded with adecline in fraudulent claims and resultant payouts.

“Sure, I hear from people who experience issues related tosilos,” Bergeron said. “Because [Esurance] is smaller and sotech-friendly, we have been able to adopt a holistic approach andencourage communication across all departments. The more you cancultivate that rapport—that free flow of information—the better offyour SIU and organization will be.”

Other topics of discussion included identifying patterns, theusing specific reports, as well as other techniques for insurancecompany claims professionals and investigators at virtually everylevel. Bergeron did, however, tailor a healthy portion of thepresentation to assisting supervisors and managers.

Esurance's national director joined Bergeron to round out thesession with an exploration of the company's homegrown solutions,offering advice about prosecuting fraudsters and positioning yourorganization for success in fraud deterrence.