LC HomeConsultingSTARSPUBLICATIONSTrainingLC SHOPHan van loonNewsContact UsStarswebworxPhotoPhoto web

Subpage menuPeople Process Product
12 Thinking Modes
Leadership and Management
Star people
Star Managers
Star Leaders
STARS Knowledge Management
STARS Socio-cultural management
Team Building Techniques
Symphonic Collaboration
STARS and complexity science
Change Networks
Problems with PDCA
Improved PDCA
Walter Shewart
W Edwards Deming

Problems with PDCA

PDCA is the acronym for the Plan-Do-Check-Act cycle. See here for a brief overview of PDCA/PDSA. There are several problems with PDCA or PDSA. While some may appear trivial, they can have strong cultural, personal and enterprise implications. They are described below and summarised here.

First, PDCA oversimplifies the entire process of improvement. The reason for this lies in its origin. The Japanese absorbed the lessons taught by Walter Shewart for how to improve a tight loop production process with process control using continuous measurement and statistical control techniques. Here I mean by a tight loop production process the set of activities in one production workstation, and not across multiple workstations/processes. They naturally described this in Japanese and not English. They were subsequently asked to described their quality control loop in English (note that even today most Japanese have poor command of English and it was even rarer for Japanese in Shewart's time in Japan to be proficient in English) and came up with the simplest possible explanation which became: Plan Do Check Act. In fact the actual cycle is more detailed than that in the original japanese version. In real life, trying to use PDCA rarely succeed because it is a simplified abstraction and does not reflect the various activities required for successful improvement activities. For example, most serious improvement cycles have aspects such as sponsorship, leadership, coordination, communication, education, benchmarking and establishment of the modified process. In the ISO 15504 suite of standards, this need to handle many aspects in an improvement program results in a nine phase improvement process (see Team Based Business design Improvement and Process Assessment and Improvement book for realistic improvement approaches).

Second and related to the first problem, there is no real way to have continuous improvement. By definition, continuous means without  interruption. If an enterprise or any part of it is subject to continuous improvement, this implies continuous change. In such a situation people will become rapidly confused about the current status of processes and procedures and/or follow outmoded practices. There is also the real problem of "change fatigue": when people who have to adjust continuously to change (improvement) then become confused, disenchanted or stop following changes. Due to change fatigue, progress tends to grind to a halt within an enterprise. Change requires motivating people, and hence needs to recognize change fatigue and manage it. Change management needs to cover people, process and product.  In reality, Shewart and Deming meant Continual Improvement, a cycle of improvement with pauses to consolidate the improvements (changes). These are then subject to checking or study to determine if they have actually improved the area of change, before the next improvement is started. At enterprise or organizational level, improvement is a time and effort consuming activity that often requires substantial resources as well as time to consolidate change. In fact, the international standards (ISO) corrected this in the previous version of ISO9000, changing continuous improvement to continual improvement in ISO9004:2000. In ISO15504-Part 4, a similar position concerning improvement is described.

Third, Do and Act have the same meaning in English. The Compact Oxford dictionary provides the following relevant definitions:

  • Do  • verb 1 perform or carry out (an action). 2 achieve or complete (a specified target). 3 act or progress in a specified way. 4 work on (something) to bring it to a required state. 
  • Act   • verb 1 take action; do something. 2 take effect or have a particular effect. 3 behave in a specified way. 

So PDCA could just as easily be PDCD or PACA! In other words: Plan-Do-Check-Do or Plan-Act-Check-Act!  This is confusing, and PDCA can thereby seem trivial especially to uninformed personnel. In reality, the real meaning of Act is Improve or Correct (when failing to achieve the desired state - see definition 4 of Do above)! Hence the cycle should read Plan-Do-Check-Improve or Plan-Do-Study-Improve. So the acronym would be PDCI or PDSI! Why did not Shewart and Deming improve their PDCA cycle? Because they felt it better to have people at least remember a simple idea about quality improvement rather than none at all.

Fourth, Shewart assumed that management or experts would act, not the ordinary worker. This is completely outdated thinking in most enterprises today.

Fifth, Plan has a limited range of meaning [Compact Oxford: • a detailed proposal for doing or achieving something. 2 an intention or decision about what one is going to do. ]. It  was not intended to cover aspects such as creative or innovative thinking (also called "outside the box" thinking) or handle complex adaptive systems. There are at least 12 modes of thinking, and Plan only covers a couple of these modes of thinking. This is an unrealistic restriction to the possible ways to improve.

Sixth, PDCA has an inherent circular paradigm (look at the Wikipedia entry if you don't believe me). It assumes that everything starts with Planning. I know that I do not plan everything. I challenge you to Think about times you did not plan something. Did you take an action that was based upon thought but no planning? Was the outcome acceptable or unacceptable? I can recount many times that I did something without planning that was successful, but I cannot think of any time that I did something without some thought (even if it was reactive thinking). In fact, in areas such as innovation and invention it is sometimes better not to plan because this can stifle creativity! So I caution people to avoid assuming that PDCA is only used in a circular mode, in other words do not assume you must 'Plan Do Check Act'. It is possible to improve without doing everything in the assumed circular order [I may be branded a heretic for stating this :-) ].

Seventh, PDCA was only intended for incremental continual improvement, hence it is not a good methodology for large scale change, especially in large complex organizations.

Eighth and most importantly, people are not explicitly mentioned in the basic PDCA cycle. There is an assumption that people will use the cycle and make things happen. I am sure I am not the only person to have noticed that this assumption is not always true! Naturally, motivated people will look to improve but many people are neither motivated not interested in improvement. In fact the way people act and improve has strong cultural foundations. The modern Japanese worker has a completely different attitude to improvement from their forebears of little more than half a century earlier (when workers did their job and were not allowed to suggest changes, and Shewart created PDCA). In that era, only senior managers could suggest and make changes. The modern Japanese worker of today also has a different attitude to improvement to an American or Portuguese or Russian or German worker today. The cultural differences mean that modern Japanese methods like the Toyota Production System that has been translated in America into Lean Manufacturing do not work as they do in Japan. One only has to look at the differences between the same products built in Japan versus USA to be aware of this (cars, TVs, etc.). This is also true between German and US car plants. The differences have little to do with the processes and the equipment(products) used, and everything to do with the people implementing the processes and running the equipment. It is the way people interact that determines how well processes run and products are created!

Therefore, as Deming himself encouraged, we should not just accept what we are told, but should learn, understand and intelligently adapt  and improve what we learn to suit what we want to achieve, including PDCA. Use PDCA when it is suitable, use another method (STARS, DMAIC, IDEAL, TRIZ, TBBDI, EBBDI) or another form of Improved PDCA (more here) when PDCA is unsuitable.

This page and related PDCA pages are published under GNU  Free Document license provisions.

Your thoughts:
Your comments:
Contact me:

Copyright LC Consulting, LC Publishing © 2005 - 2016