Executing a large plan flawlessly can still produce something nobody wants, on time and on budget. Woody Zuill's framing: "We learn what to do next by doing the current thing." Releasing in small increments and getting feedback makes the original plan moot, because the plan was a guess made with the least information you'll ever have. This isn't an argument for chaos. Discipline stays essential; what goes away is the traditional tactical plan with its estimates, milestones, and big upfront designs. A flexible strategic goal still matters: "strategy provides a direction, not a fixed route." Design and architecture still matter, just not as work done entirely before you have the information to make good decisions. Exceptions exist: huge organizations and regulated environments need more planning, though even there the required rigidity is less than most people assume. ## Cross-Domain Applications **Software delivery**: sits in tension with [[Agile Degeneration Into Speed Theater]] -- that note argues agile degenerated by cutting design and completeness for velocity; this note argues the fix is disciplined iteration, not more upfront planning. Both agree discipline, not planning volume, is the actual lever. ## Source - [[Planning is overrated]] -- Allen Holub, X, 2026-06-29 -- https://x.com/allenholub/status/2071639049782087738