By John W. DeWitt

|

Special Report sponsored by:

Ernst & Young

Insurance carriers certainly feel the pressure on top-linegrowth in the current underwriting cycle - but more focused andeffective agency management is a perennial challenge in anyeconomy. Simply put, how do you more accurately evaluate, and thenmaximize production of, distribution channels - without incurringsubstantial additional costs?

|

"Traditional agency management approaches face limits ascompetition intensifies in a finite market," notes Gary Ciardiello,a New York-based principal at Ernst & Young who leads thefirm's predictive modeling group in the Insurance and ActuarialAdvisory Services Practice. "Innovative insurers increasingly arerecognizing the need to augment production management with a datarich, metric based more analytical framework."

|

To this end, Ciardiello and his colleague Benny Yuen, seniormanager at Ernst & Young, have been working to developanalytics-driven processes that complement traditional agencymanagement practices and improve the way carriers evaluate andimprove agency performance. They describe this as a holisticapproach to agency management - one that incorporates objectiveanalytical tools in conjunction with more subjective approaches tomaximize agency potential.

|

"Analytics do not replace what companies are already doing tomanage their agencies," Ciardiello explains. "But analytics canprovide a more disciplined, and rigorous framework forunderstanding agencies' performance and maximizing their potentialfor improvement." The ultimate goal, Ciardiello and Yuen believe,is to create some set of standardized metric that calculates"agency potential."

|

To illustrate the analytical approach and its value, considerthe example of two independent agencies, in different geographies,that place personal auto insurance for the same insurer (seechart). Their current premium volume is the same - $250,000 - andtheir market potential is similar. A traditional evaluativemeasure, such as premium volume, would indicate that the twoagencies in this comparative example have the same growthpotential. But premium volume is a lagging indicator, whereas theAgency Potential Index that Ernst & Young has developed tells amuch different forward-looking story. Agency B scores 22 pointshigher in the composite Agency Potential Index.

|

To generate this metric for Agency Potential, several indices,based on a 1 to 100 scale, have been developed and applied to eachagency. The final composite score indicates each agency's forwardpotential - providing empirical insight that can help the insurerallocate its marketing dollars more productively. (Note that thisexample uses simple averages. In practice, statistically derivedweighting methodologies would be applied to each index to arrive atthe overall score.) A comparison of individual indices explainssome of the reasons why Agency B scores higher - because of factorssuch as better infrastructure, more competitive rates in itsmarket, a stronger brand, and superior alignment with theinsurer.

The Agency Potential Index: A ComparativeExample
MetricType of FactorAgency AAgency BReason codes
Current Performance (Premuim)Internal$250k$250k
Internal IndexInternal4075Agency B has much higher growth potential and betterinfrastructure (number of employees, CSR ratio) than Agency A.
Rate Competitive IndexExternal6075Insurer's rates in Agency B area are more competitive.Insurer's ranking for Agency A regarding premium volume is lowerthan Agency B.
Brand Recognition IndexExternal6080Insurer's brand is stronger in area where Agency B placesbusiness.
Market Size IndexExternal5050Both areas are similar in terms of insureds per agency andagencies per capita.
Agency Alignment IndexExternal4080Agency B customer demographics are highly aligned with insurer.Agency A customer demographics skewed to customers that insurer isnot preferred writer.
Average Score5072



|

The Limits of Traditional Agency ManagementApproaches
Compare the analytical approach in the preceding example withtraditional, tried-and-true approaches that carriers employ tomanage their agencies. Insurance companies and their sales andmarketing teams often focus on production metrics (such as premiumvolume), profitability metrics, compensation strategies, and agencyplant strategies. Some also address back-office metrics - such asover order vendor reports, up-rating frequency, and target markets.In addition to tracking agencies' written premiums, currentproduction metrics can include new business counts, inforce counts,and retention ratios. More sophisticated approaches can encompassthe number of quotes and conversion rates along with averagepremiums.

|

To influence these metrics - for instance, to increase thevolume of quotes per agency - insurance companies rely upontraditional levers, such as changing commission structure,underwriting guidelines, or pricing, or investing in sales andmarketing. However, while these approaches can be effective,"pushing any of these levers inevitably results in expense orprofit implications," Ciardiello notes.

|

Traditional agency management strategies also rely more on theintuition and anecdotes of the sales and marketing team, ratherthan empirical data. This subjective process depends heavily onknowledgeable individuals, can be difficult to replicate, and isinherently imprecise, potentially leading to overinvestment orunnecessary expense. In other words, traditional levers mightimprove the performance of some agencies, but not of others - andwithout more analytical insight, it's hard to know why, or whatelse to do.

|

Developing Analytical Measures of AgencyPotential
More accurately understanding the underlying potential of eachagency helps explain why seemingly similar agencies performdifferently - and indicates how carriers can maximize theperformance of agencies with the greatest potential. "Potential canbe defined by understanding and analyzing two sets of factors -internal and external," Yuen explains. "Internal factors are thoseagency attributes that are in the control of the agency. Externalfactors are those in the marketplace, outside the direct control ofthe agency."

|

For instance, internal, agency-controlled factors to take intoaccount might include the agency's years of experience, credit,E&O limits, and size - its premium volume, number of employees,and CSR ratio. (The number of companies represented is alsoimportant when managing independent agencies.)

|

External factors to consider might include the competitivenessof the agency's rates - measured by win rates and market perception- as well as market size and brand awareness. Insurance companiesalready pay attention to these factors, Ciardiello notes - butinconsistently and in isolation rather than as part of ananalytical framework. Another important factor, less commonlyevaluated, is how well the agency aligns with the carrier.

|

"These key external factors typically are being looked at on anad-hoc basis - and typically on an inconsistent basis - maybeannually or quarterly - and are evaluated subjectively, from asales or marketing perspective, rather than by actuaries or otheranalytical executives," Ciardiello says. "There's nothing wrongwith this - but there is room for improvement by combining analysiswith subjective evaluations."

|

To analyze these various external considerations in a rigorousand holistic manner, Ciardiello and Yuen recommend that insurersdevelop a series of factor-specific indices that in turn will beintegrated into a single composite metric of agency potential.While an exhaustive list of the factors can not be practicallyconsidered, companies would get a powerful start on building arobust analytical framework by calculating indices for ratecompetitiveness, market size, brand awareness, and agencyalignment.

|

A useful and statistically meaningful Rate CompetitiveIndex could include a quote/conversion analysis and winrate analysis, whereby geographical market baskets encompasstargeted customers, inforce profiles, and census information. Theother important component to integrate into a rate competitiveindex is market perception, which can be evaluated with reasonableobjectivity via the utilization of mystery shoppers and surveys.These measures, when combined with analysis of marketing spend, canalso be used to generate a Brand RecognitionIndex.

|

The Market Size Index would encompass commonlymeasured attributes such as the number of insureds per agency, thenumber of agencies per capita, and the number of competitors'agencies per insured.

|

A measure less commonly used by most insurance companies is theAgency Alignment Index. It's also a bit morechallenging to quantify - but worth the effort, Yuen says, becauseof the often unrecognized role that alignment plays in constrainingor maximizing the potential of an otherwise well-positionedagency.

|

"An agency can have certain sweet spots in its book of business,and those sweet spots may or may not align well with the insurancecompany's sweet spots," he explains. "If the agency writes a lot of40- to 45-year-old single males, but the company is not strong inservicing that area, there is a misalignment between the agency andthe company that hampers their ability to successfully write newbusiness. The overall sales numbers wouldn't give that particularinsight."

|

These indices, and any other agency metrics, should incorporatefeedback loops - in other words, should be self-learning andcontinually refined through ongoing practice and experience. "Youwill have to adjust the dials on all these indices, take a holisticapproach when reviewing them and balance them properly forindividual organizations - and that balance can and often doeschange over time," Ciardiello notes.

|

Integrating the Indices: The Agency PotentialIndex
Given that Ciardiello and Yuen propose a more holistic as well asanalytical approach to agency management, the next step afterdeveloping a set of agency performance metrics is to integratethese metrics into a composite Agency PotentialIndex. Speaking more technically, companies would developa predictive model utilizing a multivariate regression approach -in contrast to the traditional actuarial method of analyzing onevariable at a time.

|

"If you don't look at all underlying factors in the book, youcan get a false read," Ciardiello explains. "You have to factor inall these data points and indices together to see what thepotential production would look like."

|

The individual indices and other agency metrics that encompassinternal and external factors would be combined to calculate thecomposite agency potential index. The weight for each factor wouldbe determined by the most appropriate statistical method - forinstance, logistic regression, relative scoring, or principalcomponents analysis. The statistical method would depend on thespecification of the index as well as on other considerations suchas whether a factor is objective (e.g. number of insureds peragency) or more subjective (e.g. market perception). A subset ofthe company's most successful agencies - say, the top 10% - wouldbe used to benchmark the remaining agencies.

|

Each agency in the composite agency potential index then wouldbe scored on a scale of 1 to 100. If calculated properly, eachagency that scores, say, an 84 would have the same productionpotential - even if the factors generating that composite score aredifferent.

|

Leveraging the Composite Agency PotentialIndex
Ciardiello and Yuen envision a number of beneficial applicationsfor the composite agency potential index. Embedding the index intogrowth strategies could help an insurance company focus on its bestopportunities for growth, documenting what works and what doesn'twhile creating a consistent framework for ongoing distributionchannel analysis. The index also could guide the development ofagency action plans, influencing agency appointments andobjectives.

|

Benchmarking is another potentially powerful application of theindex, which would provide a means to normalize productioninformation across agencies and then perform "apples to apples"comparisons. This approach could uncover highly useful insightsthat other approaches to agency management ignore, resulting inbusiness plans that are more analytically driven and more preciselytailored to each agency.

|

"If you're getting a lot of production out of an agency thatdoesn't have a high score, maybe you've maxed out that agency'spotential," Yuen explains. "In other cases, an agency might have ahigh score but really isn't performing based on its potential, sothere's an opportunity that exists to further improve itsperformance."

|

For more information, visit: www.ey.com/us/actuarial.

|

Leveraging Predictive Models to Maintain Profitabilityin a Soft Market
For more in depth analysis of this topic, view our webseminar! While predictive modeling applications have proven tobe very effective in underwriting, customer management, marketing,and other areas, deployment of models should be aligned andreconciled with other business considerations and strategies. Thissession touches on the concept of decile management as it relatesto the deployment of predictive models.

|

John W. DeWitt is a marketing consultant and business writerbased in New Salem, Mass. He can be reached at [email protected] or www.jwdewitt.com.

Want to continue reading?
Become a Free PropertyCasualty360 Digital Reader

  • All PropertyCasualty360.com news coverage, best practices, and in-depth analysis.
  • Educational webcasts, resources from industry leaders, and informative newsletters.
  • Other award-winning websites including BenefitsPRO.com and ThinkAdvisor.com.
NOT FOR REPRINT

© 2024 ALM Global, LLC, All Rights Reserved. Request academic re-use from www.copyright.com. All other uses, submit a request to [email protected]. For more information visit Asset & Logo Licensing.