If you have searched for what shops charge per hour, you have found the tables: fifty states, a number beside each one, published by half a dozen sites. They are the first thing everyone writes about labor rates, ourselves now included.
We are not going to print another one. Not because the format is boring, but because we went looking for the numbers behind those tables and could not find a version we would be willing to put our name on. Here is what is actually knowable, what is not, and the arithmetic that decides your rate no matter what any table says.
The one number that is honestly sourced
The 2025 PartsTech survey of general repair shops found that nearly half price their labor between $120 and $159 an hour.
That is a range from a real survey that states what it asked and who it asked. It is the most useful public figure on this topic, and notice what it is not: it is not a state table, and it is not a single average. It is the middle of a wide distribution, and the shops outside that band are not doing it wrong.
Use it the way you would use a blood pressure reading. If your rate is $95 and you are in a market with real costs, that is worth a hard look. If it is $175 and your bays are full, nothing is broken.
Why we will not print a state table
Three families of numbers are circulating, and they do not agree.
The vendor datasets. One shop-software company publishes a fifty-state table putting the national average at $132 an hour, with states running from $85 to $197. Their page says the figures come from more than 10,000 shops across North America — meaning their own customers. That is a genuinely large sample and it may well be the best data anyone has. It is also unverifiable from outside, because nobody can audit a private customer base. We are a competitor of theirs, so take this as read: we have no basis to call their numbers wrong, and no way to confirm them either.
The syndicated dataset. Several consumer-facing pages, including a motoring club’s 2026 article, present a state-by-state table sourced to an aggregator, which in turn credits a vehicle-health index. Follow that chain to the end and the underlying page reports average total repair cost per visit — a per-job total, with its labor column dated 2021. The motoring club’s article presents figures in the $211 to $281 band; the aggregator page it names currently shows $127 to $152. A per-visit cost is not an hourly rate, five-year-old data is not a 2026 benchmark, and a citing page should not disagree with the page it cites.
The editorial estimates. At least one site says so outright, and deserves credit for it: its table is compiled from posted shop rates and market listings and is, in its own words, an editorial estimate rather than a census.
Put the ranges side by side — $85 to $197 against $211 to $281 — and the gap is roughly two to one. Numbers that far apart are not measuring the same thing. We could pick whichever family flattered the story, average them into a new number nobody could check, or build our own table from shop websites and call it research. All three are how the existing tables got here, and none of them survives being asked where the number came from.
A perfect table still would not tell you your rate
Suppose the data problem were solved tomorrow. A state average would still be the wrong shape of number for this decision.
State averages blend metro and rural markets, dealerships and independents, general repair and European specialists. A shop downtown and a shop forty minutes out can both be priced correctly thirty dollars apart. The average describes a market that no individual shop actually operates in.
Worse, it invites the most expensive mistake in this business: pricing against your neighbors instead of against your costs. The shop down the road charging $115 may have paid off their building in 2004. Matching them is not competing with them. It is inheriting a cost structure you do not have.
The arithmetic that actually decides it
Your rate has to clear three things, in this order.
- Your real monthly costs. Rent, payroll and burden, insurance, software, utilities, tooling, the loan. All of it, not the ones you remember.
- The hours you can genuinely bill. Not bay-hours times opening hours. Techs take vacation, cars wait on parts, and diagnostic time goes unsold. The gap between hours you are open and hours you invoice is where most rate math quietly fails.
- The profit you intend to keep. Decided on purpose, not whatever is left.
Costs divided by billable hours is your break-even — the rate at which you work for nothing. Your door rate is that number plus the margin you chose. It is not complicated arithmetic, but it is arithmetic almost nobody does annually, which is why the free labor rate calculator exists: put in the three inputs, and it returns your break-even and a suggested door rate. No email, and nothing leaves your browser.
The number that bites: effective labor rate
Your door rate is what is on the sign. Your effective labor rate is what you actually collected, divided by the hours you actually billed.
These are never the same number. Discounts, comebacks, courtesy hours, a job quoted at 2.0 that took 3.5 — every one of them lands between the two. A shop posting $130 and collecting $104 is running a 20% discount it never decided to offer, and the sign is not the problem.
This is the one measurement worth doing weekly, and it is also the honest reason a rate increase sometimes changes nothing: if the leak is between door rate and ELR, raising the door rate raises the leak too. The calculator’s second tab works out your ELR, and comparing the two tells you which problem you have — we walk the full worked example, leak by leak, in a companion post.
When to raise it
Once a year at minimum, and whenever a real cost moves — a lease renewal, a tech raise, an insurance jump.
Rates that sit still are not stable, they are sliding. Costs drift a few percent a year, so a rate untouched for three years is roughly 10 to 15 percent underwater without anyone having decided to discount anything. That is the same shape of problem as the ELR gap: a price nobody chose.
Where BayDocket fits
Nothing above needs our software. The calculator is free and ungated, and so is the parts matrix calculator if parts pricing is the other half of your margin question — see our parts markup matrix post for how that side of the math works, and how profitable a shop actually is once both sides are added up. If your techs are paid on flag hours rather than the clock, flat rate vs. hourly covers how that changes the incentive underneath this same rate.
What BayDocket does is keep the rate you set from leaking once it hits the day. Labor lines carry your rate onto every estimate, customers approve jobs by text before work starts rather than after, and owner reports show billed hours and labor revenue per technician — which is your effective labor rate, from your own repair orders, without a spreadsheet. It is one flat price with every feature included.
If you want to see it working rather than read about it, open a loaded demo shop — real-looking data, no signup, no sales call.