Critical Path

What forms the path

Among all routes through the dependency network from start to finish, the critical path is the one with the longest total duration. Say A (5d) is followed by B (3d) and C (2d) in parallel, and D (4d) starts only after both: A-B-D totals 12 days, A-C-D totals 11. The critical path is A-B-D, and tasks on it normally have zero total float. Where constraints — fixed start dates, deadlines — collide with the logic, float can go negative — not a record that execution is late, but the schedule model's warning that the constraint date cannot be met with the current logic and durations.

Why it matters

It sets the priority of management attention. Shortening work off the critical path does not pull the finish date in by a single day — cut C from two days to one in the example above and the project still takes 12 days. Both the fight to hold the finish date and the fight to pull it in happen on the critical path. When choosing where to add people, where to consider overtime, and which tasks to ask about daily, the first question is always "is it on the critical path?"

Common misconceptions

Treating the critical path as a fixed list computed once is the misconception. When a task slips or a duration changes, the path itself can move: stretch C to four days above and the critical path becomes A-C-D. Watching only the critical path is risky too — near-critical paths with a day or two of float turn critical after one small slip. Finally, it is not a list of "important tasks" but of "tasks with no time to spare": important work can have float, and trivial work can have none. Managing the critical path therefore starts with a habit: after every recomputation, check where the path has moved.