Every new hire looks like added capacity, but onboarding is not free: existing staff must explain the codebase, correct early mistakes, and route decisions through people who don't yet share the same mental model. Fred Brooks quantified this in what became known as Brooks's Law: as a team grows, the number of communication pairs grows roughly as n(n-1)/2, so coordination overhead rises faster than the headcount that's supposed to fix it. Past a certain team size, each additional hire subtracts more senior-engineer time to training than they contribute in original output, at least for months.

Brooks discovered the effect firsthand while managing IBM's OS/360 operating system project in the mid-1960s. When the schedule began slipping, IBM's answer was to add more programmers — exactly the intuitive move most leaders still make today. The new hires needed months to learn a codebase built by people who had already internalized its quirks, and every hour a veteran spent explaining the system was an hour not spent finishing it. Communication paths multiplied faster than output did, and OS/360 shipped later than if IBM had left the original team alone. Brooks later wrote up the lesson in The Mythical Man-Month: “adding manpower to a late software project makes it later.”

The same dynamic plays out in fast-scaling startups today. A team that doubles its headcount to hit a deadline usually slows down first, not last, because the incoming engineers draw down the bandwidth of the people who could otherwise be shipping. Recognizing hiring as a short-term tax on senior capacity — not an instant multiplier — is what separates teams that stay on schedule from teams that quietly fall further behind while their org chart grows.