A protocol for last-minute cover
Someone will call in sick two hours before a shift. Deciding what happens next in advance turns a crisis into a procedure. Before absence or conduct becomes a disciplinary question, managers should distinguish a failed process from a serious offence; this discussion outlines the wider issue.
A small hospitality team will lose a shift to illness, transport failure or emergency several times a month. This is not an anomaly to be minimised; it is a normal operating condition, and it is worth planning for with the same seriousness as a fire drill. A cover protocol is a small-scale form of contingency planning.
What usually happens instead is improvisation. The manager finds out at six-thirty, starts phoning people in an order determined by who they can remember, and the first person who answers gets the shift regardless of whether they worked late the night before.
The call list, decided in advance
Write the order down before you need it. A defensible order for most properties runs like this:
- Anyone rostered on-call for that day, if you operate on-call.
- Staff on a scheduled day off who have registered interest in extra hours — maintain this as a standing list and refresh it monthly.
- Part-time staff not scheduled that day, in rotation rather than by convenience.
- Staff from an adjacent department who are cross-trained for the role.
- Extend an existing shift, only where it does not breach the rest interval.
- Run short, with a written decision about what will not be done.
The order matters less than its existence. Having it written removes the appearance of favouritism from a decision made under time pressure, and it stops the same two reliable people absorbing every gap.
Name the last option out loud
Running short is a legitimate response and it is the one most managers refuse to consider, which is why it happens by accident instead. The team ends up covering a missing person by absorbing the work invisibly, quality drops in ways nobody logs, and the business learns nothing.
Decide in advance what gets dropped when the department is one person down. Deep cleaning of public areas. Turndown service. The evening bar being open for the last hour. Whatever it is, write it as a documented reduced-service mode, communicate it, and record when it was used.
A written list of what stops when you are one person short. It converts an invisible quality decline into a visible, countable event — and a countable event can be used to justify the headcount that would prevent it.
Cross-training is the only real fix
Every other measure manages the symptom. The underlying problem is that in a small team most roles have exactly one qualified person available on any given day, so the loss of that person is unrecoverable.
Cross-training is slow, it costs productive hours, and it is the first thing cut when the season starts. It is also the only intervention that changes the shape of the problem rather than the response to it. A property where three people can run the desk competently and two can cover a breakfast service is structurally different from one where each of those is a single point of failure.
The practical version is modest: one cross-training pairing per month, during the quietest shift of the quietest week, with a named person responsible for it. That is roughly twelve additional capabilities a year, which is enough to change how a shoulder season feels.
Record every gap
Log each cover incident: date, role, cause, who covered, how long it took to fill, and whether anything was dropped. Individually these are noise. Over a season they are the most persuasive document you will have.
A pattern of Sunday morning gaps points at a rota problem, not a reliability problem. A cluster of absences in one department six weeks into peak season points at a workload problem. Neither is visible without the log, and both are actionable with it.