The importance of analytics in the insurance industry—anindustry driven by data—cannot be disputed, but with the size andnumber of insurance carriers in the U.S. market, where they are inthe process ranges widely.

Unfortunately, many carriers have yet to take the first step asthey grapple with other issues such as legacy systems andincomprehensible data. So we asked several insurance technologyanalysts the following question: What is one step insurancecarriers can take to improve their ability to use analytics withinthe enterprise?

Click through the slide show to see the responses from: MattJosefowicz, of Novarica; John Lucker, of Deloitte Consulting; MarkBreading, of Strategy Meets Action; Karen Pauli, of CEBTowerGroup; Ellen Carney, of Forrester; and Chuck Johnston, ofCelent. 

Matt Josefowicz, managing partner,Novarica 

Unfortunately, most insurers that aren't already using analyticseffectively have more than one step to take to get there.Investments in data quality, building and testing analyticalmodels, and incorporation of analytics results into businessprocesses are all required in order for insurers to improve theirresults through better use of analytics.

One step that insurers can take, though, is for business leadersto insist on data and reject business cases based on lore or gutfeel. Only when business leaders demand to know in detail whyunderwriting results are trending a particular way or customerretention is trending another way, why service costs are going upor claims productivity is going down, can analytics start todeliver value within the company.

Then the right technology, process, and training improvementscan be made. But until senior business leaders establish a culturewhere analytics are the critical foundation for taking action, ITinvestments in analytics will not yield positive results.

John Lucker, principal, DeloitteConsulting

Naming one step as most important to use analytics ischallenging because it's important to think about six steps toend-to-end analytics execution—analytic strategy, analyticsdevelopment, operational implementation, technology integration,change management, and performance management. 

However the first step in the process—strategy—if not done well,can doom all downstream activities. Too often insurers fail todevelop their analytic strategy choosing instead to focus ontactical tasks; projects that they want to do or feel they need todo. At times, models and solutions are built and then companies ask"now what?" 

Analytics needs to be part of a holistic strategy of movingtoward a more metrics-oriented, fact-based business processculture.  An articulate strategy should have manycomponents, for example" (a) areas of opportunity; (b) expectedbenefit to be realized and KPI's impacted by strategic streams; (c)how benefit will emanate from analytic driven business process andtechnology integration; (d) how strategy will be socialized andmanaged organizationally; (e) how will results be recognized andtuned, etc. 

Without a full line of sight to a strategic vision it is hard toimagine how significant and lasting analytic driven business valuecan be realized.

Mark Breading, partner, Strategy Meets Action

There's no question about it. Insurers understand the power ofanalytics to improve every part of their business. SMA researchindicates that the vast majority of insurers are increasing theirspending on analytics projects; many of them are making significantincreases.

Successfully creating and leveraging analytics capabilitiestakes a concerted effort. It involves business strategy,organizational changes, training and skills, enterprise datamanagement, and new technology platforms and tools. All of theseare important and require planning, resources, and coordinatedimplementation. But one area rises above the rest inimportance—enterprise data management.

Data management may be less exciting—and more elusive—thanstrategy, organizational changes, technology, or training. But anorganization's ability to capitalize on analytics ultimately boilsdown to a few simple but critical questions, all of which point todata management: How complete and accurate is the data? Is the dataconsistent across the enterprise? Are users able to easily accessthe required data? Is the data organized to make consolidation andaggregation straightforward?

Insurers that place a high value on data management will be ableto answer these questions positively, positioning them to gainmaximum leverage from analytics.

Karen Pauli, research director, insurance, CEB TowerGroup

The most immediate response to this question is "get your datain order."  However, from CEB TowerGroup's experience withcustomers adopting analytics, the response is "assure that yourorganizational structure is set up for an analytics driven businessenvironment." Organizations that view analytics as the sole domainof actuaries or, following a trendier new job classification—datascientists—will not get the value they are hoping for.

It is critical for insurers to reorganize around blended teamsfor full analytics value.  Representatives from fieldoperations (e.g. claims, underwriters, loss control, sales, etc.),business analysts, statisticians, analytics experts, modelingexperts, and business strategy should all be part of blendedanalytics teams.  Certainly, the exact structure willdepend on the product mix of a carrier. However, the point is thatfull value of analytics comes from multi-discipline collaborationand decisioning.   

Actuaries and analytics will always be a cornerstone ofinsurance companies.  However, the actuarial goal ofcreating perfect pricing through analytics is in conflict withdriving innovative business outcomes through analytics whichinherently has elements of creativity, imagination, andscientifically created risk.  Blended analytics teams candeliver on a goal of innovation for competitive advantage andcreating these teams is critical to get the most out of analyticsinvestment.

Ellen Carney, principal analyst, insurance e-business andchannel strategy, Forrester

Anything to do with data and analytics is getting big attentionand even bigger investment increases.  That means thatinsurers have to develop analytics strategies that demonstrate thedata science team's (or whatever term your firm uses—Big Data,analytics, or metrics) impact on the insurer's financialperformance, both directly through analytics that measureperformance against the carrier's goals as well as the analyticsthat capture an indirect association to business performance likebrand awareness or perception.

That means that insurers have to base their analytics decisionsbased on what the business objectives are. Analytics that measurebusiness objectives, such as improving lead conversion, better riskprofiling or even the basic stuff like using analytics to reduceservice and sales costs help quantify the impact of data andanalytics strategies and help to justify further investments.

Chuck Johnston. Director, Americas life/annuity and grouppractice, Celent

Insurers need to get past perfect and focus on good enough. Manyinsurers are still very focused on data quality to the point wherethey are missing out on near term opportunity.

Insurers have been doing data management and cleansing projectsfor decades and the state of industry data has gotten better but isfar from cleansed. While there are immediate customer servicesissues in using ambiguous data in transaction processing systems,analytical systems can deal with ambiguity in information,especially if there is sufficient volume of data.

Remember, the more data you are processing the greater potentialfor cross reference and use of statistical tools to actuallyincrease overall accuracy. The single most important step insurerscan take to improve their ability to use analytics within theenterprise is to get started.