?

Replacing legacy claims systems can be stressful for both theclaims and IT departments. Here are some tips from the trenches onhow best to ease the strain and maximize the gain.

BY DONALD LIGHT

Any time an insurance company thinks about replacing a coreclaims system, there always are a lot of reasons for not taking anyaction. For instance, most legacy systems are fast, reliable, andwell integrated with other core systems and infrastructure. Theprospect of losing those positive features makes many claims and ITgroups think twice about replacement.

A more subtle but equally important factor is just abouteverything adjusters do (i.e., the claims processes, tasks, andactivities) is molded by the legacy claims system. Changing thatsystem will mean almost every process becomes a candidate forchange, and some processes definitely will change. Potential oractual wholesale change creates challenges for line claims staffand every level of claims management.

In the end, though, many insurers decide the claims function isjust too important to live forever within the constraints of legacyclaims systems. For example, it can be hard to find people with theright skill sets to maintain and change themnewly hired staff looksaskance at green screens, obscure codes, and rigid workflows. Inaddition, as important as pricing, underwriting, and servicing are,people and businesses buy insurance because they want someone tomake things right if something bad happens, i.e., they want theirclaim paid quickly, fairly, and accurately. Even more compelling inthe argument for a cost-cutting replacement is the amount offinancial leverage claims exert over an insurers results. Sixty to70 cents out of every premium dollar is paid to claimants (and morein bad loss years). Another 10 to 15 cents goes to the appraisers,adjusters, and attorneys who decide how much is paid to theclaimants. A relatively small percentage change in reserves,especially for long-tailed liability lines and mass torts, can havea powerful impact on the income statement as well as on perceptionsof investors and raters.

Replacing a claims system puts significant stress on the claimsgroup, the IT department, and their relationship. The
project cannot succeed unless both groups work closely and wellwith one another. This article, based on research on severalsuccessful claims systems implementations, identifies both thosestress points and the best practices for resolving them.
There are five steps in replacing a core claims system: making thedecision to replace; choosing a vendor; creating the detaileddesign of the new system; implementing the design; and deployingit. There also is a sixth step: adding value post-implementation.While not part of the replacement process strictly speaking, thissixth step does much to determine the level of benefit the insurerreceives over the life of the new system.

1. Decide to Replace
Simply deciding to replace acore claims system is not enough. A smart insurer will take thetime to map the key business and technology drivers that havemotivated the decision. For example, beyond stopping the pain of alegacy system, does the claims group want to make its adjustersmore productive or process change easier? Does IT want to lower itsresource commitment to maintenance or to migrate all new coreapplications to a new platform?

There also are corporate objectives and strategies. Is growth animperative? If so, there are implications for improving claimantexperience and policyholder retention. Or is improving underwritingperformance the focus of management attention? Then identifyingfraud and reducing leakage will take center stage. Or have past orfuture mergers and acquisitions created the need for a claimssystem that can serve to consolidate many legacy systems?

The payoff for this type of self-knowledge comes with the nextstep.

2. Choose a Vendor
An insurer needs to balance carefully the vendor-selection team. Itshould be small enough to move rapidly and large enough to siftthrough a lot of information from a long first-round list ofvendors. Beyond the obvious need for representation from bothclaims and IT, the team should include people with a hands-onknowledge of how the current processes and systems function (andmalfunction). The team also needs people who are senior enough sothat its recommendations will carry adequate weight.
When the team creates a request for information (RFI), it shouldmake maximum use of what has been documented in the previous stepabout drivers, motivations, objectives, and strategies. These willsuggest various high-level requirements for features and functionsthat should be embedded in the RFI, for instance:

Externalized workflow and rules engines that can be used bybusiness-side analysts to develop new processes and
decisions rapidly.

Imaging and document and content management capabilities tocreate a paper-free adjustment process, utilizing digital claimsfolders and able to generate required forms and correspondence.

Portals for claimants, agents, body shops, medical providers,etc., that facilitate inquiries, simple transactions, and sharedwork spaces.

Compatibility with Java or .NET
platforms.

It is worth highlighting one issue that sometimes does notreceive adequate attention: the nonlicense costs of implementation.The total implementation project requires a large teamsometimes avery large team with representatives from claims and IT as well asthe vendor and/or consultants. Different vendors solutions maydemand dramatically different levels of resources oversignificantly shorter or longer periods. Explicit recognition ofthese costs should be integrated into the selection process.

3. Design
In this step, the implementation team makes a series of criticaldecisions, including detailed requirements, use cases, userinterfaces, workflows, and decision rules. Good decisions will go along way to making the new system a success.

One of the most important decisions will be the trade-offbetween the amount of change in claims processes on the one handand the need to control the pace and cost of the implementationproject on the other. Every insurer implementing a new claimssystem intends to improve its claims processes. During design itmay become clear some of the things the claims staff wants tochange will require significant technology resources to configureor even customize the new system. Most insurers will opt forpostponing most of these resource-intensive changes until the newsystem has been in use for some time.

Additionally, the governance structure for the entire projectshould be in place at the start of the design task. Because of thescope and complexity of new claims systems, many insurers will havea working-level governance group meeting weekly and a policy-levelgroup meeting bi-weekly or monthly.

4. Implement
Most project resources are consumed during this task. This is wherecareful, detailed design work is rewarded. This also is where agood governance structure and process will handle new or revisiteddesign issues. Some changes are expansions of scope and often areviewed skeptically. Other changes may be prompted by hittingresource constraints within scopethese require more judgment.

Conversion of closed claims files sometimes causes unforeseendifficulties during this step. How far back should the conversionprocess go? Should all records be digitali.e., no paper? How do youhandle the inevitable incompatible data fields and unstructuredinformation? Leaving some form of the legacy system in placepartially will solve these issues.

5. Deploy
There are two major issues in deployment: phasing and useracceptance. Many insurers choose initially to deploy a new claimssystem in a single region or for a product line or function (e.g.,first notice of loss). This approach allows for a shakedown period,during which systems, process, and people issues can beresolved.

In a sense, every new claims system always achieves useracceptancebecause the old system no longer is available. But userswho are briefed on the new system during its design andimplementation tasks and are well trained, preferably by theirpeers, can make a big difference in the systems value over time.Some creative evangelism also can helpone insurers rollout reliedheavily on balloons, streamers, and a cake.

6. Add Value
This is a post-implementation step. The initial design trade-offsbetween process change and technology time/resources no longerapply. Modern claims systems have a depth of functionality,workflow, and rules capabilities that enable continuous processimprovement. To realize that value, insurers must create theappropriate organizational values and incentives.

Six Things an Insurer Must Get Right

There are no guarantees of success in lifeor in implementing anew claims system. However, industry research has identified sixelements of successful implementations.

The first is to build a technology road map and business driversinto the vendor-selection process. An insurer that does not have anexplicit understanding of both sets of issues before vendorselection starts will reap the consequences during implementationand for years afterward.

The second is to balance the what and when of process changes.Any claims group should have a long list of process improvements itwants to make. The good ones know how to time those changes topush, but not punish, claims and IT staff.
The third is to solve two legacy riddles: to rip or not to rip, andto convert or not to convert. Completely ripping out the legacyclaims system has the virtue of simplicity but the cost of a muchmore formidable integration challenge. The answer to the riddle ofconverting old claims files will depend on cost and need for futureaccess.

The fourth is to engage the claims group. Of course, claimsstaff always will be involved. The question is the quality of itsownership and participation. Does the staff bring deep experienceand a vision of the future? Does it have a we, not you, attitude?Does it spark anticipation and acceptance throughout the claimsorganization?

The fifth is to be friends with your vendor. Of course vendorswant every implementation of their system to be successful. Smartinsurers take advantage of two very specific kinds of knowledgevendors bring: what it takes to implement their system successfullyand the full range of features and functions their systemprovides.

The sixth is to manage the program. Good project and programmanagement skills are essential. To meet both expected andunexpected challenges, claims, IT, the vendor, and any consultantsall must share a sense of ownership.

Increasing numbers of insurers have their claims and IT groupsfollowing some version of these best practices, with the resultthey are realizing financial and competitive gains. u

is a senior analyst in Celents insurance group and is based inSilicon Valley. His current research focuses on how technology cansupport insurers strategies, objectives, and core processes.