Tool 16 · Control

Fill in a PDCA cycle

Plan, Do, Check, Act: four boxes, each with the question that matters. Once a cycle is complete, you start the next one from Act. Print it as a card or copy it as text.

PDCA cycle
Saved automatically in your browser
You start a new cycle at the bottom, under Act.
PPlan
DDo
CCheck
AAct
P

Plan

What are you going to do and what do you expect to happen? That is your hypothesis.

D

Do

Test your hypothesis: carry out the plan and watch closely what happens.

C

Check

Compare the result with what you expected in Plan.

Your expectation from Plan: Lead time from receipt to payment goes from an average of 30 to 14 days, within 8 weeks.

Did your hypothesis hold?
A

Act

Standardise what works, or start a new cycle.

What do you do now?

This cycle is complete.

Start the next cycle under Act. Every round makes the problem smaller.

How to run a PDCA cycle

  1. Plan. Choose the problem, write down what you are going to do and what you expect to happen as a result. That expectation is your hypothesis, and it should contain a number: from how much to how much, within what period. Also agree on how you will measure it, before you start.
  2. Do. Carry out the plan and watch closely what happens. Keep track of what you actually did, even if it deviated from the plan.
  3. Check. Put the result next to the expectation from Plan. Did your hypothesis hold? That is the question, not "has it got better".
  4. Act. If it works, you standardise: record what and where, otherwise it is gone again in three months. If it does not work, or not well enough, you start a new cycle with what you have learnt.

When do you use it?

Anywhere you want to improve something and want to know whether it works. In an improvement project the PDCA belongs to the Control phase: the new process is running, and you keep an eye on whether it keeps performing. But it works just as well on a small problem in your own team. The iteration is in the name: one cycle is rarely the end.

Common mistakes

  • A goal without a number. "Faster" or "fewer errors" cannot be tested in Check.
  • Do without Plan. Just trying something is not PDCA. Without an expectation you learn nothing from the result.
  • Skipping Check. The temptation is to move straight on after Do. It is precisely the comparison with the expectation that creates the learning.
  • Forgetting Act. An improvement that has not been recorded disappears as soon as attention slips.

Tip from our e-book: track progress on a Kanban board with the columns Plan, Do, Check and Act instead of To Do, Doing and Done.