Choosing the first change

The first change is not chosen for its size. It is chosen for what it earns you, which is permission to make the second one. Get it right and everything after is easier; get it wrong and you will be relitigating the first one for a year.

Three criteria. Reversible: if it fails you can put it back on Monday and nothing is lost. Visible: the improvement is obvious to the people doing the work, not just in your spreadsheet — a shorter queue, less time hunting for something, fewer angry phone calls. And small: one process, one week, no new software, no new hire, no capital.

Then pick from the evidence you have gathered rather than from the thing that annoys you. Cross the leak sizes against the process map and look for a change that moves a number you have already baselined. And do not start with the thing your father or mother is proudest of, however wrong it is. That fight is winnable later, from a position where you have demonstrated something. It is not winnable in week three.