Revenue per employee is the pitch

Headcount used to be how a company demonstrated it was working. Hiring was evidence of traction, and an org chart that grew each quarter was the visible form of progress. That signal has inverted, and founders are now asked to defend the number rather than announce it.

The metric that replaced it is revenue per employee. Benchmarking of software companies describes a wide spread in recurring revenue per head across stages, with the median sitting far below the figures the best-performing companies report.

What moved the expectation is the tail rather than the median. A small number of companies have reached substantial revenue with tiny teams, and reporting on firms built around AI leading on revenue per employee has made those figures the reference point in a fundraising conversation.

That reference point is doing more work than it can bear. A company selling a developer tool with self-serve distribution and no field sales operates in a structurally different business from one selling to hospitals, and comparing their revenue per head compares the markets rather than the teams.

The underlying change is real even where the comparison is unfair. Work that required a person now requires a person supervising a system, and that shift is concentrated in exactly the functions that used to grow fastest with revenue: support, marketing operations, sales development, junior analysis.

Founders describe the practical version as hiring later rather than hiring less. The role gets defined, the work gets done without a hire for two quarters, and the eventual hire arrives into a function that already has a working process rather than being asked to invent one.

The cost of that discipline lands on early-career workers, and the industry has been slow to say so. The jobs being absorbed are the ones people used to take to learn the business, and no equivalent entry point has appeared to replace them.

There is also a failure mode that the metric encourages. A team held deliberately small under a growing load produces the same outcome every time: the people who understand the system leave, and the efficiency that looked structural turns out to have been borrowed.

Investors are aware that the number can be gamed and mostly look at its trajectory instead. Revenue per head rising as the company grows is a signal about the business; a high figure at twelve people is a signal about the stage.

The honest reading is that a genuine improvement has become an unreasonable expectation on the way to being a norm. Some companies are structurally more efficient than their predecessors, and most companies are being asked to look like the ones that are.

Founders raising now should be able to answer why the team is the size it is, in either direction. That question was not asked three years ago and it is now the second thing in the room.