The insurance industry has always made use of data to makebetter decisions. The underwriting department of aninsurance company has been at the forefront of data use for manyyears – using historical claims data as well as factors such asdriver age, gender, credit score, and so on toproject potential for loss, and make riskdetermination. 

Now, many insurers are making significant investments inpredictive analytics. According to the IVANS 2011 CarrierAutomation Trends study released September 2011, 37 percent ofcarriers are currently using or plan to integrate predictivemodeling and business intelligence into their organizations in thenext 12 months, with the goal of greater consistency and accuracyin business decisions in less time.^[i] 

A survey conducted by Towers Watson in October-November 2011specifically asked insurance carriers about the bottom- andtop-line benefits of incorporating predictive models into rating,pricing, underwriting, and risk selectionprocesses. Seventy-five percent of those surveyedidentified bottom-line benefits of rate accuracy, loss ratioimprovement, and improved profitability.  Over one-thirdof those surveyed also identified top-line benefits of 'expansionof underwriting appetite', 'improved renewal retention,' andincreased market share.[ii] 

How Can Analytics Help in Claims?

The average consumer today has an auto accident once every sevento 10 years, well behind the average vehicle trade-in cycle ofevery five years.  Due to the nature of the product beingsold, auto insurers typically have very limited interaction withconsumers outside of policy issue, bill time, and sometimes throughother services such as banking However, the real moment of truthfor the insurer is at the time of an accident, making claim'shandling one of the most significant opportunities for an insurerto retain or lose a customer. 

The ability for insurance carriers to advance their use ofanalytics to improve the customer experience is becoming a must,particularly as consumers demand that the claims experience be assatisfactory as with any other interaction they have with companiesthey frequent.

Many claims organizations have made good use of analytics tohelp track performance and improve collaboration with businesspartners, including:  

• The use of electronic appraisal reviews and shared guidelineshave helped to ensure business partners have the information neededto fulfill work in a transparent, compliant, and completemanner.
• The use of management dashboards have helped facilitate claimsperformance reviews in a concise, targeted manner, enablingmanagers to address specific areas of performance, adjust levers,and evaluate the impact in real time. 

These tools are necessary and valuable, but they're the tip ofthe iceberg. Advancements in technology have enabled the next waveof analytical tools that lead to more actionable information andless reliance on human interpretation of data. Today's tools alsowork to more directly enhance the customer experience. 

What's the Problem?

Two big challenges remain in today's claims process: slow totalloss identification and inefficient workflows, beginning right atfirst notice of loss (FNOL).

To identify total losses, insurance carriers have historicallyused a series of questions, or a 'decision-tree' method todetermine whether a vehicle is repairable or a total loss at FNOL.This is problematic as many total losses aren't identified at FNOLand are advanced through the claims process – inspections areconducted and estimates written – before a formal vehicle valuationis requested and, only then are a majority of total lossesidentified. These extra steps can increase claims-related expenses,including salvage, tow and rental; increase cycle time, andnegatively impact customer satisfaction.

Another challenge for insurance carriers is efficient andcost-effective claims routing. Currently, there is no systematicway to quickly assess and project the value of a claim at FNOL. Forexample, to send a staff appraiser to inspect a vehicle out in thefield may not be a cost-effective approach if the vehicle repaircost is $1000 or less.

Claims Gets Proactive

CCC has developed a predictive model that builds on work itbegan over three decades ago to create electronic processes forinsurance claims professionals and their business partners. CCCONE™ Predictive Solutions is an analytics tool that recommendsrouting damaged vehicles by applying estimated vehicle repair costdata, vehicle market values and estimated salvage costs usingguidelines established by insurance carriers. Within minutes, FNOLrepresentatives using the CCC solution can route totaled vehiclesfor salvage, while repairable claims are returned withrecommendations for routing to the appropriate appraisalsource.

CCC's automotive insurance claims domain expertise uniquelypositions the company to deliver the CCC ONE PredictiveSolution. 

CCC ONE™ Predictive Solutions – Early Results areIn

Early results from a series of insurance carrier pilots showbenefits within two key areas: early total loss identification andcost-efficient routing of claims to optimal appraisal sources. Withtotal losses, insurance carriers in the pilots have seen, onaverage, a 50 percent improvement in their ability to determinewhen a claim will exceed their established total loss thresholds,while staying within a range of 1 to 4 percent of falsepositives.  This helps insurance carriers to more quicklynotify their customers of total losses and realize savings intowing and salvage costs.

CCC ONE Predictive Solutions pilots improved insurance carriers'ability to predict low-value claims (based on their own guidelines)by 80 percent, improving their ability to route the claim to theappropriate appraisal source.

To learn more about CCC ONE Predictive Solutions, please visit:www.cccis.com

[i] Underwriting, Consumer Portals Among Top Technologies forCarrier Investment: IVANS Survey."  Property Casualty 360,September 22, 2011.

[ii] Predictive Modeling, Proving its Worth Among P&CInsurers. Towers Watson, February 2012.