Work overhead out per field hour, not as a percentage. Total every cost that runs whether or not you sell a job, then divide by the hours your crew can put on an invoice. In a two-man residential shop that comes to somewhere near $23 an hour, and it has to be in the price before wages, before material, before a dollar of profit. NAHB's 2026 Cost of Doing Business Study put remodelers' operating expenses at 23.6% of revenue for fiscal 2024, against a 6.3% net profit. Overhead is close to four times what you keep. Carry it with a 10% allowance and you are eating the gap on every job you sell.
What counts as overhead and what doesn't?
Ask one question about each cost: would it disappear if the job did? If it would, it is job cost. If you would still be paying it in a month where you sold nothing, it is overhead.
That test sorts most of it in a minute. Lumber, the tile setter's invoice, the dumpster, the permit fee - all job cost, all gone if the customer cancels. The liability premium arrives anyway. So does the truck payment, the phone bill, the accountant, the licence renewal.
Two lines trip almost everybody.
The owner. The hours you spend quoting, ordering material, chasing invoices, driving to the supply house are not on anyone's job. They are the cost of having a business. If you draw $85,000 and two days of your week go to that, $34,000 of your own pay is overhead. NAHB counts owner's compensation inside the 23.6%, and in a one-truck shop it is usually the biggest overhead line on the page.
The small stuff nobody itemizes. Sawzall blades, drill bits, sandpaper, a tube of caulk grabbed on the way out. You never write it on a job sheet because writing it down takes longer than the caulk cost. Two thousand a year is a normal number for a shop running two men, and it is invisible until you total the receipts.
One thing that is not overhead: unbillable field time. Drive time and warranty callbacks come out through the billable-hour divisor when you set your loaded labor rate. Put them in both places and you have double-charged yourself, which feels conservative and is just wrong.
What does overhead cost per field hour?
Here is the whole calculation for a two-man shop - an owner who runs the job and works most days, plus one employee.
| Line | Basis | Annual |
|---|---|---|
| Owner's pay for non-field time | $85,000 draw, 40% off the tools | $34,000 |
| Truck, all in | 18,000 business miles at the IRS 2026 rate of 72.5¢ | $13,050 |
| General liability | Insureon median for general contractors, $162/mo | $1,950 |
| Advertising, website, lead fees | $3,600 | |
| Small tools and consumables | never charged to a job | $2,400 |
| Accounting and tax prep | $1,500 | |
| Phone, two lines | $1,440 | |
| Software | accounting, estimating, file storage | $1,800 |
| Licence, bond, continuing education | $600 | |
| Total overhead | $60,340 |
Now the divisor. The employee puts about 1,650 hours on an invoice in a year. The owner is on site most days but only about 1,000 of his hours reach one, because the rest went to the list above. Call it 2,650 field hours the shop can actually sell.
$60,340 ÷ 2,650 = $22.77 per field hour.
That shop sells roughly $255,000 a year, which puts its overhead at 23.7% of revenue - within a rounding error of NAHB's 23.6%. If your own arithmetic lands miles from that, check it before you trust it.
A note on the truck line. The IRS rate is an all-in figure: depreciation, fuel, maintenance and insurance are already inside the 72.5 cents. Contractors who claim the mileage rate and then add a separate commercial auto premium have counted the insurance twice. If you would rather itemize, AAA's 2025 Your Driving Costs study puts a half-ton pickup at 98.54 cents a mile against 79.11 for a mid-size - higher than the IRS rate, and still assuming a personal-use policy rather than a commercial one.
Is "10 and 10" a typical contractor overhead and profit?
No. It is an insurance convention that escaped into general estimating, and it is low.
"Typical contractor overhead and profit" was the fastest-rising related query for overhead and profit on Google Trends over the past twelve months in the US, up 450%. Plenty of people are looking for the rule of thumb. The rule of thumb comes from first-party property claims, where a general contractor coordinating repairs charges 10% overhead and 10% profit on top of what the subs are paid - and the long-standing adjusting practice is that it applies once three or more trades are involved.
That number was built for a GC who subcontracts the entire loss. It was never a costing method for a remodeler who self-performs, and it does not survive contact with the NAHB figure: 10% against 23.6% of revenue actually spent.
There is a second bite in it. Ten percent added to your cost is 9.1% of the price you charge, not 10%, for the reason set out in markup vs margin. So a 10-and-10 job returns under a tenth of revenue toward an overhead load that is running near a quarter of it.
Should you spread overhead as a percentage or per hour?
Per hour, if you swing the hammers yourself. A percentage of job cost quietly charges your office to your material packages. Material does not consume an office.
Take the shop above. Overhead works out to about 34% of direct job cost across the year, so a 34% markup recovers the right total. Watch what it does to two jobs of almost identical size.
| Repaint, whole house | Kitchen refresh with cabinets | |
|---|---|---|
| Field hours | 160 | 60 |
| Labor cost at $42/hr | $6,720 | $2,520 |
| Material | $1,400 | $7,000 |
| Direct cost | $8,120 | $9,520 |
| Overhead recovered at 34% | $2,761 | $3,237 |
| Overhead actually consumed at $22.77/hr | $3,643 | $1,366 |
| Difference | -$882 | +$1,871 |
The paint job runs your business for four weeks and pays $2,761 toward it. The cabinet job ties up the shop for a week and a half, pays $3,237, and the extra came out of the painter's pocket.
Then it gets worse. The same distortion sets your bids. Percentage allocation makes you cheapest in town on labor-heavy work and dearest on material-heavy work. You win the paint jobs and lose the cabinet jobs, which is exactly backwards. A contractor who cannot work out why his calendar is full and his account is empty is often looking straight at this.
Per-hour allocation fixes it in one line: $22.77 goes on every field hour you quote, the same way the wage does.
What this misses
The $22.77 is one shop's arithmetic. It is a worked example, not a benchmark. Your truck, your state, your advertising spend and whether you pay yourself a real wage will move it by ten dollars in either direction. Run your own P&L. Anything else is borrowing a number you cannot defend to a customer who pushes back.
It breaks down when you sub most of the work out. A GC who self-performs 20% of a job has very few field hours, and dividing all his overhead across them produces an hourly rate that looks absurd. That contractor genuinely does need a percentage on subcontract cost - which is what the insurance convention was groping toward, even if 10% is the wrong figure. Somewhere around 30% self-performed labor is where the per-hour method stops being the better instrument.
If you pay yourself out of what's left, this understates you badly. Owner's compensation is the largest line in the table and the only one with no invoice attached. Leave it out and your overhead rate looks like $13 an hour. You find out in February.
The Trends figure is query growth, not volume. A rising query up 450% tells you the share of searches for that phrase climbed sharply. It does not tell you how many people typed it.
Every hour of that $22.77 is decided before you leave the house, and it is the number contractors skip when a homeowner is standing there waiting for a price - the same reflex behind the free estimate you gave away on a Saturday. BidWalk carries your overhead rate into the estimate while you are still walking the job, so the office is paid for before the customer hears the total.
