
The quote that looked fine until the invoice came due
A shop quotes a bracket job at $42/hour shop rate, same rate they use for everything — the CNC mill, the manual deburring bench, the assembly table. The mill time comes in as estimated. The deburring and assembly run long, because they always do, but nobody notices because the rate never changes and the labor hours get lumped into one job total. Three months later, the owner is looking at a quarter where machining jobs made money and assembly-heavy jobs quietly lost it, and the spreadsheet can't tell why — because it never separated the cost of running a $250,000 five-axis mill from the cost of a folding table and a work light.
This isn't a case for a controller or a cost-accounting textbook. It's a case for overhead allocation done at the level of precision a 5–50 person job shop actually needs: not perfectly, not by activity-based costing standards, but by work center — the smallest unit of resolution that changes the answer. This article walks through that method: what overhead allocation actually means, how to calculate a rate you can defend, and where it belongs in a routing before the job ever hits the floor.
What overhead allocation actually is
Overhead allocation is the process of spreading indirect costs — the ones that don't show up as a line item on a specific job — across the jobs that use them. Rent, utilities, equipment depreciation, supervision, quality staff, software, insurance: none of these get invoiced per part, but every part consumes a share of them. Allocation is the mechanism that assigns that share.
The alternative — not allocating overhead at all, or allocating it as one flat number regardless of what a job actually uses — is what causes the bracket-job problem above. A shop that treats a manual bench and a CNC mill as costing the same per hour to run is guaranteed to misprice something, because the two are never actually equal. The mill carries expensive depreciation and power draw; the bench carries almost none. Flattening them into one number doesn't make the difference disappear — it just hides it inside whichever jobs happen to use more of the expensive resource.
The predetermined overhead rate: the one calculation you actually need
The core tool is the predetermined overhead rate — an estimated rate, set before the accounting period starts, used to apply overhead to jobs as they're quoted and run rather than waiting until year-end actuals are in. The formula is simple:
Predetermined overhead rate = Estimated total overhead cost ÷ Estimated total allocation base (hours, or machine hours, or dollars of direct labor)
That's the entire concept. The complexity small shops run into isn't the math — it's deciding what counts as "overhead" and what the allocation base should be. Overhead should include everything indirect: facility costs, indirect supervision, quality and inspection staff not billed directly to a job, software, insurance, depreciation on shared equipment. It should exclude direct labor and direct material, which get costed separately per job. The allocation base is usually direct labor hours or machine hours — whichever more closely tracks how the overhead actually gets consumed. A shop full of manual assembly work should probably allocate on labor hours; a shop dominated by unattended CNC run time should allocate more on machine hours.
For the full walkthrough on setting this rate, including how to pick and defend an allocation base, see our predetermined overhead rate calculation guide.
Rolling overhead into a work center rate: a worked example
Here's where the flat-rate problem gets solved — not by allocating overhead once for the whole shop, but by calculating it separately for each work center. This is illustrative math for a representative shop, not a sourced industry figure; the point is the method, and every shop's own numbers will differ.
Say a shop has three work centers: a CNC mill, a manual weld station, and an assembly bench. Total shop overhead for the year is estimated, then split across the three centers based on square footage, power draw, and equipment value — the mill, being the most capital-intensive, absorbs the largest share. Divide each center's share of overhead by its estimated annual hours, and you get three different overhead rates per hour, not one. Add labor cost and burden per hour on top of each, and you get three different work center rates — the number that actually belongs on a quote and in a routing operation.
This is the mechanism behind a work center rate calculation: the shop stops asking "what does an hour of labor cost us" and starts asking "what does an hour at this specific work center cost us," because those two questions have different answers, and the gap between them is exactly what erodes margin on jobs that don't match the shop's "average" mix.
Where labor burden fits in
Overhead allocation and labor burden are related but not identical, and shops often conflate them. Labor burden is the fully loaded cost of an employee-hour beyond the base wage — payroll taxes, workers' comp, benefits, paid time off. It answers "what does this person actually cost us per hour worked," separate from facility and equipment overhead. A complete work center rate typically stacks both: the labor cost including burden, plus the work center's allocated overhead rate, rolled into one number.
Skipping burden and quoting off base wage alone is a quieter version of the same mistake as flat-rating overhead — it understates true cost in a way that doesn't show up until the job is done and the numbers get reconciled. For the full breakdown of what belongs in a burden calculation, see how to calculate burden rate and labor burden rate manufacturing.
When one shop rate isn't enough anymore
Not every shop needs work-center-level allocation on day one. A shop running one process on largely interchangeable equipment can often get by with a single blended rate for a while. The signal that it's time to split rates is usually a quoting pattern: jobs concentrated in one process type consistently come in more profitable than jobs spread across multiple processes, or vice versa, and nobody can explain why from the job cost report — because the report doesn't distinguish between the processes in the first place.
At that point, the fix isn't more accounting theory. It's breaking the shop rate into per-work-center rates, applying them to routings operation by operation, and comparing actual hours logged at each operation against what was quoted. That comparison — actual vs. quoted, by operation — is the real payoff of doing allocation this way: it turns "we lost money on that job" into "we lost money on the deburr operation specifically, because it ran two hours over a rate that was never wrong to begin with."
Putting the rate to work: routing, traveler, and the numbers that follow
Once work center rates exist, they need a place to live that isn't a spreadsheet nobody updates. In WorkTickets, a routing is built per part number with each operation assigned to a work center — mill, weld, assembly, whatever the shop runs — and each work center carries its own configured burden and machine rate. That routing becomes the traveler the job carries through the floor, and clock-in/clock-out at each operation logs actual time against the standard the routing quoted. The job-level profitability summary then rolls actual labor and the work center's rate together, so the shop can see, operation by operation, where a job's cost matched the quote and where it didn't — without anyone reconstructing the math after the fact.
That's the whole point of doing overhead allocation carefully in the first place: not a more elegant number on paper, but a rate accurate enough that when a job runs long, the shop knows within which operation, and by how much.
If you want to build this out for your own shop before touching any software, the Multi-Work-Center Rate & Utilization Model is a downloadable template that walks through splitting overhead across work centers, calculating burden, and rolling both into defensible quoting rates — the same structure described above, ready to fill in with your own numbers. For the broader picture of how allocation, burden, and job costing fit together, the job costing resource hub is the starting point.

