Few people command the same reverence in manufacturing circles as Edward Deming.  Though he was initially shunned in the United States, his work in Japan earned him substantial notoriety and prestige.  His model of Plan, Do, Check, Act, which was later adjusted to Plan, Do, Study, Act, has been a staple of continuous quality improvement programs for decades.  While this change model is very effective at the micro level, it’s efficacy for large-scale or human-centric processes has very mixed results.


The cycle is noticeably simple.  You plan to do something, do it, study – or check – the results, and act on what you’ve learned.  Deming reportedly changed the word from “check” to “study,” because he didn’t like the mechanical nature that the check implied.  The power of the cycle isn’t in the single execution of it.  The power of PDSA is the flywheel effect.  Once you get used to using it, you can use it continuously to make incremental improvements in a process repeatedly.  The results over time can be powerful.


Lean manufacturing came out of an offshoot of Deming’s work and is notable because Lean is commonly implemented with PDSA.  The direct descendant of Deming’s work was the Toyota Production System (TPS), which included continuous quality improvement through the implementation of PDSA.  It also included the concept of reduction of waste.  Said differently, anything that doesn’t add value to the process should be removed.

Lean, like PDSA, is focused on continuous improvement of processes relying on the compounding effect to deliver sustainable results over time.

Macro Changes

While PDSA is effective at micro changes, the mechanical model that it’s based on struggles at large-scale changes and organizational changes, because those problems are more like wicked problems.  This means the study aspect of the cycle is broken, because the learnings that are achieved are no longer usable by the time they’re learned.

More challenging is the fact that the cycle makes no effort to address the human or unknown aspects of any problem.  Emotions and perspectives play no part in PSDA, and as a result, humans and emotions tend to get in the way of the implementation of changes produced in a PDSA cycle.

Sometimes the kinds of changes that are needed in an organization are transformational.  Such changes are not achievable via PDSA.


The State of Indiana went bankrupt – once.  Bonds were issued for the development of the Wabash and Erie Canal among other improvements.  It was designed to make transportation in the state much more efficient and safer.  It would have done that, except it was introduced at a time when canal-based transportation was being replaced by railroad transportation.  No amount of incremental improvement in canal-based transportation would have been able to address the innovative improvements that the steam engine-powered locomotive would produce.

PDSA is therefore effective when the process is performed frequently, doesn’t change much, and isn’t overwhelmed by a competing approach that offers innovative rather than incremental improvements.