Ask ten shop owners what their hourly rate is and you will get ten confident numbers and no two methods. Press on the number and it usually traces back to one of three places: what the shop across town charges, what a trade association suggested some years ago, or what the rate was last time somebody looked, plus a bit.
None of those is a calculation. They are all guesses wearing the costume of a decision, and they are the reason a shop can be busy every day of the year and still not be sure where the money went.
The rate itself is not complicated. It is one division.
Hourly rate = annual cost the work center must recover ÷ annual billable hours. Both halves are estimates. The top half is tedious but tractable. The bottom half is where shops quietly lose money, because the obvious answer is wrong — and wrong in the expensive direction.
Build the numerator: everything the work center has to pay for
The numerator is every cost that exists because that work center exists, for one year. Six categories cover almost all of it.
1. Direct labour, with burden
The wage is not the cost of the person. Add employer payroll taxes, unemployment insurance, workers' compensation — which is not cheap on production classifications — plus any benefits and the employer share of insurance. That total is your burden, and it is commonly a fifth to a third on top of wage.
Do not take that range as your number. Pull it from your own payroll reports. Burden varies enormously by state, by classification and by what you offer, and a borrowed percentage is how a rate ends up wrong by several dollars an hour before you have priced anything.
2. Equipment
Purchase price divided by useful life, in years. If the machine is financed, the interest belongs here too. If it is leased, use the lease payment instead and skip the depreciation — you are recovering cash out the door, not a tax schedule.
3. Maintenance
The service contract if you carry one. If you do not, use your own repair history rather than optimism; unplanned maintenance is a cost of running the machine whether or not it was in the budget.
4. Consumables that never make it onto a job ticket
Printheads, cleaning solution, wiper blades, purge waste, test prints, power. These get billed to nobody in particular, which is exactly why they need to live in the rate.
5. Space
The square footage the work center occupies at your annual cost per square foot — including the aisle you need to walk around it and the area where substrate stages before and after. A flatbed does not occupy only its own footprint.
6. Allocated overhead
Administrative salaries, the rent on the parts of the building that do not print anything, insurance, software, utilities, the truck. Split it across your work centers by a key you can defend — headcount, floor space, or direct labour hours are all reasonable. What matters far less than picking the perfect key is that every dollar of overhead lands somewhere.
One rate per work center, not one rate for the shop. A flatbed carrying six figures of capital and a laminator that cost a few thousand cannot share an hourly rate without one subsidising the other. When they share, you systematically overprice the cheap work you could win and underprice the expensive work you cannot afford to lose.
Build the denominator: available hours are not billable hours
This is the half that decides the answer, and it is the half that gets a two-second treatment.
The instinct is 2,080 — fifty-two weeks at forty hours. That number describes a calendar. It does not describe how many hours you can put on an invoice.
Take it down in two steps.
First, subtract time that is paid but not worked. Holidays and paid time off. Ten of each, at eight hours, is 160 hours gone before anyone has switched a machine on. You are left with attendance hours.
Second, apply utilisation — the share of attendance hours that actually lands on a job you can bill. What eats the difference is not idleness. It is setup and changeover, cleaning and calibration, scheduled maintenance, meetings, training, reprints, waiting on artwork, and waiting on a customer to approve a proof. All of it is real, all of it is paid, and none of it is billable.
Nobody runs at 100%. A shop with steady work and disciplined scheduling lands somewhere in the seventies. If you do not know your own figure, that is the single most valuable thing on this page to go and measure, because it moves the rate more than any other input.
You get it from your own records: hours logged against jobs divided by hours paid, over a full season rather than a good week. If your system cannot tell you that, treat the gap itself as a finding — it is the same blind spot that lets stale time and cost standards survive for a decade.
A worked example
Every figure below is an illustrative placeholder, chosen to show the shape of the calculation. Substitute your own; the method is the point, not the numbers.
One work center: a wide-format printer with a dedicated operator.
| Annual cost to recover | Amount | How it was derived |
|---|---|---|
| Operator wage | $52,000 | Base salary |
| Payroll burden | $11,440 | 22% of wage |
| Equipment depreciation | $24,000 | $120,000 over a 5-year life |
| Maintenance | $6,000 | Annual service contract |
| Power and consumables | $3,000 | Not billed per job |
| Floor space | $4,800 | 400 sq ft at $12 per sq ft |
| Allocated overhead | $18,000 | This center's share |
| Total | $119,240 |
Now the hours. Start at 2,080. Subtract 160 hours of holiday and paid time off, leaving 1,920 attendance hours. Apply 72% utilisation and you have 1,382.4 billable hours.
Divide:
| Method | Hours used | Resulting rate |
|---|---|---|
| Divided by billable hours | 1,382.4 | $86.26 |
| Divided by the calendar | 2,080 | $57.33 |
The gap is $28.93 an hour. Put another way: dividing by 2,080 recovers about 66% of what the work center costs to run. Charge the lower number and every billable hour is roughly a third short — which on this example's 1,382 billable hours is a shortfall near $40,000 a year, from one work center, entirely invisible on a busy shop floor.
That is the whole argument for taking the denominator seriously. It is not a rounding error. It is the difference between a shop that works hard and a shop that makes money.
What utilisation does to the number
Because utilisation carries so much weight, it is worth seeing the sensitivity directly. Same $119,240 of cost, same 1,920 attendance hours, utilisation varied:
| Utilisation | Billable hours | Hourly rate |
|---|---|---|
| 60% | 1,152.0 | $103.51 |
| 65% | 1,248.0 | $95.54 |
| 70% | 1,344.0 | $88.72 |
| 72% | 1,382.4 | $86.26 |
| 75% | 1,440.0 | $82.81 |
| 80% | 1,536.0 | $77.63 |
| 85% | 1,632.0 | $73.06 |
Fifteen points of utilisation is thirty dollars an hour. Two conclusions follow, and they pull in different directions.
The first is that guessing at utilisation is not acceptable, because the guess is worth more than most of the cost line items you carefully researched. The second is that improving utilisation is a pricing lever as well as an operations one — every point you claw back from setup, waiting and rework lowers the rate you need in order to break even, which is a competitive advantage you can choose to keep or spend.
This is a floor, not a price
What you have calculated is the rate at which the work center breaks even. It contains no profit. Selling at it means running the machine all year for the privilege of covering its own costs.
Profit goes on top, and the arithmetic of putting it there is its own trap — adding 30% to a cost does not produce a 30% margin. That distinction, and the way it compounds across a quote, is covered in our guide to pricing wide-format work.
Two more caveats worth holding onto. The rate is an average, so rush jobs, awkward substrates and press-check work genuinely cost more than an average hour and should not be sold at the average. And a rate is not a quote: a quote is material, plus labour and machine time at these rates, plus finishing, plus freight, plus margin. The rate is one input to the estimate, which is the argument for having the estimate assembled from your costs rather than reverse-engineered from a target price.
Recheck it on a schedule
Rates decay. Wages move, a lease ends, a machine gets paid off, insurance reprices, volume shifts and utilisation shifts with it. None of those events announce themselves in your pricing, so the rate stays right until one day it quietly is not.
Recalculate annually at minimum, and immediately after any of these: an equipment purchase or payoff, a lease change, a wage change, a move, or a real change in the volume of work going through the center. Ten minutes of division protects a year of quoting.
The shops that do this well are not the ones with the most sophisticated model. They are the ones whose number came from their own data, recently, and who can say out loud where each half of the division came from.