Projects

The longest path isn’t the whole story

Mohit Saini · 17 Jun 2025 · 3 min read

The longest path isn’t the whole story

Every project management textbook teaches that the longest path of dependent tasks determines a project’s duration. Every practising manager knows that delays on that path delay the project. So it seems logical to focus on the longest path above all else.

Managers do exactly that. In planning, they look for ways to move tasks off the longest path onto side paths, so the plan shortens. In execution, when there is a choice, they favour the longest path, even if side paths slip a little.

Yet when delayed projects are analysed, a different picture appears. The longest path keeps moving. Tasks that had plenty of slack in the original plan suddenly become critical. Small, low-effort items, non-critical drawings, minor resources: all of them hold up projects surprisingly often.

Why? Because in organisations where the same people work across many projects, “longest path first, no matter what” has a hidden flaw.

A simple example

Imagine a project with three tasks, A, B and C, each done by a different specialist. Call them the designer, the analyst and the builder.

Looking at the dependencies, the planner notices that B doesn’t need all of A to start. Only the last part of A is needed before C. So A is split:

  • A1 must be done before B can start.
  • A2 must be done before C can start, but not before B.

The longest path now runs A1 → B → C, and A2 sits on a side path with slack. The plan is shorter. Everyone is pleased.

Now add more projects

But the designer doesn’t work on just this project. She has similar tasks waiting on several others.

The split creates a window: after finishing A1, she has some time before A2 is needed. If she has nothing else to do, she might simply do A2 early. That is fine.

But suppose she uses the window to do A1 for Project 2, which is on Project 2’s longest path, so that seems like the right priority. Project 1’s window shrinks.

If she then does A1 for Projects 3 and 4 as well, each on its own longest path, Project 1’s window runs out. Its A2 is now late. The side path has become the longest path, and Project 1 is delayed.

Multiply this across many people and many projects, and you get the familiar picture: side paths turn critical one after another, people drop everything to rescue them, priorities shift daily, and nearly every project is late.

Why detailed schedules and simple rules don’t fix it

Detailed schedules with exact start dates for every task seem like the answer. But task durations are uncertain, and the dates start clashing within days.

“Always do the longest path first” creates exactly the cascade described above.

The fix: limit how many projects each person has open

The rule that resolves this is a work-in-progress limit for each critical resource group.

In our example, the designer might be limited to two open projects at a time. Within those two, she works on longest-path tasks, and then returns in time to finish their side-path tasks. Only when a project leaves her active list does a new one enter.

This means that, at times, she will work on a side-path task of an active project while longest-path tasks of other, waiting projects sit untouched. That feels wrong. But it is what keeps the longest path from shifting and keeps her active projects on track.

In a multi-project world, the longest-path rule must give way to the work-in-progress rule.

Without a limit, focusing on the longest path of many projects at once eventually makes every path critical.

Related reading

Mohit Saini

Strategy, supply chain & sales transformation consultant. About · Book a discussion