Your 45% Job Could Really Be a 24% Job. Here's Where the Margin Goes.

Commercial and residential service jobs get bid on a margin that feels real. The estimate says 45% gross. You win the work, you do the work, and at year-end the company margin is nowhere near 45%. The money did not vanish. It leaked out through four costs that almost never get allocated back to the specific job that incurred them.

Here is the waterfall, and how to rebuild it from your own numbers.

The 45% that becomes 24%

Start with a fictional job bid at a 45% gross margin, say $10,000 of revenue and $4,500 of planned gross profit. Watch what happens as the real costs land. The point range in each leak heading below is an illustrative planning assumption, not a measured median from any population.

Leak 1: labor that was never fully burdened (about 9 to 11 points)

Most bids price the billable on-site hour at something close to the tech's wage plus a rough markup. Two things get missed. First, the burden: payroll taxes, workers comp, and benefits add materially to base wage. BLS Employer Costs for Employee Compensation puts total benefits near 30% of compensation, and construction workers-comp class rates push the all-in load higher than that (BLS ECEC). Second, the non-billable day: a meaningful share of a paid day never reaches an invoice. The Level Index measures the invoiced end of that directly: across 963 companies with at least 100 logged hours, the median contractor invoices 97.1% of the hours logged against jobs, and the bottom decile only 66.9%. That billing capture figure divides hours invoiced by hours logged on jobs, not by total paid hours, so it is not a billable share or utilization rate. Hours that never make it onto a job at all sit outside it. When you divide the fully burdened daily cost across only the billable hours, the true cost per billable hour is well above the wage you bid against.

For reference, the BLS median wage is about $29 per hour for HVAC mechanics (SOC 49-9021), $31 for plumbers (47-2152), and $30 for electricians (47-2111) (BLS OEWS). Multiply by an illustrative 1.35 to 1.60 for burden (payroll taxes, workers comp, and benefits; your payroll register, workers comp policy and benefit invoices give your real multiplier), then load only the billable hours, and the real number climbs fast. Only the gap between this loaded cost and the labor cost already in the bid reduces margin. If the bid already used a loaded rate, do not subtract the burden a second time.

Leak 2: callbacks and warranty labor (about 4 to 6 points)

Return trips get booked as new dispatches or absorbed silently, so the original job keeps its clean margin on paper while a second crew visit eats the profit. We have no defensible public figure for callback rates, and the vendor numbers circulating for it are not sourced well enough to publish. Measure your own: count return visits against first-time completions for one month. If you are not charging that return labor back to the original job, that job's real margin is lower than the report shows.

Leak 3: material price drift between bid and buy (about 2 to 3 points)

Fixed-price bids lock the sell price, but the material cost floats between the estimate and the purchase. Construction input prices have moved several points over six-month windows in recent years (BLS Producer Price Index, inputs to construction). On a job quoted weeks before the material is bought, that drift comes straight out of margin.

Leak 4: the financing cost of slow receivables (about 1 point)

A job you collect on in 35 to 55 days (an illustrative range) is a job you financed for the customer. At a bank prime rate around 6.75% when this was written (Federal Reserve H.15; check the current figure) plus a small-business spread, the carry on that receivable is a real cost that never shows up on the job P&L. Fictional arithmetic: a $10,000 job collected in 45 days at a 10% all-in cost of capital costs 45 / 365 x 0.10 x $10,000 = about $123, or about 1.2 points of revenue. It is small per job and large across a year of open receivables.

Add the four illustrative leaks and the fictional 45% bid margin lands between about 24% (top of every range: 11 + 6 + 3 + 1 = 21 points) and about 29% (bottom of every range: 9 + 4 + 2 + 1 = 16 points). On the $10,000 job, that is $2,400 to $2,900 of profit instead of $4,500. The AR carry is a financing cost below gross margin, so the result is an economic job margin, not the gross margin on a standard P&L, and it is profit, not cash in the bank.

How to rebuild your true margin from your own books

You do not need new software to check this. You need one job and four adjustments.

1. Start from the quoted gross margin on a representative recent job (say 45%).
2. Re-cost the labor. Take the field wage, multiply by 1.35 to 1.60 for burden (payroll taxes, workers comp, and benefits), then divide the day's cost across only the billable hours (compute this from your own paid, job-logged and invoiced-hour records, with matching periods), not the clock hours. Recompute the job at that rate, and subtract only the difference between this loaded labor cost and the labor cost already in the bid. If the bid already used a loaded rate, that difference may be near zero.
3. Subtract callbacks. Multiply your own measured callback rate by the average return-trip labor cost, and spread it across your jobs, or charge each return visit back to the original job.
4. Subtract material drift. Compare the purchase-order or vendor-bill cost with the estimate line for the same materials on the same job. The difference is that job's drift.
5. Subtract the AR carry. Take (average days to collect divided by 365) times your annual cost of capital (roughly 9 to 11% all-in is an illustrative assumption; use your actual borrowing cost) times the job's revenue.

The gap between the bid margin and the rebuilt margin is where the business quietly loses money on jobs that looked fine. This is the same phantom-margin problem that appears when direct costs are not allocated to each job, and it traces back to systems that never reconcile, which is why your field platform and accounting never agree.

Why this is directional, and why that is fine

The 45-to-24-to-29% walk is an illustration of the mechanism, not a measured population statistic. Wage and benefit references have named external sources; callback rate and billable share need your own measured inputs, and the illustrative point impacts are not published population measurements; the specific point-impact of each leak depends on your trade, mix, and discipline. The point is not the exact number. The point is that the gap is real, it is large, and it is invisible until you rebuild the job with fully loaded costs.

Related reading:

• Why you don't know your real job margin
• Why QuickBooks job costing is broken for contractors
• Gross profit per hour: the complete guide
• Callback cost calculator