Telling Somebody Before They Cross

Almost every organisation discovers a working time breach in arrears: a report is run, a figure is over, and the period in question is already closed. At that point the only available actions are a record, an explanation and perhaps a payment.

The workflow in “Telling Somebody Before They Cross” becomes more reliable when scheduled hours, actual time and later corrections can be distinguished. For teams exploring employee monitoring software, employee monitoring software can add operational time and project context, provided data collection is proportionate, permissions are limited and every important exception receives human review.

The same information, produced a fortnight earlier, is an ordinary rostering constraint. Nobody is in breach, somebody works a different shift, and nothing has to be explained to anybody. The entire difference is when the report runs.

For an independent reference relevant to “Telling Somebody Before They Cross”, consult the ILO working-time resources. Use it to test working-time definitions, recordkeeping, access, retention and exception handling against the organisation’s real process rather than treating one software report as conclusive.

The threshold to alert at

Not the limit. A system that flags people at the ceiling is flagging them too late to do anything.

Pick a figure with enough room to act in: an average within two hours of the limit, or a projected average that crosses it within three weeks at the rostered pattern. Those are arbitrary numbers and they are better than no number, because they convert a legal boundary into an operational one that a rota planner can work with.

Who should be told

The person rostering, first, because they are the only one who can change anything. The individual, second, because the hours are theirs and because they may know things the system does not — a second job, a planned absence, a reason the rostered pattern will not happen.

The line manager, if different from the rostering person. And nobody else: a list of people approaching a limit is not a performance report and should not circulate as one.

What the conversation is

Short and practical. Your average is at forty-six against a limit of forty-eight; at the pattern you are rostered for, you cross it in two weeks; here is what we propose to change.

What it is not: a request that the person work less of their own accord, or a discussion of whether they want the overtime. Most people near a limit want the hours, which is exactly why the constraint has to be owned by the organisation rather than delegated to the individual's judgement.

The right to decline, said out loud

In most regimes a worker may decline hours that would take them over a limit, and in many they may decline overtime generally unless the contract provides otherwise. Whether people believe that in practice is a different question.

The test is not what the handbook says. It is whether anybody has declined in the last year and what happened to them. If the answer is that nobody has, the right is theoretical, and a compliance arrangement resting on people exercising a right nobody exercises is not an arrangement.

Where the figures should live

In the rostering tool, next to the person's name, at the moment the shift is being assigned. Not in a monthly report that arrives after the rota is published.

This is usually possible and rarely configured, because the limit data and the rostering data sit in different systems with different owners. Joining them is a piece of integration work with a very clear payback: it moves every one of these decisions from after the fact to before it.

The pattern behind the individual

Somebody who is repeatedly near a limit is not a scheduling anomaly. They are a vacancy, a skill that only two people hold, or a department running below establishment.

The per-person alert handles this week. The count of people who hit the alert, by department, month over month, is the thing to put in front of whoever owns the headcount, and it is a far stronger argument than the usual one, because it is a constraint rather than a preference.

What to do when it is already crossed

Record it accurately, including the cause. Give whatever rest or compensation the regime requires, within the window it requires. Do not amend the hours to make the figure fit, which is the one response that turns a compliance problem into a records problem and then into something worse.

Then ask why it was not seen. In almost every case the answer is that nobody was looking at the right number at the right time, and that is a fixable thing rather than a reason to blame a rota clerk.

The shift that was already promised

The awkward case is somebody rostered weeks ago for hours that now take them over, who has arranged their life around the earnings.

Changing it late is a cost to them whatever the reason. Where the organisation's own monitoring failed to catch it in time, the decent course is to cover the shift and pay it, rather than to remove the hours and let the person absorb a planning error. That is a policy decision worth taking in advance, because taken on the day it will go the other way.