
In a recent discussion among planners on projectmanagement.com, the question was simple: is the critical path enough to keep a project under control? Almost everyone said no: you also have to watch resources, risks, decisions. All true. But the most useful line came at the end: delays usually start on paths that were not critical yesterday.
Anyone who updates plans every week recognises this at once. Everybody watches the critical path, and that is exactly why the surprise rarely comes from there. It comes from a task that had ten days of slack and lost them, update after update, without anyone noticing.
Three checks help you see near-critical tasks in time, and keep the critical path a tool for decisions, not a red label on the Gantt chart.
In short
- Logic first: if the plan has no reliable links, the critical path you see tells you nothing.
- Then the threshold: “critical” does not have to mean zero slack; a threshold tied to the update cycle turns near-critical tasks into a list to review.
- Finally the trend: slack that shrinks at every update matters more than slack that is small but steady.
1. Does the logic hold?
The critical path is a calculation: Project derives it from the links between tasks and from their durations. If the logic has holes, the calculation returns a precise and wrong result.
The most common holes are three. Tasks with no successor, which end in a void and push nothing. Hard constraints, such as “Must Start On”, which pin a date and stop delays from travelling. Dates typed by hand over the links, which do the same damage without showing it.
There is a test that takes a minute. Pick a task in the middle of the plan, stretch it by one week and watch the project finish date. If it does not move, and that task has no slack, the critical path you are looking at is decoration. Then put the duration back as it was.

If you want to go deeper, How to review a Microsoft Project plan you did not build lists the checks in order, fastest first; the article on task dependencies covers the rest.
2. Widen the definition of critical on purpose
For Project a task is critical when its total slack is zero or less. That is the right definition for the calculation, but too narrow for whoever has to look ahead: a task with two days of slack, in a plan updated every two weeks, can turn critical between one update and the next without you seeing it coming.
The threshold can be raised. In Project go to File > Options > Advanced, scroll down to the calculation options for the project and change “Tasks are critical if slack is less than or equal to”. Primavera P6 has an equivalent setting in the project settings.
Don’t pick the value by gut feeling: tie it to the update cycle. A simple rule is to look two updates ahead. If you update the plan every two weeks, two updates make four weeks, that is twenty working days. That is the threshold: every task with less than twenty days of slack goes on the list.
So “near-critical” is no longer an impression. It becomes a list to review in the meeting, and every row carries the name of whoever owns it.

A word of warning: the threshold changes the Critical field across the whole plan, and with it the red bars and the filters. If the Gantt chart goes to someone who expects the critical path in the strict sense, tell them, or keep the higher threshold only in the copy you work on.
3. Watch the trend, not the value
A task’s slack, read once, tells you where you are. Read across several updates, it tells you where you are heading.
Take two tasks. The first has thirty days of slack but loses ten at every update: three updates from now it is critical. The second has three days of slack and has been sitting there for two months. With the threshold from point 2, at the first update the second is on the list and the first is not. Yet it is the first that is about to cause your delay.

To see it, compare slack between two consecutive updates. Before updating, copy Total Slack into a custom field (a Number or a Duration), then update and put the two columns side by side. The tasks where slack drops most are where the next critical path is forming. For this to work, the plan has to be updated with a status date: otherwise the slack you compare is stale.
The three doors
Resources, risks and decisions matter, as most of the discussion said. But they almost always reach the finish date through one of these three doors: logic that does not pass delays on, a definition of critical that is too narrow, slack that runs out while nobody watches it drop.
In your plan, which of the three is open?
Frequently asked questions
What is a near-critical task?
A task with little total slack, above zero but below a threshold you choose. It is not on the critical path today, but a small delay can put it there.
Where do you change the critical path threshold in Microsoft Project?
In File > Options > Advanced, under the calculation options for the project: “Tasks are critical if slack is less than or equal to”. The default value is zero days.
How can I tell whether a task’s slack is shrinking?
Copy Total Slack into a custom field before each update and compare it with the new value. What counts is the difference, not the value.
If you want the same question, “which way is it heading?”, asked of every delay in the plan, GanttDerivata compares status dates and tells you, for each delay, whether it is widening or closing.
#MicrosoftProject #CriticalPath #Scheduling #ProjectManagement