As businesses grow and become more complex, the information systems that support them require continuous updates to keep up with the changes.  Moving to an ERP (Enterprise Resource Planning) system doesn’t reduce the need to change.  The implementation of the ERP system reflects a static point in time, and supports the business processes and data requirements at the time of implementation.  Supporting these “dated” business practices costs money, and more importantly doesn’t allow a company the agility to change as the business environment changes.  The difficulty of changing the myriad of complex systems, and the interrelationship among the many enterprise systems, leads many companies to just leave things as they are, quickly giving the current ERP system the characteristics of a legacy system.  When do ERP systems become legacy systems?

Legacy systems are often thought of as green-screened hardware comprised of applications that required hard-coding and continuous additions to that code to provide new features, lack of documentation for the changes, and unenforced data integrity.  For example, database field formats may differ in many of the different applications or modules.  Consider that a large part of the Y2K crisis was a result of format inconsistencies in the “date” field of legacy applications across the globe.  Billions of dollars were spent making this single modification, and thousands of similar issues exist across the many systems of today’s enterprises.  Forrester Research reports that “76 percent of IT budgets [are] earmarked to maintain existing applications.”¹

Both the complexity and lack of quality of these systems force an abundance of ongoing maintenance and make it difficult for them to support agile business changes.  Even within a single system, there may be over 2 million columns, each of which could be related to any other creating a cascading maintenance effort that results from making a single change to a specific part of the system.  Making one change requires making changes everywhere in the system due to the multiple relationships, potentially resulting in millions of changes that must be made if a company, for example, changes its business from a product business to a service business.  This complexity is exacerbated many times over when companies have not just one ERP system, but many integrated systems with inconsistent data.  A change in one system impacts many different systems, often in unpredictable, undocumented ways.

“The Hackett Group estimates that the average $1 billion company maintains 48 financial programs, along with nearly three ERP systems,” according to CFO magazine. “So it’s little wonder,” says Randy Whitchurch, CFO at bar-code maker Zebra Technologies, that “if you’ve got a lot of far-flung locations on disparate accounting systems, [documenting controls is] a problem.”²

It can easily take hundreds of people to extract and refine data from multiple sources, or coax the information from multiple modules within the ERP system and reconcile them to the current business processes. Employees build elaborate spreadsheets, reports, and data warehouses that need to be maintained, compounding the initial reporting expense over time.

Expensive, time-consuming changes make an organization rigid and not agile enough to accommodate the ongoing nature of change.

The problem is, once an ERP system has been implemented, changes to the system are necessary if the data is to correctly represent the business and its processes.  There is no way around it.  So, when do ERP systems become legacy systems?  I think the clear answer is the moment after implementation.

1 Murphy, Phil. “APM Tools Will Reach $500 Million to $700 Million By 2008.” Forrester Research. July 22, 2005.

2 Goff, John. “Sarboxing.” CFO Magazine. February 1, 2004.