Your Repeating Specific Weaknesses: Finding Them, Why They Are Bad, Eliminating Them, and Keeping Them from Coming Back - Integrating CMMI and Agile Development - page 243

Again, you can't use my weaknesses—too many organizations try to do this.
Many just try to address everything in the CMMI model equally, or pick data
to monitor because they have heard it
is what other organizations are using.
Recall the lesson from LACM in
Chapter 3. I actually found myself
slipping into this common weakness
multiple times with my golf project.
That is, I tried to work on too many
different things that really were not
key to my unique weaknesses. I
found that real improvement resulted
when I limited myself to primarily the
core issues.

How to Find Effective Checkpoints

Use the following criteria to help iden-
tify effective checkpoints to counter
your repeating specific weaknesses:
• Nonintrusive

• Support rapid feedback
• Support continual small corrections

At GEAR, 10 the signs of trouble often do not occur until integration, but the
root cause is traceable to the seemingly imperceptible “small” decisions that
occurred along the way. Examples at GEAR include decisions related to com-
pletion criteria for a design, or completion criteria for a peer review. Many
people knew the “minimum acceptable standards ,” but because they had not
explicitly documented them or put the processes in place to enforce them, it
became too easy to allow one situation after another to slip by without meet-
ing the known organizational standards.
Each decision on its own often did not appear to be significant. A specific
example I heard about during the gap analysis at GEAR was a decision to
move forward with coding a software design despite open issues that had
not been resolved relative to ambiguous requirements . I also heard about a
decision to proceed into coding despite a design review not having been
held. In some of these cases, the decisions could be valid given specific cir-
cumstances, but without a clear written standard, and an objective
compliance process, it became too easy to allow such cases on a regular
basis with no one stepping back to observe, measure, and report the cumu-
lative effect on the overall project.

Like my golf swing, these gradual almost imperceptible variances from the
planned process tend to have a cumulative effect that we often don't notice
until the project reaches a point where we no longer have control to bring it
back into alignment.

10. The GEAR case study is discussed in Chapter 7.