Revenue Per Employee Benchmarks by Stage

Private SaaS companies report a median of roughly $130K to $190K of ARR per employee depending on the survey, with $200K the median bar for well-run companies and $300K and above best in class. Public software companies sit around $400K, and the largest technology companies exceed $2M. The useful benchmark, though, is not any of those figures. It is the ratio of revenue per employee to fully loaded cost per employee: below 1x is burning, around 1.5x is typical for a venture-backed company at Series A or B, and 2x and above is what capital-efficient software companies reach.

This page gives the published numbers for venture-scale software companies, states which definition each one uses, and then shows how to turn them into a benchmark that actually applies to your company. Small businesses outside software are a different comparison, covered in revenue per employee for small businesses. If you want the calculation itself and its caveats, start with what revenue per employee is.

The published benchmarks

Each row below comes from a different dataset with a different definition, and each reports a median rather than an average, because a handful of outliers move an average of this ratio a long way. Read the third column before comparing your own number to the second.

Two things stand out. The private median has moved sharply in a single year, which is unusual for a ratio that normally drifts, and the whole increase is on the headcount side. And the gap between the private and public medians is roughly two to three times, which is structural: a public company has amortised its product and go-to-market investment across a revenue base large enough that the fixed costs dominating a startup are a rounding error.

What the number means at each stage

Benchmarks by ARR band are more useful than a single median, but they still hide what is happening inside the company. Here is the pattern by stage, and what a founder should take from it.

Pre-seed and seed, under $1M ARR. The ratio is mostly noise. Twelve people and $600K of ARR gives $50K per head, and one enterprise deal moves it by half. Do not benchmark. Track headcount against the milestones the next round requires, and watch whether each hire was a choice or a reflex. The question at this stage is how many people you actually need, not what the ratio says.

Series A, $1M to $5M ARR. The number starts to mean something, and it is usually at its lowest here, because the company is hiring ahead of revenue on purpose. Expect $80K to $150K. The signal is direction: if the ratio is falling as ARR grows, the company is adding people faster than it is adding customers, and the reason should be a deliberate bet rather than a habit.

Series B, $5M to $20M ARR. This is where the benchmark becomes a management tool. The published band puts the median around $150K, and the well-run companies are at $200K or above. A Series B company below $120K is usually carrying roles that exist because someone was busy rather than because the work required a person, and this is also when the hiring plan is largest, so the cost of not asking is highest.

Series C and beyond, $20M+ ARR. $250K to $300K becomes the expectation, and the companies that reach it are the ones that treated headcount as a design variable from Series B onward. The ones that do not tend to discover it at the next fundraise, when an investor puts the ratio next to a comparable and asks why.

How much should an employee generate?

This is the version of the question that founders actually mean, and the external benchmarks do not answer it, because they say nothing about what your people cost. The number that does is the ratio of revenue per employee to fully loaded cost per employee.

Fully loaded cost is salary plus payroll taxes, benefits, equipment, software seats, and the share of office or tooling spend that scales with headcount. For a US software company it typically lands between 1.25x and 1.4x base salary. If the average base across the company is $150K, the loaded figure is around $200K, and that is the denominator that matters.

The worked version: a Series B company with $12M of ARR and 70 people is at about $171K per employee, roughly the survey median for its band. If its loaded cost per employee is $190K, the ratio is 0.9x, and the "at benchmark" headline hides the fact that the company is paying more for people than they generate. Another company at the same $171K with a loaded cost of $120K is at 1.4x and in a different position entirely. Same benchmark, opposite conclusions. That is why the external figure is a starting point and the internal ratio is the metric.

Using the benchmark without being misled by it

• Match the definition. ARR against ARR surveys, trailing revenue against public company figures. Contractors doing ongoing work count as headcount in your number even when the survey excluded them.
• Benchmark the trend, not the level. A company moving from $110K to $160K over four quarters is doing something right regardless of where the median sits. One sitting at $200K and falling is not.
• Never improve it by cutting. Reducing the denominator moves the ratio instantly and removes the capacity that produced the numerator. The levers that work are covered in how to improve revenue per employee without cutting headcount.
• Ask the question before each hire, not after. The ratio is most useful as a filter on the hiring plan. If a proposed role does not plausibly raise revenue per employee within four quarters, the alternative is automation, a restructured team, or a super IC absorbing the work.

Common questions

What is a good revenue per employee for a SaaS startup?

For private SaaS companies the survey median sits in the low-to-mid $100Ks of ARR per employee, with $200K now the median bar for well-run companies and $300K and above best in class. Below $5M ARR the number is usually lower and far noisier, because a company of twelve people is one hire away from moving it 10%.

What is a good revenue per employee ratio?

Measured as revenue per employee divided by fully loaded cost per employee, a ratio of 1.5x is typical for a venture-backed company at Series A or B, 2x and above is capital efficient, and anything below 1x past $5M ARR means people cost more than they bring in. Measured against published medians, the average revenue per employee for private SaaS sits in the $130K to $190K range, with $200K the bar for a well-run company.

How much money should an employee generate?

Enough that revenue per employee covers fully loaded cost per employee with room for the margin the business needs. As a working rule: below 1x loaded cost the company is burning to grow, around 1.5x is typical for a venture-backed company at Series A or B, and 2x and above is what capital-efficient software companies reach. The ratio to your own cost base is more useful than any external figure.

Why is revenue per employee so much higher at public companies?

Scale, mostly. A public software company has amortised its product and go-to-market investment across a much larger revenue base, and the fixed costs that dominate a startup are a small share of headcount at scale. The public median is roughly two to three times the private one, and that gap is structural rather than a sign the startup is inefficient.

Do AI-native companies have higher revenue per employee?

The early evidence says yes, and by a wide margin, but the samples are small and definitions vary. The mechanism is what matters: AI-native teams keep the denominator flat while revenue grows, which is exactly what the ratio measures. Whether a given company achieves that depends on how the work is structured, not on which tools are on the invoice.

Should I use ARR or revenue when benchmarking per employee?

Use whichever the benchmark you are comparing against uses, and say which. Most private SaaS surveys report ARR per employee. Public company figures are usually trailing twelve-month revenue. Mixing the two produces a comparison that looks precise and means nothing.