Every guide, in one list

Fixed, rotating, split and on-call patterns each solve a different problem. Picking one before you know which problem you have is how rotas go wrong. A general hospitality industry overview helps place the operational guides in their wider sector context.

The rota should be a translation of expected demand into people and hours. Most are a translation of last week's rota.

Splits move labour cost off the payroll line and onto the employee's day. Whether that trade works depends almost entirely on geography.

Most operational failures in a hotel happen at a shift boundary. The fix is a structured handover, not a more conscientious team.

Every rota contains hours people would rather not work. How those hours are distributed does more for retention than most benefits.

Someone will call in sick two hours before a shift. Deciding what happens next in advance turns a crisis into a procedure.

Below a certain headcount, scheduling stops being an optimisation problem and becomes a constraint-satisfaction problem with almost no slack.

Attendance, task duration and productive capacity are three different things. Most properties collect the first and make decisions as though they had the third.

Paper, tablet, mobile and biometric each fail in a characteristic way. Choosing well means picking the failure you can live with.

Resistance is usually a reasonable response to an unanswered question. The question is what the data will be used for.

The technical capability to monitor continuously has outrun the operational reason to. Proportionality is the line worth holding.

When the workplace is a boat, a trail or a minibus, most of the assumptions built into time tracking software stop holding.

Most payroll discrepancies come from a handful of recurring mistakes, and almost all of them are process problems rather than honesty problems.

The housekeeping standard time is the number every rota depends on. It is also the number most properties have inherited rather than established.

Desk work is not spread across the day. It piles into two windows, and staffing to the daily average guarantees failure in both.

Between the checkout deadline and the first arrival there is a fixed window in which every room has to be turned. Everything about housekeeping scheduling follows from its width.

A hotel restaurant with a captive, arriving-together clientele has a demand curve unlike a high-street site, and staffing it like one produces a bad evening.

Linen is the quiet governor on how many rooms a property can turn in a day, and it is rarely represented in the rota at all.

Planned maintenance competes with occupancy for the same hours. Scheduling it deliberately is the difference between quiet upkeep and a drill at nine in the morning.

Seasonal operations compress a year of hiring, training, peak delivery and wind-down into a cycle that repeats before anyone has recovered from it.

On an island, the supply timetable is not an input to planning. It is a hard constraint that determines what can be scheduled at all.

A programme built on the assumption that scheduled activities run is a programme that fails on exactly the days guests most need it.

The weeks either side of peak are the only time a seasonal business can change anything. Most of them are spent recovering instead.

Distance from a labour market, a supply chain and a spare part changes every scheduling assumption borrowed from urban hospitality.

A spreadsheet is a perfectly respectable scheduling tool up to a point. Knowing where that point sits saves both a premature subscription and a late one.

Most well-known time trackers were built for agencies and remote teams. Understanding what they were designed for explains where they fit a hotel and where they do not.

Scheduling tools differ far more in how they handle change than in how they handle creation. Change is the part that happens every week.

The occupancy data that should drive the rota already exists in the booking system. Getting it to the schedule is usually a reporting problem, not an integration project.

A fortnight of testing a scheduling tool in the quietest month proves nothing. Trials should be designed around the conditions that break things.