What a burdened rate actually includes
Ask where the hourly rate on the grinding cell came from and the honest answer on most floors is that it came with the floor. Someone built it years ago, someone else added 5 percent to it once, and it has been quoting parts ever since.
Every cycle time on every estimate gets multiplied by that number, which makes it the highest leverage figure in the building, and also the least examined one. The figures in this piece are illustrative throughout, because rates are specific to a floor by definition, and that is precisely the point of rebuilding yours.
The stakes rise as the mix moves. A rate built when the floor ran long repeat jobs meets a present full of short runs, tighter tolerances and more changeovers, and every one of those shifts pushes the true cost of an hour up while the quoted figure stands still.
A burdened rate is the real cost of an hour on a specific work center, everything included.
The reason it exists is that a machine’s hour costs far more than the operator standing at it, and a quote priced on wages alone is subsidised by the rest of the business without anyone deciding to subsidise it.
What the rate is supposed to carry
Build it from the machine outward. Ownership first, the purchase amortised over its working life, or the lease payment if you lease, plus the interest either way.
Then the space it stands on, its share of the rent or the building cost, the power it draws under load rather than at idle, and the compressed air, coolant and chips handling that follow it around.
Maintenance comes next, both the planned kind and a realistic allowance for the unplanned kind, because a rate built on a machine that never breaks is fiction. Then perishable tooling and consumables as a running share, distinct from job-specific tooling, which belongs on the quote itself.
The rate is the multiplier.
Then the operator, fully loaded with taxes, insurance and benefits, at the fraction of their attention this machine actually gets. A person tending 3 machines puts a third of a loaded hour on each.
Above the cell sits the share of people who never touch the part but exist because it does, supervision, scheduling, quality, purchasing. And across the whole build sits utilization, the honest fraction of the year the machine spends cutting.
Divide by hoped-for hours and the rate flatters you. Divide by real hours, setup, changeover and waiting included, and the rate tells the truth.
One floor, one rate, wrong twice
The other inherited habit is the blended rate, a single hourly figure applied across the whole floor because it is easy to remember and easy to defend in a meeting. A blended rate is wrong in both directions at once.
It overprices the hour on the old saw and underprices the hour on the 5-axis cell, because the 2 machines do not cost the same to own, power or staff, and averaging them prices neither.
The commercial damage comes from how the market responds to that split error. Buyers compare quotes line against line, so the work you overpriced goes somewhere else and the work you underpriced arrives in volume.
The floor running a blended rate slowly fills its most expensive machines with its worst-priced work, and the mix drifts toward the jobs that lose money without a single quote looking wrong on the day it went out.
A blended rate is wrong twice.
A rate per work center is more arithmetic once a year, and it is the difference between choosing your mix and having it chosen for you.
What quietly gets left out
The inherited rate is almost never wrong about the machine. It is wrong about the edges. The utilization assumption is the usual offender, a denominator set in an optimistic January and never revisited, so every quoted hour silently assumes a fuller year than the floor ever runs.
Maintenance is the second, carried as a token figure until the spindle rebuild lands in a single quarter.
Setup scrap is the third, the first pieces of every changeover that were always going to be scrap and appear in no rate at all. And inspection time on tight work gets paid by somebody, whether or not the rate admits it.
Changeover is its own category of leak. The hour spent tearing down and setting up is an hour the machine earns nothing, and it has to live somewhere, either inside the utilization denominator or as setup time on the quote itself.
The trouble starts when it lives in neither, counted as productive capacity in the rate and priced on no line of the estimate, which is the default state of an inherited rate table on a floor whose batch sizes have been shrinking for a decade.
Each omission is small. Their sum routinely runs to a meaningful slice of the hour, and every omission moves the quote in the same direction, downward. An overstated rate loses you jobs visibly, on price, and the market corrects you fast.
Understating wins bad work.
An understated rate wins you jobs and loses you money on each, which no report surfaces, because the jobs are busy and the invoices are paid and the leak only exists in the difference between 2 numbers nobody compares.
The arithmetic of understating
Run the shape of it on the worked part this site prices everywhere else. The honed bore on that spool valve carries $0.135 a piece at 5.1 seconds of cycle. Understate the hone cell’s rate by a tenth and the line drops by about a cent.
A cent looks like nothing on a quote and survives every review. At 1,200,000 pieces a year on a 6 year program, that cent is roughly $84,000 across the program, from one line, on one part, and a floor quoting from the same rate table repeats the error on every part that crosses that cell.
The same family of numbers appears at larger scale in the history that example draws on, where a related valve quoted at $1.28 ran at $1.37, and 9 cents across 3.4 million pieces came to $306,000. Nobody chose that loss.
It was the accumulated interest on unexamined inputs, and the rate is the most common unexamined input there is. Program work makes the compounding worse, because the price is fixed for years while the understated costs are not, and the gap you signed in year 1 is still being paid in year 6.
Rebuilding it without a project
The rebuild is smaller than it sounds, because the inputs already exist. Ownership and lease figures sit in the ledger, power is on a meter, loaded labor is in payroll, and the honest utilization number is in the schedule if anyone divides by it.
One afternoon per work center, once a year, with the controller and the person who runs the cell in the same room, produces a rate you can defend line by line.
Date the rate.
Start with the cells that matter, which is a shorter list than the machine inventory. The work centers that carry tight tolerance work at high utilization are where a wrong rate does program-scale damage, because their hours are expensive, fully booked and multiplied across the largest volumes.
The old manual mill that runs 200 hours a year can keep its inherited rate for another season without costing anyone much. The hone cell cannot.
The temptation to resist is precision theater. A rate carried to 4 decimal places from a model full of guesses is less useful than a round number whose every component has a source.
The test of the rebuild is that when a buyer, an auditor or your own general manager asks where the hourly figure came from, the answer is a one page walk from the ledger to the cell, dated, with nothing in it labelled miscellaneous.
Two habits keep it honest afterward. Date the rate, the way you date a bar quote, so an estimate built on a 3 year old hour is visibly built on one. And close the loop against reality, because the only proof of a rate is what jobs actually cost.
Every closed job is a test of the table that priced it, and a floor that compares the 2, the way the build in how to quote a machined part ends by comparing quoted against ran, finds its rate drift in months instead of discovering it at the end of a 6 year program.
Any quoting tool you evaluate should make that comparison effortless, which is one of the tests applied in the buyer’s guide.