What should a machine shop charge per hour?

Ask what a machine shop should charge per hour and you will be told anywhere from $60 to $200. Both are true somewhere, and neither helps you, because a shop rate is not a market price you look up. It is an arithmetic result of five things you already pay for.

The five are: what the machine costs you as it wears out, what the person running it costs with burden on top, the electricity it draws, the share of your fixed overhead that machine has to carry, and the consumables it eats. Add those and you have your break-even hour. What you charge is that plus whatever markup you decide — but until you know the break-even, the markup conversation is theatre.

The one that gets it wrong: which hours you divide by

Depreciation and overhead are fixed per year. To get an hourly figure you divide them by hours. The question is which hours, and this is where most shop rate math goes wrong.

If your doors are open forty hours a week, it is tempting to divide by two thousand hours a year. But your machine does not cut for forty hours. Between setups, programming, inspection, waiting on material, and the hour that vanishes every time a customer calls, a high-mix job shop is often cutting twenty to twenty-five hours a week on a given machine.

Divide by open-door hours instead of spindle hours and you understate your rate by 30 to 50%. Every job you quote is then priced on a machine that does not exist.

The consequence is worse than a flat error, because it moves. The same shop, with the same rent and the same wages, has a materially higher cost per hour in a slow month than a busy one. If you calculated your rate when you were flat out and you are now at half capacity, everything you quote is underpriced — and nothing in your accounts tells you until the year closes.

Working through a real one

Take a $100,000 machine with ten years left, cutting twenty-five hours a week, fifty weeks a year. That is 1,250 spindle hours a year and 12,500 over its remaining life, so depreciation is $8.00 an hour. The operator is on $30 with 40% burden, so $42.00. Ten kilowatts at ten cents is $1.00. Your fixed overhead is $6,000 a month shared across three machines, so $24,000 a year on this one, which over 1,250 hours is $19.20. Consumables run $5.00.

Now run the same shop at forty spindle hours a week. Depreciation falls to $5.00 and overhead to $12.00, and the break-even drops to $65.00. Run it at fifteen hours and depreciation is $13.33, overhead $32.00, and the break-even climbs to $93.33.

Same machine, same wage, same rent: $65.00 an hour busy, $93.33 an hour slow. A 44% swing, driven by nothing but how much work is in the door.

What to do with the number

• ·Recalculate it quarterly, not once. It moves with utilization and with your electricity bill.
• ·Run it per machine, not for the shop. A manual lathe and a machining center do not carry the same depreciation or the same overhead share.
• ·Keep the break-even and the price separate in your head. Discounting is a decision about margin, and you cannot make it if the two numbers are fused.
• ·If a customer's work only clears break-even, that is not a customer, that is a hobby with paperwork.

What this number still will not tell you

A rate is a cost per hour. It says nothing about how many hours a particular job takes, and that is where most quoting errors actually live. A shop with a perfectly calculated $95 rate will still lose money if it estimates six hours on a job that takes nine.

The only fix for that is to compare what you estimated against what the job actually took, job after job, and let the real number replace the guess. Most shops have that data on a timecard somewhere and never close the loop.