How to Handle Last-Minute Sick Calls: A Manager's Playbook

The 5:47am Problem

It is 5:47 on a Tuesday. Twelve people on today, and the text you just read says one of them cannot make it. The shift starts at 7:15.

What happens next, in most Australian workplaces, is a phone-around. You scroll contacts, ring the three people most likely to say yes, leave two voicemails, and wait. Then you ring three more. Somewhere in there you are also supposed to be opening up, doing handover, or driving.

The thing that makes this painful is not the sick call. Staff get sick, and a workplace that resents that is a workplace with a bigger problem. The painful part is that the phone-around is serial. You are asking people one at a time, and each no costs you several minutes you do not have. Ten workers reached sequentially is a fundamentally slower process than ten workers reached at once, and no amount of being organised fixes that, because the constraint is the method rather than the manager.

So the fix is not a better contact list. It is changing the process from serial to parallel.

The Process, Step by Step

This describes how the process should work in any rostering tool that supports unassigned shifts and staff notifications. Capability varies between platforms, so check what yours actually does before you rely on it. If you are on a spreadsheet, this section doubles as the business case.

Mark the worker absent first. Before anything else. This takes seconds and it prevents the quiet downstream problem where an absent worker stays on the roster and flows through to payroll as though they attended. Do it while you are still reading the text.

Convert the shift to an open shift. Date, time, site and role carry over from the original shift, so there is nothing to re-enter. The shift now exists as a gap the system can advertise.

Reach everyone eligible at once. This is the step that does the actual work, and it is the capability worth testing hardest when comparing tools, because implementations differ enormously. The goal is that every worker who is available for that window and not already rostered learns about the shift at the same moment, rather than one after another.

This is also where keeping availability data current pays off. If the system knows who genuinely cannot work Tuesday mornings, the notification only reaches people who might say yes, which keeps the alert meaningful rather than something staff learn to ignore.

Allocate once, then close it off. Whoever takes it goes on the roster and everyone else needs to know it is gone. Done well this avoids the phone-around's most awkward moment, ringing back the person who said yes second.

Close the loop. Check the replacement actually knows the specifics, particularly if the site or client has anything unusual about it. The system confirms a booking; it does not brief anyone.

How long all this takes depends entirely on your pool, the time of day and how attractive the shift is. A Tuesday mid-morning with twelve available casuals fills quickly. A 6am Sunday in a regional town may not fill at all. Anyone quoting you an average fill time is guessing.

When Nobody Takes It

Some shifts do not fill on the first pass. Work down this ladder rather than jumping straight to the bottom of it.

Step 3 has a trap worth naming. Splitting a seven-hour shift between two casuals does not cost you seven hours of wages, because each of those short shifts may attract a minimum engagement under the relevant award. Sometimes the split is still the right call for coverage. Just know you are paying for it. Our guide to minimum shift lengths in Australia sets out what applies where.

If you are reaching step 5 regularly, the sick call is not your problem. Your casual pool is too small relative to your permanent roster, and no notification system fixes an empty bench. That is a staffing model question, and the data on how often you escalate is the argument you need to make the case.

Reducing How Often It Happens

Here is the uncomfortable part. A proportion of last-minute sick calls are genuine illness and nothing will change them. Another proportion are people avoiding a shift they resent. Managers tend to assume the split is more favourable than it is, and the second group is the one you can actually do something about.

Roster fairly, and be seen to. If the same three people always get the closing Friday and the Sunday, and the roster never rotates, you will see it in the absence pattern eventually. The fix is unglamorous: rotate the shifts nobody wants and make the rotation visible, so people can see it is a system rather than a preference. Schedules built around how people actually live hold up better than schedules built purely around coverage.

Give people a legitimate exit. This is the highest-leverage change most managers can make. If a worker has a conflict and no sanctioned way to resolve it, calling in sick is the path of least resistance, and it is worse for you than a swap would have been, because it arrives with no notice. A visible, pre-approved shift swap process converts a 5:47am text into a Thursday afternoon swap request.

Apply the policy the same way every time. Whatever your evidence threshold is, it has to be the same for everyone. Inconsistency here does more damage than a lenient policy, because it signals the rule is really about who is asking.

Look at patterns, carefully. Repeated absences attached to the same day, or always adjoining a rostered day off, are worth a private and genuinely curious conversation. Approach it as finding out whether something is going on, not as an accusation. Sometimes it is a caring responsibility, a transport problem or a second job, all of which have solutions that a disciplinary process does not reach.

One legal point to be careful about. Personal leave is an entitlement under the National Employment Standards, and section 107 of the Fair Work Act 2009 lets you require reasonable evidence. It does not entitle you to a diagnosis, and treating someone adversely for taking leave they are entitled to can engage the general protections provisions. Manage the pattern, not the illness. Broader obligations are covered in Fair Work rostering requirements.

Systems help with the scramble. They do not fix a roster people are trying to get out of, and it is worth being honest with yourself about which problem you have before buying software to solve it. If the pattern is concentrated on particular shifts rather than particular people, the roster is telling you something. Filling gaps faster is covered in more depth in open shift management.