The Mythical Man-Month
Essays on Software Engineering
The big idea
Brooks ran the OS/360 project at IBM in the 1960s — thousands of people, years late, one of the largest software efforts anyone had attempted. When it slipped, he did the obvious thing and added programmers. It got later. He spent the next decade working out why, and the answer became Brooks's Law: adding manpower to a late software project makes it later.
Why it matters
The word 'man-month' is the lie in the title. It assumes people and time trade off cleanly — that twelve people finish in a month what one finishes in a year. That works for picking cotton. It doesn't work when the parts have to talk to each other. Brooks made the point with the obvious biological example: however many women you assign, a baby still takes nine months. Two costs eat the new hires. They need training, and the only people who can train them are the ones already behind. And communication paths grow as n(n-1)/2 — go from 4 people to 8 and the pairs who need to stay in sync jump from 6 to 28. The work doesn't split; it multiplies.
Use it today
Next time something is late, notice that your first instinct is to add bodies. Ask instead what you can cut. Brooks's real conclusion is that scope is the only lever that moves fast — headcount is the slow one, and past a certain point it goes backwards.
Drop it in conversation
“There's a fifty-year-old book about IBM's mainframe OS that basically ended the idea that you can buy your way out of a deadline — the guy who ran the project added programmers when it slipped and it got later, because every new person needs training from the people who are already behind.”
In real lifeA restaurant kitchen slammed on a Friday night doesn't get faster if you push three more cooks in — someone has to stop and tell them where the plates are, and now four people are reaching for the same pan. A group project where two people were doing the work rarely speeds up when four more join a week before the deadline; it turns into a meeting about who's doing what. A house renovation on a tight timeline doesn't finish sooner with a second crew if the drywall still can't start until the wiring passes inspection. In each case the honest fix is the same and it's the unpopular one: decide what isn't getting done.