Overtime that never goes away is not about demand
Everybody expects overtime when a big order lands. The overtime worth worrying about is the kind that shows up on the same line, week after week, no matter what the order book looks like. That is not demand. That is a line that cannot make its number in the hours it has, and the overtime is quietly paying to cover the gap. It is one of the most expensive losses on a floor because it looks like hard work instead of a problem.
The hours are hiding a daytime loss
Overtime is almost never created at the end of the shift. It is created during the shift, every time the line stops, waits, or makes something twice, and then the hours at the end are spent catching up. So the overtime is a symptom. The cause is the changeover that ran long, the material that showed up late, the small stops nobody logged, the rework that jumped the queue. Add those up across a shift and you have the exact number of overtime hours you are paying for.
Chase one line for one week
Pick the line with the most recurring overtime and, for one week, write down why it ran long each day in one sentence. Not a formal study, just an honest note at the end of each shift. By Friday the pattern is almost always obvious: the same changeover, the same missing part, the same machine. Overtime feels random when you only see the hours. It stops feeling random the moment you write down the reasons next to them.
Why it is worth fixing first
Overtime is money you are already spending, every week, with nothing to show for it but the same output you should have made in regular hours. Recover it and the savings are immediate and permanent, not a one-time project. And the fixes are usually the same ones that speed up the whole line: faster changeovers, material arriving on time, fewer small stops. You fix the daytime loss and the overtime disappears on its own.
The one you should not cut
One caution: not all overtime is waste. Overtime that genuinely covers real, profitable demand is fine, and cutting it would cost you orders. The point is to tell the two apart, which the one-week log does cleanly. Cut the overtime that covers a loss; keep the overtime that covers a sale.
Telling recurring, loss-driven overtime from real demand on an unfamiliar floor is exactly the kind of thing a walkthrough surfaces fast.
Request your free walkthrough and we will find where your floor is losing the most time, no charge.
