Why Deadlines Are Set Before Anyone Understands The Work

The delivery date was fixed at a point when nobody had examined the problem. This is not carelessness. The date was needed for something other than predicting the finish.

The date arrives from somewhere else

Commitments are made in budget rounds, contracts, regulatory timetables and public announcements, all of which require a date long before the work is scoped.

Those processes cannot proceed without one. A plan with no date cannot be funded, compared against alternatives, or approved by anybody.

So a number is produced, and it is produced by the people who need it rather than by the people who will meet it.

Estimating requires doing part of the work

Reliable estimation depends on understanding the problem, and understanding the problem is most of the effort in anything genuinely uncertain.

Early estimates are therefore drawn from analogy with previous work, which is reasonable when the work resembles what came before and useless when it does not.

Uncertainty at the outset is enormous and narrows as work proceeds, which means the most confident-sounding date is produced at the moment of least knowledge.

The first number becomes the anchor

Once a figure has been said aloud in a meeting, every subsequent discussion is a negotiation away from it, and moving far from an anchor requires justification.

Later estimates produced with far better information are received as slippage against the original rather than as improved forecasts.

The person delivering a revised date pays a reputational cost for accuracy, which is a strong reason to keep quiet and hope.

Padding sets off an arms race

Teams learn to add contingency, managers learn that estimates contain contingency, and the estimate is trimmed on the assumption that it was padded.

Teams then pad more heavily to survive the trimming, and the numbers exchanged stop carrying information about the work at all.

What remains is a ritual in which both sides adjust a figure according to what they expect the other to do with it.

The date is doing coordination work

A fixed date lets other groups plan around it, marketing prepare, suppliers commit and finance schedule, and none of that can happen against a range.

Understood that way, the deadline is a synchronisation device rather than a prediction, and it is doing that job even when it is wrong.

The damage comes from treating it as both at once, so that a coordination tool is defended as a forecast, and the work is compressed to protect the announcement.