The capacity you are looking for is already on the floor
Every busy shop hits the same wall. The work is stacking up, the dates are slipping, and the obvious answer is more: another machine, another shift, another operator. All of those cost money you may not have, and all of them take time you definitely do not have.
Before you spend any of it, look at your setups. There is a good chance the capacity you are about to buy is already sitting on your floor, lost to changeover time you have never actually measured. A week of honest data often turns up 10 to 30 percent more capacity inside the schedule you already run. That is not a rounding error. On three machines that can be most of a fourth, and you already own it.
The reason it stays hidden is simple. Cutting is loud and visible. Setup is quiet, and most shops never write it down as its own thing. So it disappears into a vague bucket called downtime, and you cannot fix a number you cannot see.
Why changeover is the time that hides
Ask most owners where their time goes and they will talk about run time, because run time is what they watch. The part is cutting or it is not. Setup is different. It is a person walking for tooling, tearing down the last job, finding the program, dialing in the first article. It does not look like production, so it does not get counted like production.
It gets lumped into one blurry bucket
When a machine is not cutting, the time tends to get logged as downtime, if it gets logged at all. Setup, waiting for tooling, changeover, and a hundred small interruptions all land in the same pile. Until you split that pile apart, you genuinely cannot say whether your machines are busy making parts or busy getting ready to make parts. Those are very different problems, and only one of them is fixed by buying iron.
The bottleneck is not the machine you think
It is tempting to assume the bottleneck is the slowest or loudest machine. In a job shop it usually is not. The real constraint is the point where work stops moving predictably, where jobs pile up waiting for a setup, a tool, or a missing part to show up. A machine that spends half its day being changed over is a bottleneck even if it is the fastest thing you own, because it is not the cutting that is slow, it is everything around the cutting. This is the same root cause behind a lot of late jobs: the time leaks where nobody is looking.
How to get the capacity back
You do not need a lean consultant or a stopwatch on every station. You need to see the setup time honestly, then sequence the work so you do it less often. Four moves do most of the job.
Track setup and changeover on their own for one week
This is the unglamorous first step and it is the one that changes minds. For one week, log setup and changeover separately from run time. Not perfectly, just honestly. At the end of the week add it up. The number is almost always bigger than anyone guessed, and that surprise is what makes the rest of this worth doing. You cannot manage time you refuse to measure, and right now most shops are managing it blind. The same instinct that drives machine utilization applies here: if it is not measured, it is not real.
Sequence by shared setup, not by whoever shouted loudest
Most schedules are built by urgency. Whatever job is screaming today goes on the machine next, even if the machine has to be torn down and rebuilt to run it. That is how a shop changes over three times in a day for jobs that could have shared one setup. Family sequencing fixes it. Group the jobs that use a similar setup and run them back to back. Brake families in the morning, small batch changeovers after lunch. You still hit your dates, you just stop paying the setup tax three times when once would do. The hours you save are real and they cost nothing.
Stage the next setup while the current job runs
A lot of changeover time is spent doing things you could have done while the machine was still cutting. Pulling the next program, gathering the tooling, staging the material, reading the print. If the operator only starts hunting for the next job after the spindle stops, the machine sits idle for work that did not need the machine at all. Stage the next setup during a known window while the current job runs, and the changeover itself shrinks to the part that genuinely needs the machine stopped. This is the oldest trick in setup reduction and it is still the highest return.
Limit how many jobs you open at once
Plenty of shops create their own bottleneck by starting everything. Ten jobs get opened, all ten are half done, and the floor is a maze of partially finished work waiting on a tool or a setup that is tied up elsewhere. A simple rule helps: finish the critical work on open jobs before launching new ones. Less work in progress means less thrash, fewer setups interrupted halfway, and parts that actually move through the shop instead of sitting. It is also far easier to see what is really going on when the floor is not buried, which is half the battle when you are tracking jobs across machines.
Measure first, then decide if you really need to buy
Buying a machine to fix a setup problem is an expensive way to make the same mistake with more steel. The new machine changes over just as often as the old one, and the capacity you were missing is still hiding in the same place, now spread across one more spindle.
Start with the cheap move. Track setup and changeover honestly for a week, sequence the work so you tear down less often, stage what you can while the machine runs, and stop opening more jobs than the floor can move. Most shops find a real chunk of capacity they already paid for. If you still need another machine after that, at least now you are buying it for the right reason, with numbers behind the decision instead of a gut feeling.
All of it depends on one thing: being able to see, in one place, where your machine time actually goes. The cutting is the easy part to see. The setups, the changeovers, the half finished jobs waiting on a tool are the part that decides whether you ship on time, and they are exactly the part a whiteboard hides. For machine shops running high mix, low volume work, where you change setups all day long, that visibility is not a nice to have. It is the capacity.