Fable
The miller no longer counted how many times the waterwheel turned, but how many night shifts he had saved on hiring
The mill at the foot of the mountain used to rely on night-shift workers to sort grain sacks. Some checked labels, some weighed, and some picked out torn bags. The work was not hard, but it was fragmented and plentiful, and extra workers always had to be hired in peak season.
Later the mill installed a sorting wheel that could read. It could read labels, judge weight, and send most grain bags to the correct bin. The miller was delighted and on the first day had the accountant record how many turns the wheel made.
The wheels turned very frequently, and the ledger looked good. But when the month-end calculation was done, the night-shift workers had not decreased by much. Wet labels needed human review, mixed grain sacks needed human judgment, and every night someone still had to sort and carry the errors back.
The miller changed again to look at how many bags the machine processed. The number was still beautiful, but the foreman reminded him: some bags only passed through the machine and were ultimately finished by hand; some machines only saved the first step, and the next three steps remained the same. Passing through does not equal replacement.
In a peak-season review, the bookkeeper broke each type of task apart: reading labels, weighing, allocating to warehouses, identifying torn bags, handling disputes, logging exceptions, notifying customers. Then marked which were fully handled by the wheel, which were drafted by the wheel and confirmed by a person, and which still relied on manual work.
This table exposed the truth. The sorting wheel replaced most label reading, but replaced very little exception handling; overall night-shift work fell by only thirty percent. Those lively metrics from before had not answered how much of the original human work was actually gone.
The mill then focused improvements on the most labor-intensive steps. It added lights for wet labels, set rules for mixed grain sacks, and left an escalation channel for disputed items. In the second month, the night shift was finally cut in half. The miller also changed how he read the accounts: no longer asking how busy the wheel was, but asking which complete jobs people had done with less.
The accounting office also looked at the night-shift hours saved as money. Reading labels saves small money, while exception handling saves the big money; if they watched only the work most easily swallowed by the wheel, the mill would mistakenly think it had already saved a lot of labor. Real progress had to show up in shift schedules, wages, and delivery pressure. Later, every time the mill bought a new device, it first laid out the old shift schedule. If it could not let some type of person work one less shift or make a review step disappear, the accounting office would not count it as a true replacement. Later, during retrospectives, everyone wrote this lesson into daily rules: first see where the on-site result truly changed, then decide what to keep, repair, or limit.