Should we switch to remote work as soon as a warning is issued?

It is one possible answer and it is not automatic. It assumes two things verified in advance: that the work is genuinely doable remotely for the people concerned, and that power and connectivity will hold — which is precisely what a weather event calls into question. A serious protocol therefore has two levels: remote work, and a deliberate stop when conditions do not allow it.

Who should decide to stop an activity?

One named person, and only one, with a designated deputy. It is the most important point in the whole protocol and the most often left vague. When the decision has no owner, each manager decides for their own area, instructions contradict each other, and staff call three people to get three different answers.

How do you catch up a lost day?

By deciding in advance what does not get caught up. That is the useful counter-intuitive part: a team that tries to recover everything accumulates a debt that weighs on the following weeks. Naming, during preparation, which activities can slip without consequence and which absolutely have to be recovered lets the decision be made calmly rather than under pressure.

Should we avoid scheduling training during this period?

Not necessarily, but the question has to be dealt with at booking. A postponement clause agreed in advance — in which cases, with what notice, at what cost — turns an incident into a formality. The same conversation held on the morning of a warning goes badly for everyone.

Managing a team through cyclone season

A manager gets a message at twenty to seven in the morning. The warning has moved to class III. They check their phone, they call their leadership who are not answering yet, they write their team an instruction they will have to correct an hour later. In the meantime half the team has set off and the other half is waiting at home, not knowing whether their day counts.

This will happen another three or four times before April. And again next year.

A recurring constraint is not an unforeseen event

That distinction settles most of the subject. An unforeseen event is one whose occurrence could not be anticipated. A cyclone season is not unforeseen: it returns every year, over the same months, with a graded warning system everyone knows.

A recurring, predictable event belongs to preparation, not to reaction. Handling it by reaction is a choice — rarely a conscious one — and it is that choice which costs, far more than the interruption itself.

The lost day is visible, so it gets counted. What does not get counted: the hour spent reaching everybody, the contradictory instructions in circulation, the people who set off for nothing, and the conversation with the client where you announce a delay without a restart date to offer.

The one-page protocol

What is needed is not a forty-page business continuity plan. One page is enough, written in October, and it answers five questions.

Who decides. One named person, one designated deputy. It is the most important point and the most often vague. With no owner for the decision, each manager rules for their own area and the company issues three different instructions on the same morning.

By what time. A cut-off before which the decision is communicated — half past six, say — so that nobody sets off while waiting for information.

Through which channel. One only, known to everybody, tested once before the season. Work email is the worst choice: nobody reads it at six in the morning, and it assumes connectivity. A messaging group works, provided everyone is actually on it and the test has been run.

Who works remotely and who stops. By name where possible, by role otherwise. The reachability rules and rituals that implies need deciding in advance. It means having checked in advance what is doable from home — and accepting that for part of the team the answer is no.

What does not get caught up. The counter-intuitive point, and the most useful. A team that tries to recover everything accumulates a debt that weighs on the following weeks. Deciding calmly which activities can slip without consequence avoids deciding it under pressure, when the temptation is to hold everything.

What gets settled at booking

For anything scheduled — a course, a seminar, a trip, a meeting with people travelling in — the postponement clause is agreed at signature, not on the morning of the warning.

That applies to our own sessions: every course booked in this window comes with a postponement clause agreed at signature.

Three lines are enough: in which cases postponement applies, with what notice it is confirmed, and at what cost. Does a session interrupted on day one resume at the next module or from the beginning? That question has a simple answer when asked beforehand and becomes a dispute when asked afterwards.

Few organisations do this, and the argument against is always the same: it invites bad luck, or it reads as negative in a proposal. In practice a provider who raises it unprompted signals that they know the ground.

The human point

One last element, and it is not logistical. A warning does not weigh the same on everybody: it depends on whether you have children whose school closes, an elderly parent at home, an exposed house, or a route through an area that floods.

The most constrained people are also, almost always, the ones who will ask for least. They will come in, or they will manage in silence, and the manager will learn nothing.

A change of rules announced in October is received; the same one announced on the morning of a warning is not.

That is the practical reason — not the moral one — why an explicit rule beats case-by-case flexibility. Flexibility granted on request benefits those who ask. A written rule benefits everybody, including the people who will never say they needed it.

Cyclone season: decide before the warning, not during it