Skip to content
Log in
← Glossary

float

Also called: slack, total float

How long a task can slip without moving the project's end date. Tasks on the critical path have zero float; float is shared along a chain, so using it early leaves none for later tasks.

Float (also called slack or total float) is how many working days a task can slip without moving the project's end date. Tasks with zero float form the critical path: every day they lose, the whole project loses. Tasks off the critical path have room, and float says exactly how much.

You get it from two passes over the schedule. The forward pass gives each task its earliest start and finish; the backward pass, starting from the end date, gives its latest start and finish. Then float = latest start − earliest start (the same as latest finish − earliest finish). Float is calculated for each task, but it depends on every other task in the network: change one duration or one link and the float of other tasks changes too, so recompute it after every re-plan.

Two things make float easy to misread. First, it is shared along a chain: if two tasks in a row each show 10 days, the pair has 10 days between them, not 20, and whoever uses them first leaves nothing for the next task. Second, float is measured against the schedule's own end. When a fixed date comes earlier than that end, the plan is late before anything slips, and no amount of float on side tasks fixes it.

In practice float tells you where to look. Critical tasks get attention every week; the others get a check on how much of their float is left. Float is room for surprises, not permission to start late: a task that spends its float "because it can" becomes critical before anything has gone wrong.

Example

Eight tasks of Klasna's physics launch, in working days from Monday 2 November 2026 (day 0). The forward pass gives an earliest finish of day 51, not the 82 days of all durations added up. The backward pass from day 51 gives the latest starts:

Task
Duration
Needs first
Earliest start
Latest start
Float
A Curriculum units drafted
22
none
0
0
0
B Curriculum review
10
A
22
22
0
C Tutor sourcing and screening
18
none
0
14
14
D Demo lessons
8
B, C
32
32
0
E Tutor onboarding
6
D
40
40
0
F Landing page
8
none
0
33
33
G Booking by subject
5
F
8
41
33
H Launch email and ads
5
E, G
46
46
0

Critical path: A, B, D, E, H = 22 + 10 + 8 + 6 + 5 = 51 days. Screening (C) can start as late as day 14, because demo lessons cannot begin before the review ends on day 32. The landing page and booking show 33 days each, but it is one 33-day float: if the landing page takes 20 days longer, booking has 13 left.

With all 17 tasks the picture changes. The diagnostic test chain becomes critical, the schedule ends on day 70 (Friday 5 February) against a limit of day 65 (Friday 29 January), and the tutor chain has 17 days of float. Those 17 days don't help the 5-day gap: the fix has to come from the critical path. A second reviewer plus a question bank written in parallel moves the end to day 62, three days before the limit.

Common mistakes

  • Adding up the float of tasks in a chain. Two tasks in a row with 33 days each have 33 days together, not 66.
  • Starting late "because there is float". Float spent on purpose is gone when the real surprise arrives.
  • Reading float only against the schedule's own end. If a fixed date comes earlier, the plan is late before anything slips.
  • Speeding up a task that has float. Days saved off the critical path move nothing; put the effort on the critical path.
  • Keeping old float numbers after a change. Float depends on the whole schedule; recompute it after every re-plan.

Learn it in the course

  • The critical path and float · A forward pass and a backward pass by hand on eight tasks of the physics launch: the earliest finish, the chain with no room, and how much the other tasks can slip.
  • Case: the schedule to 1 February · Seventeen tasks, one fixed date. Find the critical path and the float, see that the plan lands a week late, compare cutting, parallel work and an extra reviewer, and write Leo the schedule decision.
  • Effort is not duration: when adding people helps · Why ten days of review don't become two with five reviewers: the parts that can't be split, the cost of coordinating, and the waiting no extra person removes.