The Spindle Hours You Paid For
A shop buys a machining center for a quarter of $1 million. The justification, written down or just carried around in the owner's head, was that the machine would run. Not that it would sit on the floor looking capable. That it would be in the cut enough hours a year to pay for itself and leave something over. Almost no one goes back a year later and checks whether it did.
Real spindle utilization is the share of the hours you paid for that the tool actually spent cutting metal. It's an unforgiving number because it throws out everything that feels like production but isn't making chips. Most owners have never seen it, partly because the machine's own hour meter misleads them. The meter counts power-on time. It counts the machine idling while an operator hunts for a fixture, warming up, holding at the end of a cycle for someone to open the door, waiting on a program that isn't proven out. Power-on looks like work. It isn't the same thing as cutting.
What the good data says
Godlan's 2025 benchmarking, built from a full year of discrete-manufacturing data, put average overall equipment effectiveness at 66.8 percent across nine sectors. World-class usually gets pegged at 85. The figure that should stop you, though, is not the average. It's where the losses come from. Setup and changeover by itself accounted for 28.7 percent of all loss, and close to 63 percent of total loss traced back to availability rather than to running slow or scrapping parts. The machine isn't underperforming while it works. It's not working.
Smaller shops usually sit below the average, not above it. A lot of small and mid-size shops live in the 30 to 60 percent range on OEE, and the reason that number matters runs opposite to how it feels. Pulling a shop from 45 to 60 percent is the rough equivalent of adding a quarter to a half again as much capacity, and you do it without buying anything. The capacity is already there. You already wrote the check for it. It's locked inside the machine you own, behind the setups and the waiting.
Setup is where the capacity goes to die
The part nobody costs honestly is changeover. A machine that runs long production jobs and a machine that runs short lot sizes with a fresh setup every few hours are two completely different economics on identical iron. The second one can burn a third of its available time getting ready to cut and never show a red light, because getting ready isn't downtime in anyone's system. It's just how the day goes.
Programming and proving out live in the same blind spot. An operator standing at the control editing offsets is a skilled person doing necessary work, and the spindle is stopped the entire time. None of that is waste in the sense of somebody loafing. It's purchased capacity, spent on everything except the one activity that pays the note. MachineMetrics has reported that shops adding automated loading on a cell can move spindle utilization from roughly 50 percent to 85. Some of that is the robot. Most of it is that the robot doesn't stop to answer a question, walk to the tool crib, or take a break the machine was quietly taking with it.
Count it before you buy the next one
You do not need a monitoring platform to start. Take your busiest machine and, for 1 week, have somebody log when the spindle is actually cutting versus everything else. Setups, waiting on material, waiting on a program, waiting on an inspection, idle. You will not enjoy the number. Owners rarely do the first time. It usually lands somewhere far under what the machine felt like it was doing.
Then look at what the low number is telling you before you sign for machine number two. If the spindle you already own runs 45 percent of its paid hours, the shortest path to more output is not another quarter-million-dollar center that will also run 45 percent. It's the setups, the fixturing, the tool presetting, the programming done offline instead of at the control. Those cost a fraction of a machine and they attack the exact hours you're already paying for and throwing away.
Measure the spindle before you buy iron to sit next to it. The most expensive capacity in the building is the capacity you already own and can't see.