Lets say your business strategy demands that you implement a newpolicy administration system. The current system wont fill the billbecause it wasnt designed to take advantage of the Internet, moderndatabase systems, or XML. Or youre putting up a new line ofbusiness and it makes sense to use a modern platform. Or you wantconsumers and agents to have access through your Web site and thatsa stretch with your current technology.

What do you do?

One alternative is to use what someone else has already puttogether and take advantage of the vendors economies of scale. Buteven with rich configuration options, the package may not suit yourneedsthat is, your business strategy. You could surrender part ofyour strategy to fit a package, but whats the point? Or you couldcontract for custom modifications (or do them internally), but thenyou may not be able to take advantage of new releases of thevendors base system. You may be stuck out on a limb.

Another alternative is to do it yourself. But how? Do you havethe technical talent and can you keep them? What happens if theproject gets out of hand and goes on for years without delivering aworking system? Should you contract out to a big-name consultingfirm? Can you afford it? Will they deliver?

Buy a package? Do it yourself? Get somethingbut not what youwant? Or try to get what you wantbut end up fired. Its aproblem.

The Toolset Approach

For at least a generation, smart technical people have beenaware ofand trying to offer solutions tothe buy/build dilemma ininsurance and other areas. Third- and then forth-generationlanguages offered a leg up. Object-oriented programming and add-inmodules were another step. Remote Web servicesthe ability to accessdata and functionality from a third partypromises to obviate theneed for some development projects. And modern developmentenvironments from Microsoft, IBM, and others can be a big help. Butthese are all generic approaches and dont provide specificinsurance-space leverage.

Another approachand one weve seen several times in thisindustryis for some clever person to point out that it makes nosense to do insurance from-scratch development over and overbecause we already know a good bit about the problems andsolutions. It would make more sense to create a toolkit thatalready has these elements available along with a configurator thatwould allow insurance business analystsnot programmersto supply thespecifications. The toolkit would then create the actual softwareneeded to run the applications.

The toolkit wouldnt be a software development environment. Itwould be a policy administration system generation environmentapowerful, insurance-specific wizard you could use to create customsystems. Then you could have it allwhat you want and with littlerisk. A pretty attractive idea, dont you think?

Of course there are some potential problems and risks withtoolsets, and well look at them below. But first, its time tointroduce Severan and its new toolkit for creating Web-based policymanagement systemsFocalPoint V2.

The Severan Approach

Severan (www.severan.com) was formed in 2000 by migrs fromMerchants and Business Mens Mutual Insurance Company. Severan CEOCharles Talmadge had served as CEO of Merchants, and vice presidentof development Bill Gregory had been VP of IT. The remainingSeveran staff of 10 consists of carrier and vendor veterans, somewith 30 years in the field. Software development itself is doneoffshore under the direction of Gregory. Severan is privatelyheld.

Severan was founded with the idea of creating a toolset to makeit easy for carriers to create their own full-function policyadministration systemsusing business analysts rather thanprogrammers. Its first product, FocalPoint V1, was written for theAS/400 marketplace and is in use by three carriers today. Late in2000 the Severan folks concluded, correctly I think, that the realopportunities lay in the Web world. They took what theyd learned increating V1, focusing on delivering the same broad functionalitybut with a different approach to both the toolset and how thesystems would be delivered to the end user. The new product,FocalPoint V2, is scheduled for release this month.

Severan says its market focuses on P&C carriers with directwritten premium below $750 millionnot because the software wouldntcreate appropriate and scalable systems for larger carriers, butbecause large carriers often have large IT staffs, who have anappetite for doing their own development. The software usesMicrosofts SQL Server or IBMs DB2 and, with load balancing servicespart of the Web server environment, the applications should bescalable to high policy and user volumes. (Severan isnt in thebusiness of Web hosting, but it can put carriers in touch withorganizations who make that their business.)

Because FocalPoint V2 creates Web-based applications, carriersusing it can solve three problems at once: ubiquitous access bycarrier staff, agent access, and insured access. Security settingsand the ability to have multiple versions of the same online formmake it possible to serve a number of different constituencies fromthe same system and configuration base.

The FocalPoint V2 Approach

One could think of FocalPoint V2 as having two main elements.The first is the Configurator, written in VB and the tool thatbusiness analysts use to specify what a system will contain and howit will act. The second is a set of server-side Java code thatactually runs the application from conventional Web servers.

What kind of systems is FocalPoint V2 intended to create? In aword, comprehensive. The system handles rating, ACORD and policyforms, billing, commission accounting, statistical reporting,ceding reinsurance, claims, access to external data sources, ACORDXML-based interfaceand even out of sequence endorsements.

To set up a new line of business, a user begins by definingrelevant elements of the database that are used later in rating,forms, and other areas. Next, he describes the various entitiesthat may participate in the processthe carrier (or carriers),agents, bureaus, reinsurers, and so on. Given these basicdefinitions, he can lay out the policy itself, including the ratingprocess, forms, and statistical reporting specifications. Thencomes claims handling, producer contract definition, reinsurancearrangements, billing plans, and security. Once thesespecifications are in place, the Configurator can generate the Javacode to run the application on the Web server.

Under consideration for a follow-on release of FocalPoint V2later this year is automated underwriting, imaging, workflowmanagement, and a comprehensive set of line of business templates.That last point is worth more comment.

FocalPoint V2 is a toolsetnot a finished system. Thats part ofits virtue, but that means carriers who use it have to go throughan exhaustive specification process and thats not for everyone. Inorder to give carriers a head start, Severan will create and offerline-of-business templates that carriers can use as a startingpoint for their own policy specifications; BOP and Commercial Autotemplates are ready now. In most cases, that should substantiallyreduce the time carriers need to spend, while still allowing theflexibility thats attractive about the product. And for thosecarriers who would like to outsource the configuration processaltogetherat least for the first go-round or twoSeveran isestablishing relationships with third-party consultingorganizations who can supply those services.

What are the downsides? That Severan is a small startup, ofcoursebut it has done something thats very clever and shouldnt beignored simply because its new. And some obvious concerns, such asthe lack of support for automated underwriting and workflow, arealready in the development queue. The environment needs moresensitivity to the management of the specification process,especially regarding maintenance over time. The system can trackspecification versions, and it does have a publish/lock feature,but real management of the ongoing use of the product still has tobe handled outside rather than within the system. In other words,the toolset may be great for defining policy systems, but it doesntpay enough attention to the management and documentation of thatongoing process. (After hearing me describe this concern whilewatching a V2 demo, Gregory now reports hes begun writingspecifications for a layer to answer the need.

John Ashenhursts company, Sound Internet Strategy, providesconsulting, Web site evaluation, and seminar services to carriersand their trading partners. He can be reached at [email protected]or (360) 376-1090.