Five questions to ask before your next improvement project

If the person doing the work and the person paying for it describe the problem differently, you don't have a problem statement — you have two. Keep refining until one sentence satisfies both.

Delay, rework, overtime, lost orders. If the answer is "we don't know", that measurement is the first project. Use comparable units: a mistake rate per million chances for quality, and for delay, the simple rule that time-in-system equals work-in-progress divided by completion rate. Measured numbers replace anecdotes — and the priority list usually surprises people.

A project ends; the process continues. If nobody owns the process afterwards, the improvement decays within months and you've run a demonstration, not a change.

Every added checkpoint and report spends capacity — and queues punish high busy-ness brutally: waiting time grows like $\frac{\rho }{1-\rho }$, which in plain words means a team at 95% busy waits about five times longer than one at 80%. An improvement that only adds steps is usually waste wearing a badge. The best interventions remove.

Decide now which number you'll check in six months, and against which baseline. A real gain shifts the average by more than the normal month-to-month bounce — and stays shifted. This is why the Chido method measures its final chapter against the baseline taken in its second.

Five honest answers before you start beat any amount of heroics after. To pressure-test a project you're about to launch, book a discovery call — thirty minutes, no jargon.