Kaizen — small improvements, compounded, and the two ways organisations get it wrong

Here is the arithmetic that makes kaizen more than a slogan. Suppose a team improves something by just 1% each week — trim a step, remove a handoff, fix a form. Improvements compound, so after a year:

In plain words: fifty-two tiny 1% gains don't add up to 52%. They multiply up to about 68% better — because each week's improvement is applied to an already-improved process. Two per cent a week compounds to roughly 180% in a year. That is the quiet engine of kaizen: modest, relentless, multiplying.

Compare the big-bang alternative: one heroic 30% improvement project, launched once, that then decays. The general shape of that decay is

which in plain words says the gain shrinks steadily over time unless something holds it in place. Big projects without daily habit lose ground from the day the consultants leave. Kaizen is the habit that stops the decay — and keeps adding.

Every kaizen improvement, however small, runs a tiny experiment: plan a change, do it on a small scale, check whether the number moved, act — keep it, adjust it, or drop it. This is the PDCA cycle, and the "check" step is the one that separates kaizen from tinkering. A change nobody measured is a guess wearing overalls.

The measurement doesn't need to be grand. A run chart on the wall — the number, plotted over time, with its median drawn across — is enough to see whether this week's change actually moved anything, or whether it's just the usual bounce.

The most common corruption of kaizen is the suggestion scheme. Staff post ideas; a committee reviews them monthly; a few are approved; most vanish. This is kaizen with its engine removed. The whole point is that the person with the idea tries it, quickly and safely, and checks the result — not that they lobby for permission.

The test: how long from idea to trial? In healthy kaizen it's hours or days. If it's weeks, you don't have kaizen; you have a slow-motion complaints process.

The other corruption is the kaizen blitz — a three-day workshop that produces an impressive wall of sticky notes and a before-and-after photo. Blitzes have their place: for a stubborn bottleneck, a concentrated push works. But an organisation that only does kaizen in events has confused the ceremony with the habit. The 68%-a-year arithmetic above only works if the improving never stops.

Worse, events rarely include the boring final step: locking the gain in place with a new standard and a chart that watches it. Without that, the decay formula takes over on Monday morning.

• Frequent. Improvements measured in days, not quarters.
• Local. Made by the people doing the work, on the work they do.
• Checked. Every change has a number, and someone looks at it.
• Held. A good change becomes the new standard way, written down, so the next improvement stands on it rather than replacing it.

The people on the floor already know where the waste is. Kaizen is simply the discipline of letting them remove it, one small piece at a time — and the maths of compounding is why that patience wins.