In search of the knowledge that can extract profits out of astaggering amount of available data, larger insurers have madeimplementing business intelligence (BI) tools a required course forIT and business units. Smaller carriers have been slower to applythis lesson, but as the amount of information increasesexponentially, theyre realizing the time is now to graduate to a BIsystem.

By Robert Regis Hyle

Is business intelligence in your future? If you are a largeinsurance carrier, it likely already is part of your operation.Mid- and smaller-size insurers have been slower to get on boardwith it, but that could be changing. Its growing quickly, says MarkSchwartz, managing consultant with Accelerated Consulting Group.The majority of Fortune 500 companies are using it and are havingsuccess with it.

The amount of data available to insurers today is increasing ata ridiculous rate, and the ability to spin that data into betterproducts, higher sales, improved customer service, and less riskmeans insurers of all sizes need to examine how they use their dataand where they can go with it. We have enormous amounts of dataacross our organization. The challenge is turning [the data] intouseful information and delivering it to the right executive at theright time so that the correct business decisions can be made, saysMark Coleman, vice president, field administration, for ThriventFinancial, which has $152.4 billion life insurance in force.

Insurance carriers have looked with pride at their reputationfor acquiring extensive data on their policyholders, risks, andexposures, but their great success in collecting it has created anequally great dilemma in compiling and leveraging it.

Intelligent business decisions based on the information acarrier has amassed can be lost in the data deluge. Companies havespent a lot of time gathering data and formatting it into a usefulformat that can be passed on to someone else, Schwartz says. Theynever had time to analyze the data.

That was prior to the introduction of business intelligencetools into the marketplace. Carriers may believe they have beendoing some form of business intelligence over the years as theyanalyze their own data, but if they think thats the same thing,they might be fooling themselves. There is an order-of-magnitudedifference, says Keith Downs, vice president, operations andtechnology, for Vanguard Fire and Casualty.

Data, Data Everywhere

Heres some data on data. A recent study by the University ofCalifornia at Berkley reports the world produces between one andtwo exabytes of unique information per year, which is roughly 250megabytes for every man, woman, and child on earth. An exabyte is abillion gigabytes. Clearly, we are becoming overwhelmed withinformation, says Schwartz.
While many people assert there is no such thing as too muchinformation, there is too much data to look through if you have todo it the old-fashioned way, says Downs. Tools help in puttingscope to the data, he adds. Theres never too much policy data ordata about the areas where were writing [policies]. Demographicschange all the time.

With all this data available, IT departments are faced withseveral challenges: Part of the problem is we have a lot of datacoming at us that is irrelevant for specific decisions that have tobe made, says Schwartz. Information is being rekeyed, which isexpensive in terms of time and cost. The tools are allowingcompanies to work a little smarter and spend their time analyzingthe data because it already has been gathered for them.

Gather and Analyze

Tom Chesbrough, executive vice president of software supplierThazar, says insurance carriers are looking for a depth ofknowledge they ordinarily couldnt find. They are looking deep tounderstand the factors involved in identifying risk and pricingcharacteristics and more predictive characteristics, and they keepdriving in that direction on both the policy administration/pricingside and the claims reserving/analysis side, he says.

While much of the effort has gone into data collection,crunching that information is what makes it an effective tool forcarriers. Companies are becoming enabled to progress from gatheringto analyzing the data, says Schwartz.

With BI tools, we can see what is going on and communicate whatis going on, says Gabriel Fuchs, senior manager, sales andmarketing, for La Suisse Insurance in Switzerland. BI gives us anoverview with details behind it. We can be more proactive.
According to Downs, analysis is what makes Vanguard stand out fromother insurers writing property coverage in the volatile Floridainsurance market. Vanguard was started five years ago in the wakeof Hurricane Andrew. In our market, catastrophe expertise is theonly thing that is going to help you be profitable, he says. Themarket is sensitive to how well you can spread your risk around thestate, the types of dwellings you are writing.

Downs believes that is where BI tools can make the difference.The more information you have about the business that is availableto write and the business you do write, the better off you aregoing to be, he asserts. If we are going to become experts in thiscatastrophe market, we need all the tools that can help us betterunderstand our current business and the available business that isout there.

Setting Standards

While the decision to use BI tools based on their benefits mightbe a slam dunk, implementing them is not without some difficulty.Think of it as an insurance company Tower of Babelwhile insurersmay possess the same data, different parts of a company may havetrouble communicating basic things such as terms and definitions.Our operational systems organize our data very differently acrossour lines of business, says Coleman. The field called transactioncode in our annuity system is not the same thing as transactioncode from our mutual fund processor. It is an arduous process tosort out all of these data definitions.

Most of the terms and definitions have been designedindividually in their own silos, but as the quest for more andbetter information continues, insurers have found the need to pulldata together from those different silos. The business people canttell you where the data is stored, and the IT people dont reallyknow what the data means in a business context, says Coleman. Soyou have two people talking cross-wise to each other, and the ITperson ends up guessing about what looks most like what thebusiness person is saying. The tools are pretty powerful, but thisis a big hurdle to get over.

To jump the hurdle without tripping, a carrier needs a strongdata architecture group, according to Coleman. Such a group isresponsible for managing data across the enterprise, settingstandards within the company, and managing code values. Settingstandards and having a dictionary to explain the terms, Colemanmaintains, brings some logic to the situation.
Joe Flynn, assistant vice president, application development, forthe PMA Insurance Group, with net premium written of $603.6million, agrees with the importance placed on adapting a standardlanguage for the BI tools. Documentation and a data dictionary, hesays. You need to define all your terms. We sat down with ourunderwriting managers and designed a data mart to pull togetherinformation most important to them.

But Coleman adds a word of caution regarding data marts.Separating data into its own departmental data marts has beentried, but Coleman believes carriers need to keep a close eye onhow one department uses data acquired from a separate department.In my experience, once you give people a little bit of data, theyare going to want a lot more, he says. If you havent thought thingsthrough, youll be forced to pay the data tax sooner or later.

Managing the Data

Thrivent is looking to better manage data in many areas, Colemansays. One of these areas is performance management. This isinformation that helps our managers track progress against goalsand identify trends early. These scorecards will help keep ourmanagers focused on what weve all agreed is important, he says.Compliance and compensation modeling also are areas that willutilize business intelligence technologies. New systems and toolssuch as Centives EIM, Callidus TrueComp, and Siebel are beingevaluated that will help us identify, investigate, and react totrends quickly, he says.
Supplying information directly to the end user not only speeds theprocess, Schwartz says, but allows business users to be morecreative in combing the data. There are reporting tools that takereporting from a data-centric modeldealing with fields andtablesand brings it into a business-centric model that deals withcustomer information, he notes. It takes [the work] from atechnical task to a business-user-friendly task, and that ease ofuse enables business users to interact with the information andbypass the information-gathering pointthe IT department.
IT is not completely out of the loop. It still is charged withsetting up the infrastructure in a way to protect the business userfrom making errors on the database, for instance, pulling upqueries that bring the database to its knees, says Schwartz.
Having the input of the users is the only way to ensure they getthe data they need, adds Flynn. We sat down with our underwritingmanagers and designed a data mart to pull together information thatwas most important to them, he says.

Starting to Mature

Schwartz has noticed BI tools have been maturing since he beganworking in the field in 1999. The progression of what the toolswere then and what they are today is just incredible, he says.Todays tools are much more sophisticated and user friendly.

One thing insurers have to battle is the tendency to look at thedata in the same old way. You have an incredible opportunity whenyou move from the old to the new to take advantage of much morefunctionality, says Schwartz.

Business users likely might be accustomed to studyinginformation off an Excel spreadsheet, and Schwartz believes it isimportant to change their mindset, something that takes a littletime. We change the mindset from How can I make this look like anExcel spreadsheet? to What information do I really need to makethis decision?

BI also leaves less room for error, Schwartz asserts. Thetypical problem with a spreadsheet is the numbers tend to beomitted or added in, he says. It is subject to human error. Whenyou automate a report, the numbers are populated from the datawhere you have some control. You imbed business rules that areaccepted across the organization.

As an example, Schwartz points to an innocent enough area suchas sales. You can have three executives come to a meeting, and theyall have different sales numbers based on how they recognize salesand when they recognize sales, he says. When you implant a BIsolution, you can imbed those business rules, such as when yourecognize a sale, in the underlying data mart or within thereporting function. At that point, everyones reports are the same.A hidden benefit, he adds, is people begin to work more closelybecause they are all on the same page.

Where to Begin

The depth of knowledge needed by insurers, Downs points out, canno longer be satisfied by simply asking the IT department to createqueries they can run against the data. We now need tool sets thatcan monitor the data as it enters the warehouse and provide alertson a day-to-day change basis to better understand how we analyzethe details.

Companies need to attack a BI proj-ect one department at a time,Schwartz believes, and should aim to see results within 90 days.You dont want something that is going to take months or years toimplement, he says. That just wont work.
Insurance carriers should start small when beginning a BI project,he adds. Pick a couple of pain points, focus on them, deliver, andthen move on, he says. A place to start is in an area where thereare problems creating reports, getting the numbers that are needed,or where there is a shortage of analysis. An incremental approachis much more effective in deploying BI solutions, he advises. Youare changing the mindsetgoing from static reports to something moredynamic. People start to understand the true potential of BI afterthey start using their own data.

The Black Hole

For Companies With at Least 1,000 Information Managers

$3 millionamount of money
wasted digging through or
recreating information

$15 millionamount of potential revenue opportunities missed

Source: International Data Corp.

Tech Guide: Business Intelligence

Acorn System
Houston, Texas
713-963-9000
www.acornsys.com

ALG Software
Atlanta, Ga.
888-374-4321
www.algsoftware.com

Appfluent Technology
Arlington, Va.
703-284-0800
www.appfluent.com

Brio
Santa Clara, Calif.
800-879-2746
www.brio.com

Bristol Technology
Danbury, Conn.
203-798-1007
www.bristol.com

Castek
Toronto, Ontario
416-777-2550
www.castek.com

Centive
Bedford, Mass.
781-778-8000
www.centive.com

CNR
Reno, Nev.
775-851-2829
www.cnrsearch.com

Corporate Systems
Amarillo, Texas
800-927-3343
www.csedge.com

Digital Sandbox Inc.
Reston, Va.
703-390-9770
www.dsbox.com

Guidewire Software
San Mateo, Calif.
650-357-9101
www.guidewire.com

The Haley Enterprise
Sewickley, Pa.
412-741-6420
www.haley.com

HCL Technologies of America
Schaumburg, Ill.
847-303-7200
www.hcltech.com

Infoglide
Austin, Texas
512-532-3500
www.infoglide.com

Information Builders
New York, N.Y.
212-736-4433
www.informationbuilders.com

Millbrook, Inc.
Bethlehem, Pa.
610-867-7400
www.millbrookinc.com

NCR/Teradata
Dayton, Ohio
937-445-5000
www.teradata.com

PeopleSoft
Pleasanton, Calif.
925-225-3000
www.peoplesoft.com

Pinpoint Solutions
Livingston, N.J.
973-716-0723
www.pinpnt.com

SAP
Newtown Square, Pa.
610-661-1000
www.sap.com

SAS Institute, Inc.
Cary, N.C.
919-677-8000
www.sas.com

Service Integrity
Newton, Mass.
617-965-0281
www.serviceintegrity.com

Skywire Software
Frisco, Texas
972-377-1110
www.skywiresoftware.com

Spotfire
Somerville, Mass.
617-702-1600
www.spotfire.com

Thazar
Overland Park, Kans.
913-327-7881
www.thazar.com

Universal Conversion Technologies
Addison, Texas
214-348-2000
www.uctcorp.com

Valen Technologies
Denver, Colo.
720-570-3333
www.valentech.com

Visibillity, Inc.
Chicago, Ill.
888-484-7424
www.visibillity.com