The Four Foreboding Factors of CMDB Initiative Failure
Ten years ago, I helped launch their CA CMDB product. We built our Configuration Management Database (CMDB) from the ground up to be a strictly ITIL process centric solution because we saw every other commercial offering at the time was flawed in some way. Some had no formal change management process to approve updates and additions to the database, some supported operational systems which were proud to be described as being real-time, while others were simply asset management databases that provided little support for change and configuration management activities.
CMDB technology has evolved, along with IT data quality awareness over recent years. Despite this, CMDB initiatives continue to be challenging. Understanding why these initiatives often fail will help readers make better decisions and avoid some of the pitfalls. The core reasons for failure of CMDB initiatives can be attributed to four factors:
- Solution Definition & Organizational Scope. Managing scope-creep is as important to a CMDB initiative as it is to any other IT project. The scope must be clearly defined and managed. Whenever there is a change in your scope, the team needs to reset all expectations and communicate them often.
- Data Quality & Accessibility. For me, this is the primary reason why CMDB projects fail. Maintaining the quality of the data in the CMDB is both the hardest and most important factor governing its success. Fortunately, the technology needed to manage data quality at-scale now exists.
- Lack of Governance. Implementing a formal governance model can be challenging, but is essential to success. The Service Asset and Configuration Management (SACM) Governance model enables intelligent decisions and improved business outcomes. Without it, data quality suffers and undermines the entire effort.
- Expected Business Value of Return. Most initiatives do not adequately speak to the business value that is expected of the CMDB implementation. Descriptions of the anticipated value of a CMDB need to be expressed in business terms to the partners who are funding the effort, otherwise it will fail.
Hopefully the above points resonate with your experience. I would love to hear of other factors you have observed. The above factors are spelled out in a paper that Blazent recently published. You can download the Failed CMDB Initiatives White Paper for a deeper discussion on this subject.