PDCA Template: Plan-Do-Check-Act Worksheet
Test one change: predict, try it small, check, then decide.
Example loaded: scrap on Line 2.
Filled boxes are tinted. Click a box to go to it.
Get the Plant Floor Briefing: one free template like this every 2 weeks
One guide, one free tool, one problem fixed. Every 2 weeks by email. No spam, unsubscribe any time.
How to run a PDCA cycle
- Plan: state the problem and the test. Current number, the cause you suspect, the change you will try, and what you predict will happen, with a number.
- Do: test small. One shift, one line or one product, for a set time. Record what you actually did and anything that went differently.
- Check: compare with the prediction. Before and after on the same metric, measured the same way. Note what you learned and any side effects.
- Act: decide. Standardize and train if it worked. Adjust and run another cycle if it partly worked. Drop it and test the next idea if it did not.
PDCA example: scrap on Line 2
One filled cycle from a machining line. The prediction was only partly met, which is a normal and useful result.
| Phase | What the worksheet says |
|---|---|
| Plan | Scrap on Line 2 is 4.2% (4 weeks of data). Suspected cause: coolant concentration drifts low, so tools wear faster. Test: check and correct concentration at the start of every shift. Prediction: scrap falls to 2.5% or less. |
| Do | Day shift only, two weeks. Refractometer at the line, reading logged each shift. Night shift kept as the comparison. |
| Check | Day shift scrap 2.9%, night shift unchanged at 4.1%. Better, but short of 2.5%. The log showed tool changes were still made on breakage, not on a count. |
| Act | Adjust: make the concentration check standard on both shifts, and run cycle 2 testing tool changes on a part count. |
The night shift comparison is what makes the result believable: the drop happened only where the change was made. Carry a finished cycle into an A3 template when the story needs to go to management.
Why this tool matters
Most improvement fails because teams jump from idea to full rollout. PDCA adds discipline: predict what will happen, test small, measure, then decide. The value is in Check: did reality match the prediction? If not, you learned something about the process, and that is often worth more than the fix.
Facilitator tip
The most common PDCA mistake is skipping Check. Teams implement, declare victory and move on. Without before and after data, you do not know whether the change worked; you only think it did. Force the comparison: what was the metric before, and what is it now?
Frequently asked questions
What are the 4 steps of PDCA?
Plan: state the problem, the cause you suspect, the change you will test and what you predict. Do: run the change on a small scale. Check: compare the result with the prediction. Act: make it the standard if it worked, adjust and run another cycle if it partly worked, or drop it.
What is the difference between PDCA and PDSA?
Deming later preferred Plan-Do-Study-Act, because "check" can sound like a pass or fail inspection while "study" means learning from the gap between prediction and result. The steps are the same.
How long should a PDCA cycle take?
As short as the metric allows. A scrap or downtime test may need one or two weeks of data; a layout change at one station can be checked in a shift. Short cycles learn faster.
How is PDCA related to an A3?
An A3 is a PDCA written on one page. The left side and the countermeasures are Plan, the implementation plan is Do, and the follow-up is Check and Act.
Why write a prediction in the Plan step?
Without a prediction, any result looks like success. Writing "scrap will fall from 4.2% to 2.5%" before the test is what lets you learn when it falls to 2.9% instead.