Dr. W. Edwards Deming

Chasing fast and cheap

Organizations have had a choice to make over the last few decades on how to survive. One choice is to improve their current operations. This has led to things like Lean Six Sigma. The other is to outsource. This has led to unemployment.

Both of these are chasing fast and cheap. They are both lazy. Improving an organization looks like hard work and feels like taking a big risk. So, people avoid it. Outsourcing just requires suspending disbelief and fooling yourself. It also allows blaming the employees, who are then layed off without guilt.

An article available from InformationWeek (May 2011, Outsourcing 2011: The Incredible Shrinking IT Team) indicates that most are taking the lazy path and not getting the results they hoped. They are locked into the Frederick Taylor management philosophy and cannot see their way out. They cannot see their way to command and control their way out of their plight.

Readers that see themselves in this position do not see the correct role of a manager. It is to understand the 'system' that everyone works in and convey its meaning and its purpose. To help everyone see their role in the system and how to work together for its aim. To help each person maximize their contribution to the system and find joy in their work. There is more, but note that bossing or commanding is not to be found.

Methods like Lean Six Sigma promise improvement, but they are rarely sustainable. They are a catechism. W. Edwards Deming captured the true faith in his System of Profound Knowledge. Anyone who truly wants to improve their organization needs to learn and employ it. And, that requires unlearning a lot of everything else.

False paths to improvement and outsourcing lead to cost-cutting and the general demise of the organizations that choose them. Deming's SoPK is really the less risky because it has, at least, a chance of working in the long run and being sustainable.

Software Engineering Institute fails to understand process improvement

I contributed the following to a question about the CAR (Causal Analysis and Resolution) practice area the SEI has in their CMMI (Capability Maturity Model - Integration). The person asked

Why is CAR at Level 5...?

Pretty much all the other responses were rationalizations as to why it was there. They all assumed that the SEI was correct. I disagree below.

Everyone seems to have a piece of the answer. But, let me be blunt. The SEI does not understand variation and its role in process improvement. CAR is at the wrong level. The SEI flubbed the CMMI and is trying to fix it without the required understanding. Organizations sense this imperfectly, but often implement Levels 4 and 5 simultaneously to get around the conflict.

In CMMI version 1.3 they just removed Generic Goals 4 and 5 from all the process areas (PA) because they were so messed up. Level 4 should have been 'process stabilization' and Level 5 should have been 'process optimization'. CAR should be in Level 4 along with a PA to evaluate process performance. Level 5 should have a PA for optimization. And, there should be a PA with appreciation for the organization as a system. IPPD tried for this but failed and got pulled.

Unstable processes cannot be optimized. They must first have all their Special Cause problems resolved. Once they are 'stable', meaning they are predictable, their performance can be evaluated. If variation is too much, or the mean is not appropriate, the process improvement can be initiated to change the process. PDSA or DMAIC or DOE or some other process improvement method can be used. The situation or organization will usually dictate which.

Like martial arts, all process area improvement is like climbing a mountain. All the paths converge at the top and the view is great in all directions. Following the CMMI without understanding variation you will just circle the mountain. As a management consultant I try to lead my clients upwards, through the fog.

Rick