The Mythical Man-Month

by Frederick P. Brooks ยท 9 ideas

  1. The Mythical Man-Month

    People and months can't be swapped one for one. Every new person needs training and adds more lines of communication, so the work grows with the team. Never plan as if twice the people means half the time.

  2. Brooks's Law

    Adding people to a late project makes it later. Newcomers slow down the experts who must teach them, and coordination costs rise. When you're behind, cut scope or move the deadline instead.

  3. Optimism Bias

    Builders assume each task will go as planned, so estimates come out short. Problems you can't see yet always show up. Add real buffer and treat your first estimate as the best case.

  4. Late One Day at a Time

    Projects rarely slip in one big disaster. They fall behind through many small delays that each seem harmless. Set sharp milestones that are clearly done or not done, and react to small slips right away.

  5. Conceptual Integrity

    A system feels simple and coherent when one mind, or a small group, shapes its design. Many designers produce a patchwork of good but clashing ideas. Keep the core design in few hands, even if it means leaving out some good features.

  6. Surgical Team

    Organize a team around one lead builder, like a surgeon, with others supporting them in clear roles. This keeps the work in few heads while still adding capacity. Small, focused teams beat large crowds of equals.

  7. Second-System Effect

    Your second version tends to be the most dangerous. You stuff in every idea you held back the first time, and it becomes bloated. Be extra strict about features when building a follow-up.

  8. Plan to Throw One Away

    Your first version mostly teaches you what you should have built. You'll end up rebuilding it whether you plan to or not. So expect it, and design your process and system to handle change.

  9. No Silver Bullet

    No single tool or method will make building software ten times better. The hard part is understanding the problem and its complexity, not the typing. Improve through many small gains instead of waiting for a miracle.

More on work