Published: 9/26/2007
According to a newly published study, enterprises considering wholesale replacement of current mainframe applications using Adabas and Natural may achieve greater return-on-investment (ROI) from the targeted use of legacy modernization to ‘update and extend’ – rather than ‘rip & replace’ – these existing mission-critical systems. The study was conducted by Forrester Consulting on behalf of Software AG, a global leader in business infrastructure software, who commissioned the research.
In their analysis of four current Software AG customers using the company’s Adabas transactional database management system and its Natural programming language for most or all of their critical applications, Forrester determined that legacy modernization would produce a five-year return-on-investment of 331% on a risk-adjusted basis, or €3,986,414 at net present value, with payback achieved in less than four months. In speaking with these customers, Forrester found that “[r]ip and replace initiatives were seen as less compelling, and unlikely to be cost-effective, after investigation into the relative ease of Web-enabling existing Natural applications supported by data in Adabas.”
Customers interviewed for the study included the European subsidiary of a global automaker; a health insurance provider operating in the United States; a federal agency of the Australian government; and a French financial services firm. Based upon their real-world decision criteria, use-case scenarios and cost-structure, Forrester Consulting undertook a thorough analysis of the business case for both ‘update & extend’ and ‘rip & replace’ using its Total Economic Impact (TEI)™ Methodology. The findings were presented in the form of a composite case study.
Forrester found that the case for legacy modernization for the interviewed customers, in terms of return-on-investment, was most often the better choice over wholesale application replacement. Furthermore, as users began to factor in the risk factors associated with replacing proven mission-critical systems with untested solutions, these arguments became even more compelling. The study also pointed towards additional value within the current implementations that could be easily unlocked using existing, off-the-shelf tools, such as those associated with service-oriented architecture (SOA).