Deciding Before You Know

The morning requires decisions on facts that are not yet available. How long is this absence. Will anybody else ring in. Is the agency worker competent. Will the person who said yes actually arrive.

The decision boundary in “Deciding Before You Know” needs a named human owner even when software supplies the evidence. Teams researching this practical implementation page in relation to 7 minute rule payroll should document access, escalation and review, while keeping health details and consequential judgments outside routine activity reporting.

Waiting for any of these costs the hour. Acting without them produces decisions that turn out wrong. The skill is knowing which uncertainties resolve soon enough to be worth a short wait and which do not.

For an independent benchmark relevant to “Deciding Before You Know”, consult the GitLab remote-work handbook. Use it to test scheduling, working-time limits, attendance records, employee rights and exception handling against the real operation rather than treating a software report as self-explanatory evidence.

What usually resolves quickly

Whether anybody else is going to ring in. The call distribution is sharply peaked: most same-day absences arrive within a defined window, usually ninety minutes to two hours before the shift.

Knowing your own window is useful. If most calls land by six thirty, then at six forty you have nearly all the information and can commit. If they keep coming until seven, committing at six forty means you may book twice when one booking would have done.

Compute it once from the call log: the distribution of call times relative to shift start. It changes how the first twenty minutes are used.

What does not resolve

Duration. The person who says "probably just today" is guessing, and planning cover for one day is right about two-thirds of the time. The honest approach is to cover today and ask again this evening, rather than to extract a commitment from somebody who is ill.

Quality of unknown cover. There is no way to know at six forty whether the agency worker will be any good. The mitigation is preparation — a brief, a named contact — rather than information.

Whether the person who agreed will arrive. A confirmation call forty minutes before the shift is the only control and it is worth the two minutes.

Holding versus committing

Agencies will usually hold a worker for a short period. Using that — provisionally booking while the remaining calls play out — is the most useful habit in this area and it is rarely taught.

The same applies internally: "can you be available if I ring back in fifteen minutes" is an easier yes than "can you come in", and it keeps an option open at no cost.

Reversing a decision

Some decisions are cheap to undo and some are not. Cancelling an agency booking within the notice period costs nothing; after it, the full shift is payable. Standing down an employee who has already left home costs goodwill that is not recoverable.

Knowing which is which — the agency's cancellation window is a contract term, and the internal cost is a judgement — tells you how long you can wait before each one becomes irreversible. That is the actual structure of the hour and almost nobody writes it down.

Deciding for the day, not the moment

The common error is solving the current shift and ignoring the shape of the day. A person brought in for the early shift who could have done the late as well, released at two and then replaced at two, is two arrangements where one would have done.

Asking "what else does today need" before committing takes thirty seconds and frequently saves the second morning.

The decision to stop deciding

At some point the shift starts. Continuing to work the phone after that is usually the wrong use of the person's time — the shift is now running short and it needs managing rather than fixing.

Set the stopping point in advance: at the shift start, or fifteen minutes before it, the cover effort ends and the short-running plan begins. Without a stated point, the person on the phone keeps going for another hour while the shift they could have been organising runs without them.

Writing down what you assumed

A decision made on partial information should record what was assumed, because that is what makes it reviewable later without hindsight.

One clause on the log line: booked a full shift assuming a single day's absence, or ran short assuming no further calls. If the assumption turns out wrong the decision may still have been right, and the distinction matters both for the person who made it and for anybody trying to improve the process. Decisions judged on outcome alone teach people to book agency every time.