BANGALORE, INDIA: Informatica Corporation, a leading provider of data integration software, has announced findings from an independent research study it commissioned Bloor Research to undertake on the data migration market. The total budget for data migration projects undertaken by just Global 2000 companies in 2007 is set to break the $5 billion mark, rising to $8 billion by 2012, yet 64 percent of these projects are not delivered on time and 37 percent experience serious budget overruns that boost project costs by an average of 30 percent, according to the study. The research also reveals how this trend of data migration failures can be entirely reversible.
According to the study, data migration time and cost overruns can be substantially reduced or even eliminated through preventive actions and the application of appropriate disciplines—including data governance, on-going data quality programs, focused migration methodologies, and the up-front use of data profiling and analysis techniques.
“The Bloor study paints a portrait of a fast-growing market pivotal to numerous strategic business initiatives, from mergers and acquisitions to large-scale applications projects, single business views and compliance. Yet it is a market challenged by inadequate methodologies, poor project scoping, and insufficient understanding and usage of available tools,” said Girish Pancha, general manager, enterprise data integration business unit, Informatica. “As data migration projects are invariably keystones in larger, over-arching application projects, it is imperative that proper disciplines, methodologies and tools be embraced so that the broader application projects do not also falter or fail.”
Key findings
The research involved interviewing more than 700 Forbes Global 2000 companies with respect to data migration projects budgeted at $1 million or more. The following are among the study’s key project-related findings:
* Despite the enormous criticality of data migration projects, only 16 percent of them studied could be viewed as unqualified successes—i.e.: exhibiting neither time nor cost overruns.
* Where there were time overruns, they averaged in excess of 40 percent of the total budgeted time—a fact attributed, at least in part, to improper scoping of the project.
* Where there were cost overruns, they averaged in excess of 30 percent of the total budgeted cost—again attributed, at least in part, to improper scoping. With the projects in the study budgeted originally at $1+ million, an average 30 percent cost overrun indicates large unplanned budgetary expenditure.
* A mere 10 percent of companies interviewed used data profiling tools to better understand their data and data sources before scoping their projects—thus exposing themselves to “unexpected problems,” coupled with the inability to establish realistic project timescales and budgets.
* Hand-coding is still the most prevalent “tool” in data migration projects—which can prove to be error prone and inadequate.
Recommendations
Based on these and other findings, the study recommends the following preventive measures and actions:
* Implement data governance across the enterprise so that the location, nature and condition of enterprise data is always properly understood. This includes extending data governance to any acquired organization prior to undertaking a data migration.
* Implement data quality initiatives so that data is maintained at a high level of migration-readiness, in addition to the numerous other benefits of ensured data quality.
* Profile and analyze all data sources in advance using a proper tool in order to fully understand the scope of data issues that will be encountered including how they may impact project timelines and costs.
* Ensure full familiarity with, and training in the use of, any tools, prior to commencing the project.