How to manage casual staff availability across multiple events

Catering and event businesses do not staff one event at a time. A busy Saturday means a wedding at noon, a corporate dinner at seven and a festival bar all weekend — drawing on the same pool of casuals, many of whom also work elsewhere. The scarce resource is not people; it is accurate knowledge of who can actually work. And that knowledge decays fast.

Why availability data goes stale

Most teams collect availability in bursts — a message blast before the season, a form, a spreadsheet column. The data is truthful for about a week. Then a student’s timetable changes, someone picks up a regular Tuesday job, a family commitment lands — and none of it flows back to you, because updating their availability means messaging a manager, and nobody messages a manager to say something boring. The result: you book against a snapshot, and the gap between snapshot and reality surfaces as declined offers, late drop-outs and no-shows.

The structural fix is to stop owning the data yourself. Availability should be self-service: staff record it once, adjust it in seconds from their phone, and know that planners genuinely look at it when staffing events. When updating availability is easier than declining a mis-aimed offer, the data starts staying current on its own.

Standing availability: the baseline

Standing availability is each person’s normal working pattern — the days and times they are generally free, including recurring rules like “never on Mondays” or “weekends only in term time”. It is not a promise to work; it is a filter that stops you offering the Tuesday lunch to someone who has not worked a Tuesday all year. In Keruja, availability and leave management holds exactly this: profiles with day and time preferences, recurring rules with exception dates, and leave alongside.

Leave and one-off unavailability

On top of the standing pattern sit the one-offs: holidays, exams, the fortnight visiting family. Treat these as first-class records with a simple request-and-approve flow, not as chat messages — because a holiday that lives in a thread is invisible at the moment three weeks later when someone books that person for a gala. Recording leave properly also protects staff: nobody enjoys declining a shift they told you about in June.

Per-event offers and confirmations

Standing availability says could; only the person says will. However good the data, book through explicit offers that each staff member accepts or declines per event. The pairing is the point: recorded availability tells you where an offer is likely to land, and acceptances turn possibility into commitment. Everything between — chasing, re-offering to the next candidate — is mechanical work that event-first staff scheduling exists to absorb.

Overlapping commitments across events

The multi-event failure mode is booking one person twice: the noon wedding runs long, and she is also due at the seven o’clock dinner across town. Spreadsheets cannot see this because each event is its own tab. A staffing system can: when everything is booked in one place, a new assignment can be checked against existing shifts and approved leave, and the clash surfaced before it is confirmed — while it is still a planning question rather than an apology to a client.

Keeping it current: staff self-service

All of this works only if updating availability is genuinely easy for staff. One place on their phone for shifts, hours, availability and time off — the Keruja staff app is built to be that place — means the data is maintained by the people who actually know it, at the moment it changes. Make it explicit in onboarding: keeping availability current is part of the job, and in return, offers will respect it. That trade is what keeps a casual pool answering your messages in October, not just in June.