I have a junior colleague I respect. He's a hard worker, and his character is among the best I've known in my life. In character and in work, he's far above me.
What's amazing is that even in unknown territory, from zero knowledge and experience, he keeps running PDCA until he gets to success.
Looking back you might analyze it as simple, but almost nobody around me can push to that level. At least I couldn't have done it then in the same position.
Watching him, you see the simple truth: if you keep turning PDCA honestly, results tend to follow.
(Of course some things stay impossible no matter how much you improve, so judgment is hard — but if a market exists, in theory continuous improvement should at least get you to a minimum level of working.)
Conversely, cases that succeed after only a few PDCA loops are almost nonexistent. The closer feeling is: after about a year of grinding measures, you look back and results have stacked up.
I often support new businesses and growth, and short-term wins exist at the measure level — but growth you can feel as "the business moved to the next phase" usually takes at least a year. Looking back you can see the bottlenecks and which improvements helped; beforehand you can't predict, and while you're in it you're just trial-and-error in a fog. No matter how polished the business plan or strategy, it often doesn't work.
I remember supporting a new business where we had to report in weekly meetings. Progress was so bad that putting together retrospectives, number reports, and next-measure proposals was painful; the day before each meeting I couldn't even feel like going out for drinks — I'd just sit at home exhausted facing the PC. In the end it went beyond the plan, but it took about two years, and even after it was on track there was no "strategy" — we'd just done a lot of things.
I can explain why it worked and think there's some reproducibility — but there wasn't a consistent story based on strategy.
Coming back: PDCA is a well-made, universal framework along first principles.
Whether you literally manage phases and keep records or not, any measure inevitably follows that order.
And as above, the value of the PDCA cycle comes from turning the cycle.
The goal isn't maximizing one loop's effect but increasing the cumulative sum — so speed and volume of validation both matter.
Paradoxically, spinning faster and more to grow the sum also grows each loop's effect. Cliché, but quantity turns into quality. Strategy fans claim P is what matters most in PDCA. In a sense that's right — but over the medium to long term, it isn't.
Still, within PDCA, C and A are essentially P, I think. The KPIs you design, decision criteria, measurement foundation, and related actions should already be baked into P. C and A are decided against those criteria — if you only start organizing, analyzing, and choosing next actions when you hit C, you're late.
It's also better to design not only the current cycle but how the next cycle, and the one after, will look based on this cycle's results.
(I've written "PDCA" so many times I'm tired — readers even more so.)
I wondered what each PDCA phase needs. Off the cuff, something like:
P: knowledge, experience, logical thinking, imagination
D: resources (people, things, money, time)
C: logical thinking, knowledge
A: resources
I personally think C/A are essentially P (especially C), but from the dictionary meanings of PDCA this split feels about right.
P is highly creative. Experience and knowledge help a lot (e.g. whether experienced people are on a new business matters), and without logic the plan breaks, so logical thinking is needed. You also need imagination for uncertain situations (people often think knowledge and experience cover this; I don't think they contribute that much — it feels like a separate ability).
D: with resources, you can execute. It's fine to treat it as mechanical execution. If you're hesitating, the prior P was wrong.
C: confirm results as designed in P. Same for A.
By the way, looking it up, PDCA seems to be a framework used mainly in Japan.
Whatever other countries use, the plan–do–review–improve flow PDCA points to is common to almost any measure.
The thread ran out of steam, so that's all.
Tweet© 2025 ZUUHE