A UK based market research firm recently reported that 81 percent of master data management (MDM) projects are now successful – compare that to 2008 when that number was only 54 percent. However, there is still plenty of room to grow and the mistakes still being made come down to a mix of soft and hard skill aptitude. Surely the IT staff put on to MDM projects are capable of doing the work, but they have to also follow through on it and communicate changes to the appropriate areas of the business. However, most IT organizations have some difficulties when it comes to those areas.
For instance, many master data management projects fail because the companies implementing them are underprepared for organizational changes to business processes ““ they have failed to communicate the changes, define the requirements or what success looks like, or are simply unclear on what the ROI is. Furthermore these projects fail when security requirements are neglected or companies overlook the necessary patching and fixing of the implemented MDM solution.
One mistake that is often made concerns the hardware: don’t leave the decisions concerning what hardware is needed for the last minute:
The definition and acquisition of the hardware requirements should be done as early as possible when the MDM project starts. Hardware requirements must be specified and hardware must be configured for various environment including development, testing and production hence can be time consuming.
Furthermore, don’t leave user acceptance testing for the end of the process. It should be carried out throughout the process of the project, allowing for stakeholders to provide feedback and the team working on the MDM project to take that feedback into consideration as the project progresses.