A kaizen event lasts one week, and by Friday it feels like a complete win. Before and after photos, applause, somebody thanking the team. The real result has not been decided yet.
A kaizen event is a week of focused work where a cross functional team attacks a specific problem or process, implements changes on the spot, and measures the result before closing. What sets it apart from an ordinary improvement is speed. Nothing gets documented for later. It gets tested on the floor that same week.
What actually happens in those five days
Monday defines the scope and captures the baseline with real data, not with the mood of whoever walked into the room most frustrated. Tuesday and Wednesday are for implementation: the layout moves, the new method gets tested, the first poka yoke devices go in if the case calls for them. Thursday measures again. Friday presents to management.
That sequence works, and it works fast, because the team is fully dedicated for those five days, free of the normal email and meetings piling up. That is exactly the trap. That dedication is the exception, and it does not exist next Monday.
Why Friday is misleading
Friday looks solved because the team carried it with full energy and attention all week. The issue is that level of attention cannot hold permanently, and the change just installed has not yet gone through a weekend shift, an operator turnover or a production spike.
I have seen excellent kaizen events collapse in the third week, not from bad design, but because nobody defined what happened after Friday.
What decides the result
What decides whether the change sticks is the eight weeks that follow. That is where I review with the team, on set dates, whether the change is still working, whether somebody quietly modified it, and whether the indicator that moved on Friday holds up under a normal shift.
I already wrote about 5S and how easily it collapses when nobody audits it. A kaizen event carries the same risk, multiplied, because the change went in under the pressure of one week and nobody has tested it under normal conditions yet.
We covered what happens during the week and why Friday is not the finish line. Next comes the part almost nobody schedules: follow up.
The follow up that actually holds the change
Follow up is a date on the calendar, with the same team, reviewing the same indicator against the same baseline. If the number held, it gets documented and closed. If it slipped, the team goes back to the cause before anybody tries to fix it with another full week.
That portfolio of kaizen events, prioritized inside a larger journey, is what turns one good Friday into a chain of improvements that actually compound.
- Who reviews the indicator in week four, with a name and a date on the calendar?
- Has the change already run through a shift different from the one that ran the event?
- What happens if the number drops? Is that already written down, or will it be improvised?
Start here
If you already ran a kaizen event this year, find the closing photo from that Friday and ask whether anyone has checked that indicator in the last four weeks. If the answer is no, that is your next event: fixing the follow up on the last one.
I always run kaizen events with follow up built in, never as a standalone week. If you have a specific process in mind, tell me which one and when you would like to run it.
What is a kaizen event?
It is a week of focused work where a team implements and tests changes on a specific process, with before and after data, instead of just proposing them on paper.
How long does a kaizen event last?
The event itself takes one week. The follow up that decides whether the change holds runs over the eight weeks after, with reviews on set dates.
Are kaizen and a kaizen event the same thing?
Kaizen is the philosophy of continuous improvement every day. A kaizen event is one focused, accelerated application of that philosophy, compressed into a week on a specific problem.