Profit Margin for a Service Business: Which One?

There is no single profit margin for a service business, and that is the whole problem. A service business has at least four of them: gross margin, operating margin, net margin, and net margin after your own labour is priced in. They are computed from the same twelve months of entries and they routinely land twenty or thirty points apart. The question worth asking is never "is my margin good", it is "which margin answers the decision in front of me". Pricing a job is a gross margin question. Dropping a service line is an operating margin question. Taking money out of the business is an owner-paid net question. Use the wrong one and the arithmetic will be correct and the decision will still be wrong.

That is also why the benchmark you found last night disagreed with the benchmark you found this morning.

Why do published margin benchmarks disagree?

Three reasons, and none of them is that somebody is lying.

They measure different margins. One page quotes a gross margin, another quotes net, a third quotes operating, and all three print the word "margin" in the headline. For a service business the spread between those is enormous, because service businesses carry almost no cost of goods and an enormous overhead line. A consultancy can run a 70% gross margin and a 6% net margin in the same year, and both figures are honest.

They measure different sizes of company. Benchmark tables are usually built from public company filings, because that is the only data anyone can get at scale. Those are firms with finance departments, salaried executives, and audited accounts. Their cost structure has almost nothing in common with a three-van operation.

They treat the owner's pay differently, usually without saying so. This is the big one, and it is covered in its own section below. In a public company, every person who works there is paid through an expense line, including the chief executive. In a sole proprietorship, the owner's work is typically not an expense line at all. Two businesses can do identical work at identical prices, and one reports a 5% margin and the other reports 45%, purely on this.

Here is what a sourced benchmark actually looks like. The NYU Stern School of Business publishes a margins by sector dataset built from US listed companies, analysis dated January 2026, covering 5,994 firms, or 4,822 once financial companies are excluded. Across the whole 5,994-firm sample the net margin is 9.74%. Inside it, Computer Services (64 firms) shows a 24.26% gross margin and a 4.45% net margin. Engineering and construction (48 firms) shows 15.46% gross and 5.94% net. Advertising (52 firms) shows a 36.24% gross margin and a net margin of minus 0.30%.

Read that last one again. Fifty-two listed advertising firms, in aggregate, made no net profit at all. If you run a small agency at 8% net and you have been told the industry average is 30%, the 30% figure was not describing your industry, your size, or your margin.

Every one of those companies pays its staff, including its founders, through the expense lines. So their net margin is already after labour. Yours, if you are the labour, is not.

What's the difference between gross and net margin for services?

Gross margin is revenue minus the costs that exist only because you did the work, divided by revenue. Net margin is revenue minus every cost in the business, divided by revenue. Operating margin sits between them: gross profit minus your fixed running costs, before interest, tax, and anything unusual.

The IRS keeps the same split, and it is a useful sanity check on which bucket a cost belongs in. Publication 334, Tax Guide for Small Business handles cost of goods sold and business expenses as two separate chapters rather than one pile. Materials consumed by a job sit in the first. Your insurance renewal sits in the second.

For a service business the practical test is a question: if I had not taken this job, would this cost have happened anyway?

No, it only exists because of the job. Direct cost. Subcontractor for that job, materials consumed, the parking and fuel for that visit, a job-specific permit, the merchant fee on that payment. These sit above the gross margin line.
Yes, it happens whether I am busy or idle. Overhead. Insurance, software, rent, your phone, accounting, the website, the van payment. These sit between gross and operating.

A worked example, with round numbers I have invented purely to show the arithmetic. These are not benchmarks and not anyone's real figures.

A two-person cleaning business bills $200,000 in a year. Direct costs, meaning subcontracted cover, supplies, and mileage, come to $60,000. Fixed overhead, meaning insurance, software, vehicle, phone and accounting, comes to $44,000. The owner works full time in the business and takes $70,000 out as a draw.

Measure	Arithmetic	Result
Gross margin	(200,000 - 60,000) ÷ 200,000	70%
Operating margin	(200,000 - 60,000 - 44,000) ÷ 200,000	48%
Net margin (owner's draw not an expense)	same as operating here	48%
Owner-paid net margin	(200,000 - 60,000 - 44,000 - 70,000) ÷ 200,000	13%

Same business, same year, same entries. 70% and 13% are both true. If you publish the 70% you sound like a software company. If you compare the 48% against a benchmark built from firms that pay their executives a salary, you will conclude you are extraordinary, and you are not. The 13% is the one that describes what the business earns over and above paying you.

Where does the owner's pay sit, and why does it change everything?

For most small service businesses, the owner's labour is the single largest input to the work and it appears nowhere in the accounts.

That is not sloppiness, it is structure. The IRS describes a sole proprietor as someone who owns an unincorporated business by themselves, reporting business profit or loss on Schedule C with their personal Form 1040 and computing self-employment tax on Schedule SE. The profit line and the owner's pay are not two separate things there. Money you move from the business account to your personal account is a draw, not a payroll expense. If your business is structured so that it pays you as an employee, your wage genuinely is an expense and your net margin already includes it. That is the single fact that decides whether your net margin is comparable to anyone else's, and it is the fact benchmark tables almost never state.

Where that line falls for your particular structure is a bookkeeper or accountant question, not a blog question, and anything touching filing, deductions or self-employment tax belongs with them or with the IRS publication above. What is safely yours to run is the arithmetic: pick a number that your own hours would cost you to replace, subtract it, and look at what is left.

The replacement number does not need to be precise. Ask what you would have to pay someone competent to do your hours, at the going rate for that work in your area, and use that. Get it roughly right and the owner-paid margin becomes the most honest number in the business. It answers the only question that matters over a five-year horizon: is this business earning anything beyond buying me a job? The same split runs through the three answers in is your business profitable, and it is why break-even for a service business produces two numbers that often sit four or five times apart.

Which margin should drive a pricing decision?

Here is the decision tree. Find the decision you are actually making, then read across.

The decision in front of you	The margin that answers it	What you compute	The mistake it prevents
Pricing a single job or quote	Gross margin on that job	(price - costs caused by that job) ÷ price	Winning work that loses money before overhead is even considered
Deciding whether to keep or drop a service line	Operating margin for that line	(line revenue - its direct costs - its share of overhead) ÷ line revenue	Cutting a line that was carrying overhead, and making the whole business worse
Deciding whether you can hire	Operating margin, whole business	(revenue - direct costs - overhead) ÷ revenue	Hiring out of a gross margin that never had to pay rent
Deciding what you can take out	Owner-paid net margin	(revenue - direct costs - overhead - your market-rate labour) ÷ revenue	Drawing a wage out of a business that isn't earning one
Comparing yourself to any published figure	Whichever margin that figure used	match their definition first, or don't compare	Beating or missing a benchmark that was measuring something else

Read as a tree, it collapses to four questions asked in order:

Is the work itself profitable? Gross margin, per job. If this is negative, nothing downstream can save it and no volume will fix it.
Does the business carry its own weight? Operating margin. Gross profit has to cover overhead before anything is left.
Does it pay me properly? Owner-paid net. Overhead covered, then your labour at market rate, then see what remains.
Is there anything left to reinvest or take as return? Whatever survives step three.

Most pricing errors are a step-one failure diagnosed at step three. The owner sees a thin year-end number and raises prices across the board, when in fact two service lines were fine and one was running a negative gross margin on every job. You cannot see that from a single business-wide figure. You can see it in five minutes if your entries carry a job or category tag, which is the main reason pricing your services works better from your own records than from a rate survey.

How do you build your own baseline from twelve months?

Your own twelve months beats any published average, because it matches your size, your structure, your market, and your treatment of your own pay. It takes four steps and no software you do not already have.

Tag every entry as direct or overhead as you record it. Not at year end, when you will not remember. The one-question test above is enough. If you have not been logging at all, tracking expenses without accounting software is the cheapest starting point.
Tag income by service line or job type. Even three coarse buckets beat one. This is what turns a business-wide margin into a decision you can act on.
Run all four margins monthly, not annually. Twelve data points tell you the shape of the year. One data point tells you almost nothing, and a single bad month looks like a crisis until you see it next to eleven others.
Record your market-rate labour number once and keep it. Write it down with the date you chose it. Revisit it yearly. Consistency matters more than precision, because the trend is the signal.

After twelve months you have your own band: your normal gross margin, your normal operating margin, and how far each swings between your best month and your worst. From then on the useful question is not "is 48% good", it is "why is this month 41% when my last eleven ran between 46% and 52%". That question has an answer you can find in the entries. The benchmark question never did.

Six or eight months in, the band is already usable. You do not have to wait a full year to start reading it.

When is a falling margin actually fine?

A margin is a ratio, and a ratio falls when the denominator grows as well as when the numerator shrinks. Several genuinely good outcomes look like a decline on paper.

You hired. A new employee's wage moves into overhead the month they start, and it takes time before their billable output catches up. Margin dips, capacity rises. Judge it on the trend over two or three quarters, not the first month.
You started paying yourself properly. If you moved from a draw to a payroll wage, your net margin fell by exactly that wage and nothing about the business changed. Compare owner-paid net across the change instead, and it should be flat.
You took on larger jobs with more subcontracted content. Pass-through work drags the percentage down while adding real dollars. A 30% margin on $400,000 puts more in the bank than 50% on $180,000. Margin is a ratio, and you do not bank ratios.
You invested in something that lasts. A year of new equipment, a rebuild of the site, or training lands in one year's overhead and earns across several.

The falls that genuinely warn you look different. Gross margin drifting down across many jobs means your prices are being eroded by costs you have not repriced. Operating margin falling while revenue is flat means overhead is growing on its own. And any month where the owner-paid net goes negative while you are working full hours means the business is currently funded by your unpaid labour. That is worth knowing early, which is the argument for checking it monthly alongside the numbers worth a look every Monday rather than discovering it at year end.

Frequently asked questions

What is a good profit margin for a small service business?

There is no single credible answer, and anyone who gives you one without naming their margin, their sample and their treatment of owner pay is guessing. For scale, the NYU Stern dataset cited above covers 5,994 US listed firms as of January 2026 and puts the whole-market net margin at 9.74%, with individual service sectors in that sample ranging from 4.45% net in computer services (64 firms) down to minus 0.30% in advertising (52 firms). Those are large public companies that pay every worker, including executives, as an expense. The honest comparison for a small service business is against its own previous twelve months, on the same margin, with your own pay handled the same way each time.

Should the owner's wage come out before margin?

Run both and label them. Standard net margin, the one that matches published figures for incorporated businesses, includes wages paid to employees. If you are a sole proprietor taking a draw, your standard net margin does not include your labour, so it is not comparable to those figures. Owner-paid net margin subtracts a market rate for your hours and answers whether the business earns anything beyond employing you. For decisions about taking money out, expanding, or selling, the owner-paid number is the one that matters. How your own compensation is treated for tax purposes depends on your structure and belongs with a bookkeeper or accountant.

Why is my margin lower than the industry average?

Usually because you are not comparing the same thing. Check three things in order. First, which margin the average used, since gross and net can be forty points apart in services. Second, what size of company it sampled, since most published tables are built from public filings. Third, whether owner compensation is inside or outside their number and yours. Once those three match, the comparison becomes meaningful, and often the gap has already vanished. If it survives all three checks, the useful next step is splitting your own figure by service line, because a business-wide average usually hides one line doing badly and another doing fine.

How do I improve margin without raising prices?

Work down the tree rather than across the whole business. Start with direct costs on your worst-margin service line, since that is where a small change shows up on every job in that line. Look at the jobs that consistently overrun their estimate and find out what they have in common. Then look at overhead that has grown quietly, meaning subscriptions, insurance renewals that reprice annually, and anything you signed up for once and never revisited. After that, look at mix: taking one more of your high-margin job type and one fewer of your low-margin type moves the average without a single price changing. Prices are one lever of four, and it is the one customers notice.

Sources

NYU Stern School of Business, Operating and Net Margins by Sector (US), analysis dated January 2026, 5,994 US listed firms, or 4,822 excluding financials. All sector figures quoted above are from this dataset with the firm count named.
Internal Revenue Service, Publication 334, Tax Guide for Small Business, which treats cost of goods sold and business expenses as separate categories.
Internal Revenue Service, Sole proprietorships, on how a sole proprietor reports business profit or loss.

Every dollar figure in the worked example is invented to illustrate the arithmetic. It is not a benchmark and not any real business's figures. Anything that crosses into filing, deductions or self-employment tax is a question for a bookkeeper, an accountant, or the IRS publications above.

Build your own baseline

Twelve months of your own entries gives you something no published average can: a margin band that matches your size, your structure, and the way you pay yourself. It is worth more than any benchmark, and it costs one tagging habit.

Open the money module of SMBDashboard free, tag each expense as direct or overhead as you enter it, and tag income by service line. There is no account to create, and your data stays in your browser unless you turn on Pro sync. The free tier holds 25 customers, 200 money entries, and unlimited tasks, which is a real stretch of records rather than a demo. Pro removes the caps and adds CSV export, recurring entries, and your own branding on the printed report, at $48 every six months (about $8 a month) or $149 once.