Payroll as a percentage of revenue: what's normal?

Payroll is usually the biggest number on the expense side and the one owners have the least confidence about. Everyone wants a benchmark, and the benchmarks that circulate — "payroll should be 30% of revenue" — are close to useless, because they are quoted without saying what is in the numerator.

Before you compare yourself to anything, decide what you are measuring. Then the comparison becomes worth making.

Define the numerator before you compare

Three different numbers all get called "payroll," and they can differ by twenty points of revenue:

• Wages only. Gross pay, nothing else. The lowest number and the least useful, because it ignores what employment actually costs.
• Fully loaded labor. Wages plus employer taxes, benefits, workers' comp. Typically 20–30% above wages, and the number you should actually manage.
• All labor. Fully loaded payroll plus contractors and temps. The only version that is comparable across time if you shift work between employees and contractors — which most growing businesses do.

Use the third. Otherwise you can "reduce payroll" by moving people to 1099 status and congratulate yourself on a cost saving that never happened. And whichever you pick, decide whether the owner's own wage is in or out, then stay consistent — on a small team that one choice can swing the ratio five points.

Split it before you judge it

A single payroll percentage blends two costs that behave completely differently. Production labor scales with volume and belongs in COGS. Administrative labor is a fixed overhead that does not move when sales do.

Blended together, a rising ratio is uninterpretable. Split apart, it is obvious. Production labor climbing as a share of revenue means your delivery is getting less efficient or your pricing has slipped. Administrative labor climbing means overhead is growing faster than the business — a completely different problem with a completely different fix.

What the ranges actually look like

With those caveats, here is roughly where fully loaded total labor tends to land by business shape. Treat these as orientation, not targets — your own trend is far more informative than any of them:

A restaurant at 45% has a problem. A software company at 45% is unremarkable. This is why an unqualified benchmark does more harm than good — it makes healthy businesses anxious and complacent ones comfortable.

Two ratios that beat the percentage

The share of revenue is a fine monitoring number and a poor management one, because it moves when revenue moves even if you did nothing to labor. Two alternatives tell you more:

Gross profit per employee is the honest version of revenue per employee — a business can raise revenue per head by taking on low-margin work and get worse while the metric improves. The labor efficiency ratio, gross profit divided by total labor cost, is the cleanest single read: it tells you how many dollars of gross profit each dollar of labor produces, and whether that is getting better or worse.

Read the direction, not the level

The useful question is almost never "is 44% too high." It is whether your labor cost is growing faster than the gross profit it generates, and if so, why.

There are good reasons for the ratio to rise — you hired ahead of revenue deliberately, you brought outsourced work in-house, you invested in a team for next year. Those are decisions. What you are watching for is the version nobody decided: headcount that accreted one reasonable hire at a time until overhead outgrew the business. That one only ever shows up in the trend, which is why the number is worth putting on a page you look at every month.