How to compare rostered labour costs with actual labour costs

Written by Roster Flow. Published 26 July 2026.

Most owners look at the wage bill after payroll runs and come away with a feeling about whether it was high. A feeling is hard to act on. Comparing the cost of the roster you published with the cost of the hours that were actually paid turns it into a number, and then into a short list of shifts that explain the number. Once the sheet is set up, it takes about twenty minutes a week.

This is a comparison of dollars, not only hours. The roster vs actual hours calculator shows where the hours differ. The question here is what those differences cost, and what to change so they cost less next week.

Why hours alone do not explain the cost

Two weeks can have exactly the same hours and different wage bills. What an hour costs depends on who worked it, when, and under which rules, so the gap between rostered and actual cost always has two parts:

• More or fewer hours than the roster planned: shifts that ran over, extra people called in, shifts that finished early or did not happen.
• Dearer or cheaper hours: sick cover by someone on a higher rate, work moving from a weekday to a weekend, or time the award pays at overtime rates.

The overtime part catches many businesses out. Under the Restaurant Industry Award 2020 and the Hospitality Industry (General) Award 2020, a part-time employee must be paid at overtime rates for time worked beyond their rostered hours (clause 10.13 in both). Under the General Retail Industry Award 2020, a part-time employee is paid the overtime rate for time worked beyond their guaranteed hours (clause 10.8). So a part-time employee who stays half an hour past the roster can cost noticeably more than half an hour at their usual rate.

Casuals and weekends change the cost too. Under the Restaurant Industry Award, casual employees are paid a 25% loading on the minimum hourly rate (clause 11.1), and the award sets penalty rates for work at particular times and on particular days (clause 24). Rules differ between awards, so confirm yours with the Fair Work Ombudsman’s Find my award tool and check current rates on pages such as the Restaurant Industry Award pay rates.

Decide what counts as labour cost, and keep it the same

Both numbers have to be built the same way, or the comparison measures your method instead of your week. Write down what goes in and keep to it every week.

Superannuation is the on-cost most small businesses add. The Australian Taxation Office sets the super guarantee at 12% for the year from 1 July 2026, calculated on each eligible employee’s qualifying earnings. If you include it, take the actual figure from payroll and check the ATO’s super guarantee page for what it is calculated on, rather than adding 12% to every dollar. The glossary entry for labour cost lists the other on-costs, such as payroll tax where it applies and workers compensation insurance.

Step 1: freeze the roster you compare against

Rosters change during the week. If you compare actual cost with the roster as it stands on Sunday night, every change made along the way has already been absorbed and the comparison shows very little. Save a copy of the roster at the moment it was published, with its estimated cost, and compare against that.

Keep a short note of each change made after publishing and why. A sick call covered by someone on a higher rate is a different problem from a shift that ran over, and the note saves working it out from memory a week later. If you do not have an estimate yet, the roster cost calculator costs a week by person and by day.

Step 2: pull the actual hours and cost

Actual cost comes from approved timesheets and the pay run, not from the roster with a few edits. For each person you need approved hours by day and gross pay for the week, split into ordinary pay, overtime, penalty rates and allowances if payroll can show it. Use the same week boundaries as the roster. If you pay fortnightly, compare the fortnight against both weeks of roster together.

Approved timesheets are also records the law requires. The Fair Work Ombudsman says time and wages records must be kept for 7 years, must not be changed unless the change corrects an error, and must include the hours worked by casual and irregular part-time employees who are paid for time worked. That is the best reason never to edit a timesheet to match the roster: the timesheet records what happened. Fair Work’s record-keeping page lists everything that has to be kept.

Step 3: put the two side by side

Lay the week out by day first. The daily view shows where the money went, and the shifts behind each day show why. Here is a week at a bakery cafe that trades seven days. Its roster estimate uses the business’s own average cost per hour for each day, which includes super: $36 on weekdays, $45 on Saturday and $50 on Sunday. They are this example’s figures, not award rates.

The week came in $372 over, which is 3.8% of the rostered cost, on 5 more hours than planned. Five hours at an average of about $40 would be around $200, so hours alone do not explain it. The next step shows where the rest came from.

Step 4: split each day into hours and rates

For each day, work out what the extra or missing hours would have cost at that day’s rostered cost per hour. Whatever is left is the rate effect: the same hours costing more or less than planned.

hours effect = (actual hours - rostered hours) × rostered cost per hour for that day

rate effect = actual cost - rostered cost - hours effect

Wednesday matched the roster, so it has no row. Of the $372, $198 came from extra hours and $174 from dearer hours. Nearly half the overspend happened with no extra hours at all, which a comparison of hours would never have shown. Any leave paid to the person who called in sick on Sunday sits on the leave line, not in these figures.

Step 5: decide what each difference needs

Most differences fall into a handful of patterns, and each has a different fix. The table lists the common ones.

In the example, the fix with the most value is Sunday’s sick cover. If the supervisor was the only person free who could run the floor, that is a skills gap on the team, not a rostering slip. Forecasting staffing from weekly sales deals with the busy day overruns on Friday and Saturday.

Set a threshold so you only chase what matters

Not every dollar of difference deserves a conversation. Pick a threshold and look only at what crosses it, such as any day more than 5% over its rostered cost, or any person more than 30 minutes over across the week. These are settings you choose for your business, not rules. In the example, Thursday is 7.8% over, Friday 6.6%, Sunday 6.0%, Saturday 4.5% and Monday 1.7%, so a 5% threshold picks out Thursday, Friday and Sunday.

Watch the trend as well as the week. One Saturday rush is trade. The same extra hour every Saturday for a month is a roster that is wrong, and it belongs in next week’s roster rather than in another conversation.

Make the comparison a weekly habit

1. Save the published rosterKeep a copy with its estimated cost the moment it goes out to the team.
2. Approve the timesheetsFix missing clock outs and decide every overrun before payroll, with a reason recorded for each change.
3. Cost the actual weekTake hours and gross pay from the pay run, built the same way as the roster estimate.
4. Compare by day, then by shiftSplit each day’s difference into hours and rates and write down what happened.
5. Pick one fixChoose the largest cause that is likely to repeat and change next week’s roster, swap rule or approval for it.
6. Record the percentageNote the week’s difference as a percentage of rostered cost so the trend is visible over a few months.

If you run more than one venue, compare each one separately before adding them up, because a good week at one venue can hide a poor week at another. Tracking labour cost across multiple venues covers that view, and the rostered versus worked hours calculator shows the hours side for each person.

When the week went to plan and still cost too much

Sometimes the comparison shows almost no difference and the wage bill is still too high. That is a planning problem, not a discipline problem: the roster asked for more than the week needed or could afford. Compare the rostered cost with sales using the labour cost percentage calculator, and check the roster against the wage budget before it is published rather than after payroll.

Roster Flow keeps both sides of this comparison together: availability and leave from staff phones, the roster published to everybody at once, clock ins that become timesheets, and approved hours sent to Xero, MYOB or a CSV file.

A checklist for the weekly comparison

• The roster is saved, with its estimated cost, when it is published
• Both sides include the same items, with super in or out on both
• Actual hours come from approved timesheets, never edited to match the roster
• Each day’s difference is split into hours and rates
• Part-time hours past the roster are checked against the award’s overtime rules
• A threshold decides which days and people get a closer look
• One fix goes into next week’s roster

Sources

• Fair Work Ombudsman: record-keeping
• Australian Taxation Office: super guarantee
• Fair Work Commission: Restaurant Industry Award 2020 [MA000119]
• Fair Work Commission: Hospitality Industry (General) Award 2020 [MA000009]
• Fair Work Commission: General Retail Industry Award 2020 [MA000004]
• Fair Work Ombudsman: find my award

Back to the blog