Staff scheduling: covering the hours without burning out the people

Staff scheduling — building the work rota, in the phrasing most of the world outside HR software actually uses — is the work of matching people to hours: enough cover when demand is high, not so much that you are paying for idle time, and distributed in a way the team can live with. Most guides treat it as an optimisation problem. In practice the arithmetic is the easy half — what makes a rota work or fail is notice, fairness and how changes get handled once it is published.

This article is about the practice. Choosing a tool for it is a separate question, covered in employee scheduling software, and the two get conflated constantly: teams buy a scheduler to fix a problem that was really an unwritten rule about who gets weekends.

What staff scheduling has to get right

• Demand by hour, not by day. A shop that is dead until eleven and frantic at five does not need even cover, and a schedule built on daily totals will be wrong twice a day.
• Minimum viable cover per period — the smallest number of people who can safely run the shift, and who among them has to hold a specific skill or certification.
• Notice. Publishing two weeks ahead is the single change that most improves how a schedule is received, and in several jurisdictions predictive-scheduling rules make it a legal minimum rather than a courtesy.
• Fairness, made explicit. Weekends, late finishes and the awkward shifts should rotate on a rule everyone can see, not on who is least likely to complain.
• Rest between shifts. A close followed by an open is legal in some places and a resignation letter in most.
• A change procedure: who may swap with whom, who approves it, and where the record of it lives.

Building the work rota for the week

1. Start from demand, not from people. Sketch required cover per hour from whatever history you have — sales, tickets, bookings, footfall.
2. Lay in the fixed constraints: opening and closing, skills that must be present, contracted hours, known absence.
3. Fill the hard shifts first — the early opens and the weekend closes. Filling the easy ones first is how the last two slots end up on whoever is quietest.
4. Check the rotation against last month. If the same person has three of four weekends, the rule is not working, whatever the rule says.
5. Publish, then freeze. Changes after publication go through the swap procedure rather than through the schedule being quietly rewritten.
6. Record what actually happened, including swaps and no-shows. Next month's schedule is only as good as this month's record.

The cheapest improvement available to most teams is not a better algorithm — it is publishing earlier. People arrange childcare, second jobs and their lives around the rota, and a schedule that lands on Friday for a week starting Monday imposes a cost that never shows up in the labour numbers but appears reliably in turnover.

The patterns worth knowing

• Fixed shifts: everyone works the same days each week. Predictable and popular; inflexible when demand moves.
• Rotating shifts: teams cycle through earlies, lates and nights. Fairer on paper, harder on sleep — the direction of rotation matters, and forward rotation is generally easier on the body than backward.
• Four on, four off: continuous cover with long blocks of rest, common in operations that never close.
• Split shifts: cover two peaks with a gap between. Efficient for the employer and widely disliked, which is a trade worth making consciously.
• On-call: cheap cover for unpredictable demand, and the pattern most likely to be abused if the compensation for being available is vague.

Where the schedule lives

Ettex Calendar holds published schedules as shared calendars people can actually see, with the coverage view and the change history in one place, and absence booked against the same weeks is covered in absence management. The demand arithmetic — hours by period, cost against budget — belongs in Ettex Sheets alongside it.

The boundary is worth stating plainly. Ettex Calendar is not a workforce management system: there is no availability collection from staff, no shift-swap marketplace, no time clock, no labour-cost forecasting against sales, and no compliance engine for break or predictive-scheduling rules. Where those rules apply, they are set by your local employment law and worth checking with someone qualified rather than inferring from software behaviour.

One vocabulary note, because it causes real confusion when searching for help: work rotas, staff rotas, shift schedules and staffing schedules all describe the same artefact. Software vendors tend to say scheduling; the people actually building it usually say rota. Nothing in the practice changes with the word.