The customer may be king, but insurers can't afford totreat every customer like one. Knowing who to retain, who to targetfor up-selling and who deserves royal treatment are goals behindtoday's customer analytic solutions.

While all these goals are critical, retention is still the topreason insurers look to customer analytics, and with goodreason.

“One of the things we've learned through analyzing our resultsis that customers who have stayed with us longest over the yearshave also been more profitable,” says Michael Petrarca, assistantvice president at Amica Mutual Insurance Co. “We've used analyticsto move our retention even higher.”

Mark Breading, partner, Strategy Meets Action, observes,“Because retention is a driver of profit, carriers look toanalytics first to determine when customers are at risk of lettinga policy lapse or defecting to another insurer. Determining when apolicyholder has a propensity to leave and what actions to take toprevent it are important.”

Because the insurance process has relatively few customer touchpoints compared to other financial-services industries, insurersneed to maximize every opportunity to build relationships. However,that doesn't mean every contact is the right time to pursue anygiven customer objective, and that's where analytics come intoplay.

“Companies often offer things at the wrong time,” observes KarenPauli, research director in the Insurance Practice at CEBTowerGroup. “They offer an extension or new limit when someonecomes to them inbound at a point of distress. You need to haveanalytics focused on life-event changes and other points wherecustomers would indicate they would be open hearing somethingnew.”

She also believes that insurers should use customer analytics totarget service optimization. “Carriers finally get the fact thatthey can't give absolutely the same level of service to everyone,”she says. “Companies already have different levels of coverage anddifferent products they offer to different customers. They shouldprovide extra services to customers that are the most profitable ordesirable.”

Ask the Right Questions

The divergent goals of customer analytics demonstrate thatinsurers need to be asking the right questions before slicing anddicing customer data.

When Nationwide launched its customer analytics initiative in2007, it began with an effort to better understand the customerlifecycle based on insights from J.D. Power and a detailed analysisof how customers had responded to actions and interactions from thecompany and its agents. The company worked with Ohio StateUniversity to develop an approach to analytics that would producedesired outcomes and gain the greatest buy-in by operationalunits.

Based on that research, Nationwide identified four “signature”experiences that had the greatest impact on how customers viewedthe company: welcoming new members, following up after a loss,communicating premium changes, and contacting customers to preformneeds-based assessments of their insurance products.

“Those are proactive relationship-building opportunities,” saysWes Hunt, vice president of Marketing Information Management atNationwide. “Each of those involves a direct contact, whether onthe phone, in person, or using technology, that can improvecustomer retention and sales.”

The carrier then built a customer experience management system,leveraging an enterprise data warehouse solution from Teradata andIBM's InfoSphere mater data management solution, to consolidateexisting customer data and capture customer contact data from phonecalls, Internet transactions, mail and email, and social mediainteractions. The web-based platform enables Nationwide to deliver“analytic packages”—action items within other business processes toguide actions taken by Nationwide staff and agents who areresponsible for customer contact.

Amica, which uses a variety of analytic tools from SAS, alsofound that customers who had more contact with the company weremore likely to stay customers. “Customers who experience ourunderwriting or claims service are much less likely to leave thanthose whose only contact with us is the bill,” Petrarcasays.

Powerful Tools

In addition to a wealth of data from internal systems, there ismore third-party information available for customer analytics thanever before. “There is a proliferation of commercial and governmentsources that is accessible and can be analyzed,” says MatthewJosefowicz, managing director at Novarica. “Consumer buyingbehavior, economic indicators, even weather patterns—you can lookat how all of those things affect you against your own customerdata.”

At the same time, advances in the ability to easily visualizethe results of complex queries has expanded the impact ofanalytics.

“The visual capabilities that are being offered in themarketplace today are light years ahead of what had beenavailable,” Pauli says. “Visual analytics let the business side getdirectly involved with outcomes. In many cases, people who had beendealing with 'green bar' data are now able to run complex what-ifscenarios and see the outcome on maps, bar graphs, and other visualdisplays.”

But there are still obstacles to analytics that carriers mustovercome. “The biggest challenge is getting hold of customer datathat may not be easily consolidated between different systems,”says Josefowicz. In addition, data quality and lack of anenterprise data model are barriers to the process.

Nationwide contended with numerous challenges in its effort,including different operational systems with disparate data, dataquality issues, and incomplete data history and contact rules. Thatmade it difficult to get a single view of the customer.

“Customers don't distinguish between business units regardlessof the number of relationships they have with Nationwide,” Huntsays. “It was important that we create a common customer profileregardless of the systems that data comes from.”

Nationwide used a variety of data cleansing tools to deal withdata that already existed, deployed address cleansing tools toensure that newly captured customer data was accurate and creatednew processes and system edits to ensure data elements wereaccurately and fully captured going forward.

“We focused on data elements we had been capturing but we didn'thave sufficient detail on,” says Chetan Kandhari, vice president ofIT Business Solution Applications at Nationwide. “For instance, aphone number or email might have been optional before, but werealized we needed to focus on capturing that to allow us to bettermatch customer data.”

Nationwide obtained strong central support for its datagovernance and consolidation initiative. “We established theconcept that customer data is an enterprise asset,” Hunt says.“There are stewards of data, but no one team 'owns' data.”

That type of top-level support is essential tothe success of a customer analytics initiative for a multi-productcompany. “Finding data is not a problem,” Breading says. “You haveto have the will and the skill to analyze it. You need a seniorteam that sees data as a competitive advantage, puts the rightresources in place to go after it, and staffs [the initiative]with skilled people who understand the business and technologysides to leverage analytics.”

Channel Challenges

Insurers who sell through independent agents face anotherobstacle in trying to act on their analyses of customers.

“The reason why insurers have lagged other areas of financialservices is that they have less of a direct relationship with thecustomer,” says Josefowicz. “Most companies are really organizedaround thinking about their product line and the distributionchannel, not the customer.”

Nationwide worked to demonstrate to agents that the actions theywere being asked to take in any of the four “signature experiences”would lead to results. “We focused on establishing trust, meaningthat if you as an agent are receiving information and arecommendation from us, you can trust that it's going to lead to agood outcome for you. That action could be as simple as validatinga phone number or address, or it could involve attempting to sellanother product or something more complex,” Hunt says.

“We created a 'closed loop' system where we track what actionswere recommended, what actions were taken, and what the resultwas,” Kandhari says. “We also compared that to the results we weretrying to achieve—if we were trying to improve retention of acustomer segment, we could show agents what actions and under whatconditions that would happen.”

For instance, Nationwide was able to demonstrate that agents whofollowed its recommendations for customer prioritization andproactive contact experienced a lift in renewal rate—anywhere froma 10-point increase to double the previous percentage.

Analytics Drive Results

Josefowicz encourages carriers to keep asking questions in thesearch for effective customer analytics.

“Start from a clean perspective instead of being constrained byprevious generations of thinking—don't be limited by whatinformation you used to be able to correlate. It's really possibletoday to take a look at new correlations between differentcustomers, products, losses, and pricing information and answerquestions that may not have been possible to answer before,” hesays.

“For instance, what does a small pricing change do to retention?A larger change? Or how does speed of claims resolution affectretention or an agent's future volume? You can look at a lot ofdifferent potential metrics that can drive increased business inways people had never thought to do because the process had beentoo onerous,” Josefowicz says.

“Improved customer satisfaction drives results,” Petrarca says.“Anything that we can do to increase customer satisfactionwill help our company and our policyholders succeed.”