IT Modernization: Planning to Minimize the Risk

by Richard L. Ptak, William Moran
November 20, 2008

The collapse of the bridge in Minneapolis over the Mississippi River in 2007 is just another reminder that our infrastructure is aging. Many of our bridges, water pipes, and roads have been in service far longer than their design called for. In a similar way, our computer infrastructure often consists of code created 20 or more years ago. While code doesn’t deteriorate in the same way physical structures do, it does shows its age in other ways. Code designed many years ago will need to be replaced with more modern code that reflects current capabilities and meets the requirements of today’s business.

Like bridges and roads, much of the old code has been repeatedly patched. In many cases, the programmers who made these changes are no longer available, having moved on to other projects or jobs. Documentation, if it exists at all, may be missing or incomplete. Business rules may be buried in code that no one understands well. In addition, many of the programmers who maintain these systems are scheduled to retire in the next several years. Any of these factors alone can make the task of modernizing these older systems an overwhelming effort.

However, waiting until a crisis strikes can increase the difficulty and provide a perfect opportunity for a high-risk, visible failure. Consider the state of Maine’s attempt several years ago to modernize its Medicaid payment system in response to legislative changes. The state scrapped the old system before the new one was thoroughly tested; the new system didn’t handle claims processing well, so millions of dollars in payments to doctors and other providers were delayed for long periods. This IT failure became a major issue in state politics and reveals common mistakes seen in infrastructure modernization (see www.cio.com/article/20133/Maine_s_Medicaid_Mistakes). Such infrastructure modernization, whether involving a national transportation system or an enterprise business transaction system, poses specific challenges and calls for specific action. 

The state of Maine would have been better off to examine its legacy applications without any external pressure, but you can’t always avoid surprises or evade legislated deadlines. However, you can take some actions that help to minimize risk; let’s review some positive steps you can take.

First, ensure the enterprise has a business plan and that the IT requirements flowing from that plan are clear and understood. Determine whether or not legacy applications can meet these requirements. There are tools available to help with this decision (BMC, CA and IBM Tivoli, for example, provide tools that can help by simulating the effect of increased workloads).

If the legacy system won’t meet the requirements, then an estimated budget needed for a modernization effort must be developed and reconciled with what actual funds can and will be made available. Several issues will determine these values, including the risk to the business of a system failure, competitive pressures, system stability, etc. The overall business environment also will influence the timing of the project, since that can influence the available budget.