Friday, April 25, 2008

Developing Questions for the CMMI Specific and Generic Goals Would Lead to Implementing Alternative Practices

I would be very CAUTIOUS about taking this approach to implementing the CMMI. The CMMI authors did not go to the lengths they did with defining the Expected components (Specific and Generic Practices) if all they really wanted an organization to do was start at the Goals and identify the supporting practices. In my experience as a Lead Appraiser, it is only on the rare occasion that I have identified and documented an alternative practice in a SCAMPI. 99.999% of the time, the organization’s practices are aligned with the Expected components. And I assert that the only time an alternative practice comes up is with legacy projects that are years in duration using practices that are no longer state of the art and there is no business value in changing them. For example one of the underlying assumptions of REQM is that an organization is using tools to help manage and control the project’s requirements. But if the original requirements were established before the advent of requirements tools and the organization’s documented process for reviewing and managing requirements was to manually review, trace, and track requirements through meetings and discussions re-deriving the traceability each time and the process works, it is an alternative practice. There is no requirements database, just a requirements document. There is no requirements traceability matrix or requirements tracking system, just the documented requirements and the expertise of the requirements analysts who are highly trained in how to procedurally review, maintain, manage, and control requirements. So it is difficult to evaluate SP 1.1 – 1.5 as FI, but REQM SG 1 is clearly FI.

No comments: