According to several analyst firms, investments inclaim-management systems are on the rise as organizations seek tosustain growth, service, and profitability from their claimoperations. In 2003, Downey Insurance, a managing generalunderwriter for Indiana government entities, became part of thisinvestment trend and industry-wide implementation explosion.

At Downey, we recognized that our existing claim system wasoutdated; the system lacked automation tools, web-basedcapabilities, and workflow management functions. We needed a moremodern claim-processing infrastructure to drive operationalefficiency, control costs, and improve performance far into thefuture. Once a system was selected, however, we were faced with thedaunting task of implementing the new software, converting datafrom the previous system, and customizing the new solution to meetour specific needs.

As many claim managers know, system implementations can be anightmare. Strict project management is required or the projectwill be delayed and go over budget. Organizations also need theright blend of claim-management experience and technical expertisein order to configure the system properly for optimal workflow,best practices, and data integrity.

In September 2004, Downey's implementation was successfullycompleted. By sharing our own step-by-step approach, we hope toconvey some key strategies and lessons learned to the projectmanagers who are just setting out on their own implementationjourneys.

Build a Team

Our selection process had pointed to a clear solution that wouldmeet our business needs and objectives. As we set out to deploy thesystem, we realized the vendor's expertise could be leveraged as acrucial resource throughout the implementation process.

In March 2004, we held an implementation kick-off meeting withmembers from both of the organizations. When it came time to selectproject managers, we decided on two people: one internally and onefrom the vendor. This would ensure each side was completing tasksand driving action toward the “go live” date. The team alsoincluded members of Downey's claim and IT staff. Since adjusterswould be the ultimate end users, it was important to have theminvolved in the entire process, from system acquisition toimplementation.

Establishing a clear plan of action for the team also wasimportant. Our vendor provided a sophisticated spreadsheet thatserved as the plan of action. It documented action items, partieswho were responsible for specific tasks, and deadline dates. Downeywanted to go live in September, so the team developed the timelinewith this objective in mind. The plan kept the team focused anddefined roles and responsibilities, and assigned tasks were clearlyunderstood by all parties.

Set Goals and Responsibilities

At the onset, Downey planned weekly meetings to assess whetherthe project was on track, discuss challenges that came up, and makesure problem areas were resolved in a timely manner. No one wanteda long, drawn-out implementation. The meetings held everyoneaccountable to their respective duties and created significantproject momentum. Each week, we saw tasks crossed off the actionplan and new assignments delegated.

Another key factor for success was having an early trainingsession, which we scheduled two months prior to our implementationdate. In this training session, our vendor provided extensiveinformation on how to customize the system to fit our businessworkflow needs. We learned how to tailor screen views, set upbusiness rules, and configure system security. The training alsohelped us determine the fields into which we would map data. Wealso scheduled a second training session right before theimplementation date to ensure our users were comfortable with thenew application and could immediately resume production.

Within our action plan, we also included steps to customize thesystem. In the past, system modifications had to be performed by anIT technician. Since the new browser-based system was intuitive anduser-friendly, our business users could make most of the changesthemselves. In some cases, we required technical assistance inorder to design a particular workflow or select a configurationoption. Since our existing IT staff was unfamiliar withclaim-handling procedures, our vendor was instrumental in helpingus align the system to meet our needs.

In regards to specific customizations, we configured screenviews and selected what fields and tabs should be visible. We usedbusiness rules to automate and manage our workflow, as well asrules to monitor for specific events that would require supervisoryreview. There was tremendous flexibility in customizing reports. Wewere able to add, delete, and change report formats, which ensuredour reports would be ready once we went live. We customizedspecific monthly member reports, loss-ratio reports, and set upautomatic report distribution for our internal users in claims,loss control, accounting, and underwriting.

Here are general customizations to consider:

• How do you want the system's screens to look?
• What customizations must be made to meet immediate andlong-term claim-handling and risk-management needs? Schedulechanges according to priority.
• What claim processes and workflows would benefit frompre-configured business rules?
• What reports will you need once the system goes live?

Clean-Up Before Conversion

When converting from an older system, data integrity probablywill be an issue. Files may be missing data or orphan data mayexist. All this needs to be cleaned up, particularly for fieldsthat will be required in the new system.

Since conversion problems can significantly delay animplementation date, be sure to leverage your vendor's technicalexpertise. They have assisted numerous clients with conversionsfrom various systems, so they have extensive experience in handlingdata issues.

In our case, the vendor helped us to clean up corrupt data froma previous system conversion and helped map data to appropriatefields. Our new claim system used referential data integrity rules,which helped identify errors and discrepancies, enabling us tocatch and correct them ahead of time.

After cleanup, we planned to have our existing system down for aweek for the conversion. This allowed us three days to upload ourdata into the new system and two days to test it. During the systemdowntime, Downey utilized an interim plan to handle day-to-dayclaim operations.

Once the data was converted, we performed thorough data testingto ensure important fields and files were transitioned and verifiedthat data integrity was acceptable. Once again, we utilized ouraction plan to manage the testing phase, tracking who would testwhat and when.

Our claim staff verified data by running through all of thefunctions in the claim life cycle — entering a claim, setting itup, and making payments. This ensured everything was in workingorder.

As planned, we went live on September 9, 2004. In the secondphase of implementation, we rolled out document imaging and billreview in April 2005. Our implementation team had laid a solidfoundation for success, and our vendor played a key role in keepingus on schedule and providing industry and technical expertise whenneeded. If your organization utilizes a similar approach ofstrategic planning, coordinated teamwork, and strong leadershipskills, you can expect a smooth journey through the data conversionand implementation process, too.

Louise Sparks is a claim manager at Downey Insurance, adivision of Brown & Brown, Inc. She can be reached [email protected].