How well does your company culture embrace the concept of Plan-Do-Check-Act (PDCA)…or have they forgotten what those letters truly stand for? It's challenging in our fast-paced world where we have demands from our internal customers, external customers and ourselves for near “instant” results to stop and follow a disciplined approach that (conceptually) takes time to follow and complete properly. 
What happens in the Plan step? The Plan step is similar to the Define step in the DMAIC process. This is where we determine what problem we are trying to solve (measurable problem statement) and the data we have to support the problem. Without baseline data to support your problem it is challenging to determine if the subsequent changes you make in this process will make things better or worse. In the Plan step you can also use other Lean tools such as a fishbone diagram, affinity diagram, etc to identify the root cause and determine the course of action you will try first in the Do step. Don't forget about any metrics that may be required for this process that could be included in your problem statement. How much time should you spend in the Plan step? As much time as you need to in order to clearly define your problem and gather your corresponding data. As a general rule of thumb, I normally suggest spending 50% of your time in the Plan step.
Next we have everyone's favorite step, the Do step of the PDCA cycle. This is where the rubber hits the road and we can actually start to try to fix our problem. We will try our first implementation of a change that we believe will permanently solve the problem based on all of the research we did in the Plan step. It is important that there isn't scope creep and we don't try to “boil the ocean” in this step. We need to focus on the problem identified in the Plan step to ensure that we have a way to gather data from this change.
The Check step is where the original baseline data and your new data from your change in the Do step come into play. You will evaluate the data sets and metrics to see if you received the efficiency gains that you desired, reached your performance goal, removed the waste, eliminated the issue, etc. You can do this through control charts, Pareto charts, etc.
Now it's time for the Act step. Based on the outcome of the experiment you just completed, you will either make this a permanent process change, update stand work, etc or repeat the PDCA process and start again. Start again? Yes, if what you did in the Do step didn't pan out when you reviewed the data in the Check step, go back to the Plan step and review your fishbone diagram (or other Lean tool you used) and determine another possible root cause to the problem. At first you don't succeed, try, try again.
What happens when you don't follow PDCA? When I teach this concept, I tell people it's like going to target practice and following the steps: Ready-Fire-Aim. What does that get you? Normally an off target response and an unhappy customer.
Does PDCA take time to complete? Yes, initially it will take your teams time to complete and get familiar with the process until they develop “muscle memory” and PDCA becomes automatic. The best way to do this is through training, application and promoting the utilization of the tool.  It just takes a few courageous Lean innovators to break the cycle of the way your company may have always done things to start something new!
3 Replies
I think that different methods embrace PDCA in their own fashion.  Agile and Sprints are managed in a cyclical way that also incorporates planning, execution, verification and validation, along with reviews and retrospective analysis to apply lessons learned toward future increments.

Also, the Do and Check phases are not necessarily sequential but frequently performed concurrently or intermittently with each other.  If Do and Check are seen as distinct and mutually exclusive phases, then the adoption of such a system would be impractical in 2020.  

Hi Katie, 

Thank you for sharing this PDCA refresher. Your opening sentence rings very true with me on a personal level based upon past experiences.

PDCA can be compared to numerous other structured problem solving methodologies and finding ways to convey this to others who question how "lean" PDCA is can be very beneficial in helping mold the very corporate culture you alluded to. As you've mentioned already, the Plan phase is similar to the Define step within the DMAIC methodology. The Plan phase further encompasses the Measure and Analyze phases of DMAIC methodology as well. Given my automotive experience, the 8D methodology is another method I have guided teams in using. When queried on the comparison of the seemingly "simpler" PDCA approach vs. 8D, I often respond that they are actually very much the same, with the Plan phase representing D1 - D5 of the 8D methodology. The inputs and outputs required of each are the same. This sometimes helps others understand the magnitude of the Plan phase and more easily digest the recommendation to put 50% of the total time into this phase.

The significance of the Plan phase is further demonstrated in your reference to avoiding scope creep. Two of my favorite tools to use for this purpose are the IS-IS NOT tool, and the 5W2H method. By clarifying what IS or IS NOT part of the problem, or quantifying the problem by answering who, what, when, where, why, how, and how many, we can develop a concise problem statement that can be referred back to as needed to help minimize the risk of scope creep.

The point I personally appreciate most from your post is to ensure the team goes all the way back to the Plan step if the countermeasures implemented in the Do step did not achieve the targeted outcomes. When leading teams through the importance of this, I sometimes use the analogy that if we don't return all the way back to the Plan phase, we have a higher probability of "stepping in the doo-doo," or getting stuck in a never-ending Do-Do phase loop.

I've previously been asked by people at the senior leadership level whether or not I believe PDCA is truly a lean tool given the time it can require to be executed well. I often respond that I do indeed believe it is lean and add that I also find the 8D and A3 methodologies to be lean. All three of these methodologies are firmly founded on the Deming Circle/Cycle/Wheel. I then ask them a question in return: Which is more lean: Giving the team the latitude to follow the PDCA methodology successfully and ideally succeeding on the first attempt, or not affording the team the time required to do it right the first time and thus leading them down the path of doing it over multiple times? Which approach brings more value to the organization?

Hi Katie ,
My field is pharmaceuticals, radiopharmaceuticals  and most recently  cosmetic industry . I found that ISO quality management method  though very applicable to cosmetics , like ISO 22716, are not directly applicable for pharmaceuticals;  the notions of "off target results" and an "unhappy customer" are very different from other industries. The regulatory compliance directives are dominant to lead the way of quality management.