Today, Glen Alleman posts about dealing with uncertainty in projects.
I’m never disappointed in Glen’s no nonsense approach, uncertainty is a fact, adults deal with it. But I want to make a separate related point.
“As well, uncertainty arises from the basic processes of work
- This is Deming uncertainty.
- It is the statistical “noise” built into the work process.
Both of these sources of uncertainty impact cost and schedule
- Trying to control the “noise” adds little value.”
In Statistical Process Control, (SPC) we try to eliminate assignable or “special” causes that make the process unstable. Assignable causes are accidental rather than essential. If you remove the assignable cause, your process still works as intended, only with more predictable results.
When the variation looks like noise from a consistent distribution we have a “stable” process. Any specific observation has uncertainty bounded by the natural range of the process. This variation is an essential feature of the process. (whereas assignable cause is a “bug”). Trying to control the noise doesn’t just “add little value”, it’s positively harmful. Deming has some nice examples of how chasing noise makes performance worse.
One can, however, sometimes reduce the noise. Since the noise is an essential feature of the process, you must change the process to reduce the noise. Changing the process for the better (either to improve performance or predictability) is what process improvement is all about. Unless you define your process, actually use your process, and measure your process, you have neither a basis for improvement, nor can you claim improvement with a straight face.
It all comes back to process.