Michelle Bergeron has been lauded for her notable strides as anexpert sleuth and innovator while at Esurance. In fact, shehas become somewhat of a celebrity in the field since beinginterviewed by The Wall Street Journal and working closelywith the NICB and other agencies.

Bergeron recently garnered yet another distinction at the16^th annual ACE conference in sunny Las Vegas: the2012 Claims Professional of the Year. Here, she speaks withClaims about her win and designing an innovative programfrom the ground up.

What was your initial reaction to winning this year'saward?

I was very shocked and just very humbled. My boss, Bob Cline,joked that it was like I was on Candid Camera. Even now I feel likeit is very surreal that what I started almost four years ago hasmade an impact on our process and helped others set up systems oftheir own.

What precipitated the start of the SIU analyst programat Esurance? Did your background in criminal justice help prepareyou for the challenge?

Initially, I was told of the job opportunity as a chance tobuild a program from the ground up from a colleague. I washired by Esurance in 2008. My then-boss understood the needto implement an analytics program at Esurance. The idea of havingan opportunity to try and build a better mouse trap at a very lowoverhead cost really intrigued me. Things are much more streamlinedin the online world, but it also presents some unique opportunitiesfor fraud that are different from your traditional agent-basedcarrier or even a phone sales model. I always like achallenge, and this was certainly one that pushed my limits andexpanded my knowledge base.

I have always loved technology for as long as I canremember. Throughout high school and college I worked as acertified nurses aide (CNA) and initially I got a paralegaldegree. I worked in an office where we handled everythingfrom court-appointed criminal cases to personal injury suits. When I finished my Bachelors degree in Sociology (Crime, Law &Deviance) I was looking for something where I could really apply myknowledge and experience. I happened to be selling insurancefor a major carrier at the time I graduated and I worked my waythrough the software/systems department over to the fraudunit. I think all of the things I have in my background serveme well in my current capacity.

There doesn't seem to be any 'one path' to fraud analytics. Ifyou have a natural inclination toward problem solving and a driveto investigate things, then that is a good start. Technologyis a requirement in this area of fighting fraud so technical skillsand computer literacy are a requirement. Many people say that SIUsounds like such a cool career and ask how they could get hired. Myadvice is to just get your foot in the door at an insurance carrierand keep working toward your goal.

How did you design such an innovative process from theground up?

I think the first thing that really helped me to formulate aplan was to understand our business. Every carrier has adifferent process that begins with sales all the way through to theclaims process. I always go through the process from start tofinish myself and look for any weaknesses or areas that could beexploited. It is important to understand changes to not onlythe claims process over time, but also sales and service. Anychange on the way a policy is obtained or a claim is reported canhave a ripple effect for our fraud unit. After I felt I had agood understanding of our process I reached out to other analystsin the industry to be a sounding board for my analytics design.

The combined experience of a competent peer group is a powerfulresource. We had all seen what has worked well and things thatwere not as successful in our careers. I used the informationand suggestions to help narrow down tools and processes that wouldgive us the “most bang for our buck.” Because we were arelatively small carrier, we needed impactful, low-cost solutionsthat could be maintained without monopolizing limited ITresources.

How were you able to get the buy-in of others atEsurance?

My first boss explained that we would need to start with a verymodest budget and show the returns from the program. Within thefirst six months, we were able to find some very large, organizedrings attacking our company. The successful identification of fraudthat had existed undetected allowed us to expand our platform andadd additional tools and resources. Additionally, Bob Clinewas hired as our National SIU Manager, and he is a big supporter ofour analytics program. We would not be where we are today withoutBob's assistance and the support of our senior management.

What results has your organization realized since theprogram's inception?

Organized fraud detection has been very impactful to our claimsunit. Identifying many suspect claims within the first 48hours allows us to begin investigating before payments have beenmade and to work with the adjusters to formulate a plan of actionthat is proactive instead of reactive. In the past, I haveseen claims sent for investigation when they are months old or evenwith almost all the damages already paid. We have much betterresults when we can get the investigation going in the beginning ofthe claim. Fraud is ever changing and so are our program andanalytics. As soon as we close down one area of fraud, theperpetrators will find another weakness to exploit. We cannever be content with what exists today and we must always beevolving.

How has the program evolved in these pastyears?

The fact that we have been able to add additional tools over thelast few years has really bolstered our capabilities. I seeour primary purpose as early fraud detection and assisting theinvestigators in developing their cases to a successfulconclusion. Some of the analytical tools we have are verypowerful and can compile extensive amounts of information in amatter of minutes or hours. That same information compiledmanually could take hours or days for an investigator.

One of the most significant enhancements we have made is theself-service data access we were able to build for theinvestigators. I worked as a solo act for the first threeyears in this area. I did everything from build the systemsto tactical case work. If it needed to be done I found a wayto fit it in to the program. A large volume of requests werealways about “if this vehicle was insured by us,” or “how much havewe paid to this business/provider,” and many others. Therequests were very easy to handle, but there are 40+ investigatorsasking those questions every day and only one person to field allthose requests. I knew I could free up a lot of my time formore complex work if I could find a way to allow the investigatorsto retrieve their own data.

We found an easy-to-use platform to navigate our variousinformation silos called “Intellishare,” which I set out to buildspecifically to support this need. The entire project wasbased on feedback I collected from the investigators about theirneeds and then I balanced that with speed and performance. They can now search millions of records within seconds and get animmediate response. Additionally, the system showsinterconnected data, which wasn't something that they sawbefore. Now they find me organized cases to work, which is atotal role reversal.

Fraud, especially organized rings, continue to plaguethe industry. What represents our best chance at combating theproblem and deterring future abuse?

The only way I can ever imagine a deterrent effect is to havemuch tougher legal penalties for insurance fraud. Until therisk outweighs the potential reward, we will always have to bevigilant and work to combat fraud. Additionally, moreresources need to be devoted to the prosecution of these types ofoffenses. This type of crime doesn't get a lot of mainstreampress and also lacks specific funding from most states to filecriminal cases for fraudsters. I think the public would be farmore supportive if they could actually understand how much money istaken out of their pocket to cover the cost of insurance fraud.

What specifically do SIU and claims organizations needto be doing? Are carriers sharing data and trends with eachother?

Each SIU and claims organization has a unique model andstructure. Because of the differences in processes, I don'tthink you can have a one-size-fits-all approach. At alllevels, associates who handle policies and claims need to betrained on red-flag indicators and they need to be vigilant.Underwriting, policy service organizations, and claims need to allpartner with their SIUs to sharpen the saw and develop bestpractices. The one thing we can all do is proactively investigatefraud and communicate with other carriers. I find it discouragingthat not all carriers in the industry are willing to shareinformation and some may even have a mindset that it's better toshift fraud to other carriers by sending the fraud “down thestreet.”

I don't feel that attitude truly serves any of us in the longrun. That carrier may temporarily get rid of the immediate threat,but all these groups cycle back and forth between carriers and evengeographies. So then you just pass the buck and sooner orlater they will be back to at your doorstep. If we take acollective approach then we can work to have people arrested ortogether make it so difficult that the fraudsters find it is toomuch trouble and the risks of getting caught are too high. I dofind that most carriers are very proactive and try to shareinformation where appropriate to assist in indentifyingthreats. Immunity laws vary by state so we have to understandwhat is acceptable in the venue. We also call upon NICB tohelp us exchange information between member companies. Themore of our information and trends we can aggregate, the lessopportunities the fraudsters have to perpetuate their schemes.