I'm always struck by the diversity of Queens, New York. It seems every country and culture is on display, and that makes for some tough choices at lunch: falafel, moussaka, sushi, kebab, and knish all on the same block! Like Queens, many data centers are melting pots. Stroll through yours and you're likely to find an array of platforms running a variety of OSs, databases, and legacy applications. Thanks to mergers and acquisitions, heterogeneity has become an accepted norm in most data centers. But that's an unfortunate circumstance.

Managing multiple applications over the long haul has serious drawbacks if those applications perform the same functions. For one, IT departments need broader skill sets and more employees to maintain those applications. Finding competent technical talent is neither cheap nor easy.

In addition, financial reporting is typically burdensome at companies with a hodgepodge of legacy systems. Reconciling data is a non-value-added activity that can overwhelm accountants, financial analysts, and IT personnel alike. Worse, consolidating data from multiple systems can lead to reporting errors. Multiple applications hamper day-to-day operations and customer service too. Suppose your company maintains three legacy applications that serve the same customer base. If business partners, customers, or market conditions warrant new functionality, you might have to implement it in three different ways. Adapting to change becomes much more challenging than it should be. Your company's reputation will also suffer if it can't service customers in a cohesive fashion. Nobody wants to hear, "I can't access that information. Enjoy some elevator tunes while I transfer you to someone who can!"

Finally, multiple systems present hurdles to strategic decision-makers who need unified views of the business. Data warehouses and operational data stores can help, but their price tags rise in lockstep with the number of applications feeding them.

Consolidating legacy applications makes sense. It helps companies simplify data centers, streamline financial reporting, standardize business operations, support decision makers, and provide better customer service. Why then, do so many organizations still rely on redundant legacy systems? The reasons are legion.

1. Short-Term Costs

Migrations are expensive. Modifying business processes, converting data, training users, upgrading hardware, creating documentation, bolstering functionality, and purchasing additional licenses can add up to big bucks. These expenses can keep migration plans on the drawing board for years.

2. Integration Aids

In the short-term, integrating two systems is easier than mothballing one. Thanks to a new generation of middleware and popular file formats such as XML, sharing data has never been easier. A common web interface can be slapped onto multiple back-end applications, and electronic forms software can help make documents look alike, even if their content originates in different systems.

3. Competing Priorities

Migrations require dedication over time and compete with other projects for attention. Your CEO won't be particularly interested in consolidating systems if she's eyeing another acquisition target, for example. Similarly, if your CFO is busy raising capital, a migration project will not garner much support from his group.

4. Risk Avoidance

Migrations force companies to change key processes, which can disrupt day-to-day operations. What's more, conversion errors or gaps in planning can have tremendous ramifications for your company, and your job. Few people look forward to flipping the switch on go-live day. In fact, most of us go out of our way to avoid failure, and that places countless conversion projects on the back burner.

5. Politics

Retiring a key legacy system is a sea change for the folks who built it and for those who use it. Migrations alter the pecking order in many organizations, and not just in the IT department. When coupled with our natural aversion to change, it's not surprising that politics leave many conversion projects languishing at the gate.

6. Lack of Leadership

New systems can affect virtually every employee. Like no other IT project, migrations call for across-the-board support. Without clear directives from the top of the food chain, getting buy-in from diverse constituencies is like herding cats.

7. Lack of a Project Champion with CBA

Migrations require champions who understand the costs and benefits and can present a compelling case replete with comprehensive financial analysis. Many IT managers understand the benefits of consolidation so implicitly, I believe, that they don't bother performing formal cost benefit analyses. This oversight might preclude broader support for the initiative.

Given these reasons for why migrations don't happen, it's easy to understand why so many organizations maintain redundant legacy applications. It's much harder to get them to do something about it. When you get fed up managing multiple systems and want to institute change, make sure you address the points above to ensure a better chance at success.

Scott Steinacher has broad managerial and technical expertise in application development, e-commerce, business intelligence, and document management. He is a cofounder of TimeTrade Systems, Inc. (timetrade.com), a leading provider of web-based scheduling solutions. In addition to several dozen articles, Scott is author of Data Warehousing and the AS/400 (29th Street Press, 1998). He holds an MBA in corporate finance from NYU’s Stern School of Business, and a B.S. in Information Systems from SUNY Albany.