What Negative Float Actually Tells Me

Negative float shows up in an update and the reflex is to treat it as a headline — “we’re eight days negative.” On its own, that tells me almost nothing. What matters is why it exists, and negative float has more than one cause.

What it actually means. Negative float is the network saying it cannot meet a constrained date on the current logic and durations. It is arithmetic, not entitlement — it does not, by itself, establish that anyone owes anyone time.

So the first question is: what is the constraint? Very often the negative float traces straight back to a hard constraint — a mandatory finish, a deadline imposed on the network — rather than a real driving path. Remove or interrogate the constraint and the “delay” can change shape entirely. Negative float against an artificial date is an artefact, not a fact about the job.

If the constraint is real, then it is a genuine signal. A real completion obligation with a driving path that no longer reaches it is exactly the early warning controls exists to surface — while there is still time to mitigate, re-sequence or escalate.

What it does not tell me. It does not tell me whose risk the delay is. It does not tell me the critical path is correct — a network held together by constraints can produce negative float that means nothing. And it does not tell me how much time to grant; that is a causation-and-criticality question, assessed event by event, not read off a float value.

So when I see negative float I do three things: find the constraint driving it, test whether the path is real logic or an artefact, and only then ask what it means for entitlement and for the decision in front of us.

Negative float is a symptom. Reporting the symptom is easy. The value is in the diagnosis.