The short answer: one question sorts every purchase. Would you have spent this if the job had not happened? Yes means it is a job cost and belongs against that job. No means it is overhead and belongs to the business. Mixing them corrupts your pricing and your tax position at the same time, in opposite directions.
A drywaller finishes a basement job and runs the numbers: $9,200 invoiced, $3,100 in materials, $3,800 in labor. Around $2,300 left over. Feels like a win.
The same month, his ladder rack needs replacing, the van insurance renews, his phone plan bills, the compressor gets a service, and his liability policy takes its monthly bite. None of it lands on any job. All of it comes out of the same bank account. And at the end of a month full of "profitable" jobs, the account is somehow flat.
He is not losing money on jobs. He is losing money between them, and he cannot see it, because everything he spends goes into one undifferentiated pile. The pile has a name problem: two fundamentally different kinds of cost are living in it, and every useful number in his business (his prices, his job reports, his tax return) depends on keeping them apart.
The two buckets, defined by one question
Job costs (accountants say "cost of goods sold" or "direct costs") are expenses that exist because a specific job exists. The lumber for the Hendersons' deck. The subcontracted electrician on the basement. The dump fees for that demo. The hours you and your crew put into that address. If the client had never called, the cost would never have happened.
Overhead (indirect costs) is what it costs to be in business at all. Insurance, the van and its fuel and repairs, your phone, tools and their maintenance, software, advertising, the accountant, shop rent, your license renewals. These carry on whether this month has nine jobs or none.
The sorting question is one sentence: "If this job disappeared, would this cost disappear with it?" Yes: job cost. No: overhead.
Run the edge cases through it and most of them resolve. Screws bought for this deck: job cost. The bulk box of screws that lives in the van and feeds twenty jobs: overhead, or shop stock, until used. Fuel for the special trip to the far-away site: arguably job cost; the everyday driving between quotes and suppliers: overhead. A blade worn out on one brutal concrete cut: job cost in spirit; general blade consumption: overhead. Do not agonize over the borderline items. Be consistent, and get the big categories right, because the damage from mixing is not in the edge cases. It is in the wholesale dumping of one bucket into the other.
The markup vs margin calculator converts one into the other and shows the price and profit behind both.
What mixing does to your pricing
Here is the mechanism, and it runs in both directions.
Dump overhead into jobs and your jobs look unprofitable at random. The month the insurance renews, whatever job you happened to book it against looks terrible. You conclude that bathrooms are losers, when actually bathrooms are fine and your ladder rack died. Job-to-job comparison, the thing that teaches you what work to chase, turns to noise.
Ignore overhead entirely (the more common failure) and every job looks better than it is. The drywaller's $2,300 "win" is real at the job level. But if his business carries $3,400 a month of overhead and he closes three such jobs a month, his actual position is $6,900 of job margin against $3,400 of fixed cost. That is $3,500 for the month, before his own taxes, and a far thinner cushion than three "$2,300 wins" implied. Contractors priced this way feel busy and broke at the same time, and the sensation is accurate.
The repair is to make overhead a known, priced-in number instead of an ambient leak:
- Total your real overhead for a year. Insurance, vehicle, phone, tools, software, marketing, accountant, rent, everything that is not attached to a job. Suppose it comes to $41,000.
- Count your billable capacity honestly. Not 52 weeks of 40 hours. After quoting, driving, supplier runs, admin, weather and empty weeks, a realistic solo number might be 1,300 billable hours.
- Divide. $41,000 over 1,300 hours is about $31.50 per billable hour. That is your overhead recovery rate: what every working hour must contribute before your labor earns anything and before there is any profit at all.
- Build it into your rate. If your hands are worth $55 an hour and you want a genuine profit margin on top, your charge-out rate has to clear $55 plus $31.50 plus margin (call it $100 and up) before it is doing its whole job.
Contractors who skip this arrive at rates by copying the guy down the road, whose overhead they do not know, or by working backwards from what feels chargeable. Both methods routinely produce rates that cover the visible bucket and quietly starve the invisible one.
Knowing the number also gives you the answer when a client pushes back on your rate:
"That rate isn't my wage. It carries the insurance, the licensed van, the tools and the warranty behind the work. The part I take home is a fraction of it."
The growing-crew trap
One warning for growing crews: overhead does not scale politely. A solo operator's overhead is a van and a phone; add two team members and suddenly there is a second vehicle, more insurance, more tools walking around, maybe a small shop, and software seats. Contractors who grew from solo to a crew of four often keep pricing on the overhead picture they had when they were alone, because nobody re-ran the division. Recalculate the recovery rate once a year, and any time the business changes shape: a new hire, a new vehicle, a shop lease. The twenty minutes of arithmetic is the difference between growth that compounds and growth that just makes you busier at the same bank balance.
What mixing does to your taxes
The second casualty is quieter and shows up in April.
To be clear about what does not change: both buckets are generally deductible business expenses. Mixing them does not usually change your total deduction. What it changes is whether your return is right in the details, and whether anyone (your accountant, an auditor, you) can verify it.
Timing rules differ between the buckets. Materials bought for jobs interact with inventory and work-in-progress: supplies sitting in your shop at year-end, or costs on a half-finished job that spans the year boundary, are not treated the same as the insurance bill you paid in March. If everything is one pile, your accountant has to reconstruct which pile each item belonged in, and reconstruction is billed by the hour. There are also categories with their own special treatment wherever they sit: meals at 50 percent, vehicle costs needing a business-use split, capital tools depreciating over years rather than deducting at once. A sorted ledger is what makes those adjustments findable. Exactly how each lands varies by country and province or state: confirm with your accountant.
Auditability is the bigger deal. A tax review starts with your claimed expense lines and asks what is behind them. "Materials: $61,000" backed by a clean trail of receipts each booked to a named job reads as a business that knows itself. The same $61,000 backed by a single undifferentiated expense heap (part deck lumber, part personal-ish fuel, part a mystery) invites the auditor to start pulling threads. Sorted books are not just more convenient; they are more credible, and credibility shortens audits.
And your accountant's fee tracks your mess. Every hour they spend sorting your buckets is an hour billed to do work you could have done in twenty seconds per purchase, at the counter, when you still knew the answer.
The habit that keeps the buckets clean
Everything above collapses into one operational rule: every dollar leaving the business gets its bucket assigned at the moment it is spent. Not at month end, not in April. At the counter, the question "which job is this for?" has an obvious answer. Three weeks later it has a shrug.
So: receipt captured at purchase, booked either to a named job or to overhead, one tap of a decision. Labor hours logged against the job they were worked on. Subcontractor invoices attached to the job that consumed them. When you genuinely do not know the bucket at the counter (an unusual purchase, a tool that might stay on one site), default it to overhead and note it. A monthly two-minute review of the overhead list catches the handful of items that belonged on a job, which beats holding the receipt loose until you decide. Do this and both of your critical reports assemble themselves: true per-job profitability on one side, a true overhead total on the other, and a monthly picture that finally explains where the money went.
Zeus draws this line straight through the product. Costs booked to a job (materials, receipts, subcontractor invoices, tracked hours) build that job's profitability report. Everything else goes to Company Expenses, a separate overhead ledger that never muddies a job's numbers. And because both buckets are honest, the Company Expenses report can show the figure that actually matters: profit after Company Expenses. That is your job margins with the cost of being in business already subtracted, broken down by category for the period you pick. That is the number the drywaller could not see, on one screen, every month.
Two buckets, one question at every purchase, and both your prices and your tax return start telling the truth.
Frequently asked questions
Where do my own wages go, job cost or overhead?
Your hours worked on a job are a job cost, and you should log and price them like any crew member's, at a market rate, or your job reports will flatter every job you personally worked on. What you pay yourself beyond that, the owner's draw or salary for running the business, behaves like overhead. Blurring these two is the single most common reason a solo operator's jobs all look profitable while the year does not.
Is my van a job cost or overhead?
Overhead, in almost every practical sense: it serves every job, so no single job should carry it. The defensible exceptions are job-specific vehicle costs, a rented lift for one site, an unusual haul for one contract. For tax purposes the vehicle also needs a business-versus-personal use split with its own documentation rules, which vary by country and province or state. That is a worthwhile accountant conversation.
How do I allocate overhead to jobs without faking precision?
Mostly, don't allocate it per job at all; recover it in your rate. The yearly-overhead-divided-by-billable-hours method gives you a per-hour recovery figure to build into pricing, which achieves the goal (every job contributes) without pretending you know that this bathroom consumed $212.40 of insurance. Per-job allocation matters more for large bids and multi-month contracts, where your accountant can help you pick a sane method.
Does the split actually change what I owe in tax?
Usually not the total deduction, since both buckets are generally deductible. But it changes timing in cases involving year-end inventory and work in progress, it determines whether category-specific rules (meals, vehicle, capital items) get applied correctly, and it massively changes how defensible your return is under review. Think of the split less as a tax trick and more as the difference between books that can answer questions and books that cannot.
ZEUS