Methods and phase counts
| Method | Phases | When |
|---|---|---|
| PDCA | 4 | General purpose; known cause, daily improvement. |
| PDSA | 4 | Improvement by trial; "study" instead of "check" puts the weight on learning. |
| SDCA | 4 | Holding a gain: standardise first, then do. |
| OODA | 4 | Fast-changing situations; emergency decisions. |
| DMAIC | 5 | Data-driven, measurable, recurring process problems. |
| 5S | 5 | Order of the workplace; visual management. |
| A3 | 7 | Problem, analysis, countermeasure and follow-up on one page. |
| 8D | 9 | Serious, recurring or externally reported problems; team discipline. |
| Kaizen event | 3 | Short, focused workshop. |
The selection test
Two questions are enough: is the cause known and does it recur? Choosing a heavy method for a one-off deviation with a known cause turns work into form-filling. Conversely, closing a recurring problem with PDCA means repeating the same fix every year.
What they share
Whatever the name, every method asks for the same three things: a concrete statement of the problem, an analysis that reaches the root cause, and a countermeasure whose effectiveness is later verified. Without those three you have a template, not a method.
Is choosing a method mandatory?
The standard does not require a particular method; it requires the result. A method exists to make that result repeatable.