The Configuration Management Database (CMDB) is the most expensive and wasteful dead end in ITIL. The name itself entices technology managers to ITIL implementation failure like moths to a flame. ITIL with its metrics and management concepts is just so much fluff to most traditional technical people and IT managers. It is an unfathomable set of theories with nothing ‘real’ for technical people to grab onto, everything that is except the CMDB. A configuration management database, now that is something that IT can build.
This is how the money pit is birthed, a desperate attempt to demonstrate an active ITIL initiative. Projects are created, initial schemas are designed, and coding begins. Then the scope creep begins. First the schema must be expanded as complications arise. Other groups become involved and the schema expands again. Next entire existing data stores are identified that must be included. The project expands to include rewriting existing database applications. More resources are consumed and other IT projects are delayed. Results seem always just beyond the next scope change. Eventually there are no more resources, the organization becomes disillusioned with ITIL and IT goes back to business as usual.
Why do so many IT managers fall into this trap? Everything in ITIL depends on the CMDB. It seems a natural enough place to start. However, the problem is that a straight line cannot be drawn between the CMDB and business value. To be successful with ITIL implementation, projects must demonstrate business value.
ITIL discusses a concept called Value Chain Analysis, which provides a method for associating every action or task with a business goal. The idea being that, if IT cannot draw a link between an IT task and a business goal then that task is wasteful and should be eliminated. Of the eleven core processes of ITIL, ten can be directly connected to business goals. Configuration Management, it turns out, stands alone in that it does not directly impact any business goal.
To be fair ITIL does make an attempt in identifying goals for Configuration Management that state “To be efficient and effective, all organizations need to control their IT infrastructure and services.” It also identifies some goals for Configuration Management which essentially say that Configuration Management is a support tool for the other ITIL process areas. Herein lies the problem with beginning an ITIL project by creating a CMDB.
The CMDB is a support tool. It is impossible to develop an effective support tool without knowing the requirements of the process that is being supported. This is the source of so much of the scope creep that kills CMDB projects. They are started as an attempt to design and build a generic all encompassing support tool for IT. A project of this nature effectively provides no boundaries for what is to be included in the CMDB and what is not. Further, when projects are designed this way, determination of whether they delivered business value or where a money pit only becomes apparent upon significant completion of the entire project.
IT support tools definitely add business value in every organization. However, the business value they add is decidedly indirect. Support tools allow IT staff to more effectively and efficiently perform specific tasks. When those tasks are linked to business value they in turn provide a link for the tool. In this way the business value driven by the use of an IT support tool can be directly identified and compared to the cost of implementing and operating the tool. CMDB projects that aim simply to create and implement a CMDB are missing this vital link. Therefore project expenses are ultimately unjustifiable by the business.
Too many IT departments are putting the cart before the horse when building a CMDB. It is no wonder that they have difficulty selling carriage rides to the business.