A coaching tool for teams · after W. Edwards Deming
A continuous improvement loop. PDCA stands for Plan, Do, Check, Act. The team picks a single improvement, tests it on a small scale, analyses the results and decides what to keep.
About this exercise and facilitation guide
PDCA, also known as the Deming cycle, is the classic framework for continuous improvement: instead of changing everything at once and hoping it works out, we frame one small change, predict what will happen, test it, then compare the result against the prediction. We keep what works; what doesn't becomes the next experiment.
The power is in the repetition
A single cycle rarely solves the problem. PDCA gains its power when the team closes one cycle, learns something and starts the next one straight away, aiming slightly higher. That's why the exercise ends with a decision and with what to test in the next cycle.
Facilitation guide
One improvement per cycle. If the team brings three problems, pick the one with the biggest impact and keep the others for a future cycle.
Don't start without a measurable success criterion. "Things should get better" can't be checked; "the meeting takes 45 minutes" can.
Ask for a prediction at Plan. The team learns the most when it compares what it predicted with what actually happened.
At Check, compare against the criterion — don't look for someone to blame. A poor result is information, not failure.
Always close with the verdict: adopt, adjust or abandon. A cycle without a decision is just a conversation.
Your data stays on your device. Nothing is sent anywhere.
Our experiment
Before starting the cycle, we name the change we're testing and how we'll measure success.
What do we want to improve?
How will we know it worked?
P
Plan
We describe the situation as it is, choose the cause we're tackling and set out the steps. A small, testable plan — not a strategy.
What small change could we test, and what do we expect to happen?
Questions to explore
What does the situation look like right now, in facts and figures rather than impressions?
What do we believe the main cause is, and what other causes have we considered?
What exactly are we changing, and what do we predict will happen if we're right?
How long does the test run, and who tracks the results?
The steps of the experiment
What we'll do
Who
By when
D
Do
We put the plan into practice, on a small scale, and note what happens along the way — including what wasn't in the plan.
What did we actually do, and what came up along the way?
Questions to explore
What did we manage to do exactly as planned, and what did we have to adapt as we went?
What surprised us along the way?
What data or observations did we gather while the test was running?
C
Check
We compare the result against the success criterion and against the prediction from the plan. We're looking for the lesson, not for someone to blame.
What do the results tell us, compared with what we predicted?
How close are we to the success criterion? (0–10)
not at allpartlywe've met it
Questions to explore
Does the result confirm or contradict our prediction?
What went better than expected, and what fell short?
What did we learn about the real cause of the problem?
A
Act
We decide what to do with what we've learned: keep the change, adjust it and run another cycle, or let it go.