Know which jobs make money while you can still fix the ones that don't

Why general accounting software is not job costing

This is the confusion that costs contractors the most money. QuickBooks, Xero, and the like are excellent at telling you how the whole business performed last month. They add up income and expenses across everything you did and hand you a profit-and-loss statement. That is essential, but it is the wrong altitude for running jobs.

Job costing works one level down. It asks a narrower, more urgent question: is this specific job making money, right now, with the work that has been done so far? The difference is timing and granularity. Accounting is a rear-view mirror for the company. Job costing is a live gauge for each project. You need both, and they do genuinely different jobs.

The practical failure happens when a contractor tries to force job-level answers out of company-level software. You can tag transactions by job in QuickBooks, and many do, but you are still reconciling after the fact, the field has no easy way to feed it, and change orders and committed-but-unbilled costs slip through. By the time the P&L shows a job lost money, the crew is three projects away.

What real-time job costing actually tracks

A complete job cost has four direct buckets plus overhead. Good software captures all five as the work happens, not weeks later.

Labor, and why "loaded" labor matters

The wage on the check stub is not what an hour of labor costs you. Add payroll taxes, workers' comp, insurance, and benefits and the real number is often 25% to 40% higher. That is the labor burden. Costing software applies burden automatically so an hour logged in the field lands in the budget at its true loaded cost. The calculator above does the same thing with the burden field.

Materials and purchase orders

Materials leak in two directions: receipts that never get entered, and prices that crept up between bid and buy. Tracking POs against the estimate catches both. When a supplier invoice comes in above the PO, you see the variance immediately instead of at reconciliation.

Equipment

Owned equipment has a real hourly cost, including fuel, maintenance, and depreciation, that is easy to ignore because no invoice arrives. Rented equipment is simpler but still needs to land against the right job. Either way, equipment that sits idle on one job while another waits for it is pure margin erosion.

Subcontractors

Subs are usually the largest single line on commercial work, and they carry their own change-order and billing risk. Tracking committed sub costs, what you have agreed to pay rather than only what has been invoiced, keeps the job's true position honest.

Overhead

Overhead is the cost of being in business at all: the office, the estimator, the insurance, the truck payments. It has to be spread across jobs somehow, usually as a percentage of direct cost or labor. Ignore it and every job looks more profitable than it is.

How to calculate job cost (the formula the software automates)

At its core, the math is simple. It is the doing it continuously, across every job that requires software.

1. Loaded labor = hours x wage x (1 + burden %)
2. Direct cost = loaded labor + materials + equipment + subcontractors + other
3. Overhead = direct cost x overhead rate
4. Total cost = direct cost + overhead
5. Profit = contract price - total cost
6. Net margin = profit / contract price

Margin and markup are not the same number, and mixing them up is one of the most expensive mistakes in the trades. Markup is profit as a percentage of cost; margin is profit as a percentage of price. A 50% markup is only a 33% margin. The calculator above shows both so the difference is always in front of you.

The $120,000 job, costed line by line

Here is the whole costing engine on one real job: a $120,000 contract using 640 hours at a $34 wage with 28% burden, $31,000 in materials, $6,500 in equipment, $18,000 in subcontractors, $2,200 in other costs, and 12% overhead. It costs $95,819.14 and clears $24,180.86 in profit, a 20.15% margin.

Nothing in this table is a surprise, and that is the point: every line was fixed before the first crew call. Labor is the line to watch hardest, at $27,852.80 it is 29.1% of total cost and the only bucket that grows when a job drags. The dashboard preview above runs the same math live: the Riverside remodel is 68% complete, materials are $1,900 over budget, and it still clears 18.7% because the overrun showed up while there was time to respond.

Loaded labor: why a $34 wage really costs $43.52

A crew member who earns $34 an hour costs you $43.52 for every hour on the clock, because 28% labor burden rides on top of the wage. The extra $9.52 an hour is the gap between the check stub and the real cost.

That gap has a name: burden. Payroll taxes, workers' comp, insurance, and benefits get paid in the background, none of it on the check stub. On the $120,000 job, the 640 hours at $34 come to $21,760 on paper, but the loaded line reads $27,852.80, and the $6,092.80 difference is burden on this one job. Budget at the raw wage and the biggest line is understated by 28%, which is how a job that looks fine on the bid comes in thin at the closeout. Software applies burden automatically at clock-in, so the dashboard shows the loaded number, the one that matters.

Markup vs. margin: two percentages from one $24,180.86 profit

The same $24,180.86 of profit reads as a 20.15% margin and a 25.24% markup, and the two are not interchangeable. Margin is profit divided by the price; markup is profit divided by cost. The bigger cost base is what makes markup the bigger number.

The mix-up happens at bid time. A contractor who prices with a 25% markup while thinking in 25% margin is really quoting a 20.15% margin, five points thinner than planned. Mark the $95,819.14 total cost up 25% and the bid lands at $119,773.92, about $226 short of the $120,000 price. One job, a rounding error; twenty a year, real money left on the table. The calculator above prints both figures on every run so the difference is never a surprise.

Why the spreadsheet fails at job costing

A spreadsheet cannot do job costing because job costing is a timing problem and a spreadsheet is a storage problem. The site's own numbers make the case: spreadsheet tracking runs 6 to 10 days behind the work, and unbilled change orders leak 2 to 5 percent of revenue.

• It is a snapshot, not a feed. Cost data sits in paper receipts and time cards until someone types it in, usually days later and from the office.
• There is no job-level discipline. A tab per job, formulas copied by hand, one sorted column and the workbook disagrees with itself. The 6 to 10 day lag is the best case; the worst case is the wrong answer.
• Committed costs do not exist. A purchase order signed but not yet invoiced is invisible in a spreadsheet until the invoice lands, which means a job always looks healthier than it is.
• Nothing alerts. The person who would notice the materials overrun has to remember to open the file. The demo above flags its $1,900 overrun on its own at 68% complete.

Choosing job costing software: what actually matters

Job costing by industry

The engine is the same; the vocabulary and templates change.

Construction and the trades lean hardest on cost codes, committed costs, and change orders. A remodeler and an electrical contractor track the same five buckets but organize them under different code structures. Manufacturing job costing focuses on per-unit material and machine time, often with work orders instead of projects. Professional services and agencies job-cost by billable hours against a fixed fee, where labor is nearly the whole cost. The reason a single platform can serve all of them is that budget vs. actual, per job, in real time is a universal need. Only the cost codes differ.

Cost codes: the shared vocabulary that makes jobs comparable

A cost code is a fixed label for one kind of work, and it is what makes budget versus actual mean something across jobs. When every job logs hours and receipts against the same codes, one job's actuals become the next job's estimate.

Those twenty codes are the spine of construction job costing. Rough carpentry, electrical rough-in, drywall, paint: each is a separate budget line, so an overrun shows up where it happened instead of inside a lump of "labor". A foreman logging time against Electrical rough-in instead of a vague note lines today's actuals up with the estimate, gives next year's bid real history per code, and puts two project managers on the same numbers.

A field crew tracking from a phone: how it actually works

The phone in the foreman's pocket is what makes real-time job costing possible, because a cost is only live if it is captured where the work happens. Crew members clock in against a job and a cost code, photograph the receipt at the supply house, and the entry lands on the dashboard the same day.

This is the part that surprises most owners: the crew does not change how they work, they just point the camera and tap. The foreman who used to collect paper time cards now watches hours land against the right cost code before the week is over, and the office sees an overrun line while the crew is still on site. Nobody has to remember to look, because the alert does the looking; the office still decides, but the noticing is automatic, and that quiet shift is most of the value.

When job costing software does not pay off

Honest answer: job costing software earns its keep when a job can lose money without anyone knowing until it is over. It can be dead weight for a one-person crew where the owner is the estimator, the accountant, and the labor, and for service work booked by the day where there is no real budget to reconcile against.

If every job is small, short, and priced in your head, the software solves a problem you do not have. The same goes for a firm that already closes every job's books weekly; that discipline replaces the tool. The useful test is not company size but blindness: if nothing on a job changes between bid and closeout, you may not need this. The moment one job can drift $1,900 over on materials while the crew is still on site, the comparison stops being fair. A Crew plan at $29 per user per month costs less than a single change order you never billed.

Common job costing mistakes

• Using unburdened labor. Costing a job at the raw wage understates cost by a third and makes losers look like winners.
• Forgetting committed costs. If you only count invoices you have received, a job with $40,000 in signed-but-unbilled subcontracts looks far healthier than it is.
• Never billing change orders. The single most common profit leak in the trades. Do the extra work, forget the paperwork, eat the cost.
• Confusing markup and margin. Bidding on markup while thinking in margin quietly erodes profit on every job.
• Ignoring overhead. A job that is "profitable" before overhead can be a loser after it.
• Reconciling too late. A number that is accurate two weeks after the job ends cannot change any decision.