Your labour cost percentage in 2026-27: what a healthy Sydney venue roster actually looks like

Two rosters. Same 200 hours, same crew, same $28,000 in weekly sales. One lands at 29.7% labour cost, the other at 32.7%. The only thing that changed is which days the hours sat on, and across a year that gap is about $43,600.

That's the bit the benchmark articles skip. They hand you a target percentage and stop, as if the number were something you set rather than something your roster produces. Under the 2026-27 Hospitality Award rates, the shape of your roster does more work than the size of it.

The short version: the commonly quoted 28 to 32% target isn't wrong, but it isn't sourced either. The ATO publishes real labour-to-turnover ranges built from lodged tax returns, and they shift with your turnover. For a casual under the Hospitality Award (MA000009), a Saturday hour costs exactly 20% more than a weekday day hour, a Sunday hour 40% more, and a public holiday hour double. Everything below follows from that. General information only, not financial or legal advice.

What the ATO says a healthy venue actually looks like

The 28 to 32% figure gets recycled across every POS and bookkeeping blog with nothing behind it. There's a better number, and it's free. The ATO publishes small business benchmarks built from actual tax returns, including labour as a share of turnover, split by venue type and size.

The band climbs with turnover for restaurants and cafes, because growth mostly buys you more people, not more machines. Big pubs run the other way, which is what happens when a slice of revenue comes from gaming and accommodation rather than table service. The catch, and it's a big one: the ATO defines labour as salary, wage and contractor payments, and it excludes what the owner pays themselves. If you're doing 50 hours a week on the floor for nothing, your number looks healthy for a reason that has nothing to do with your roster.

An hour is not an hour

Here's a casual food and beverage attendant grade 2 (Level 2) on the Hospitality Award, at the rates that took effect from the first full pay period on or after 1 July 2026. The ordinary hourly rate is $27.08, and the casual percentages below already include the 25% loading.

The evening loadings are the odd ones out: a flat $2.95 an hour from 7pm to midnight and $4.42 after midnight, not a percentage, so they bite hardest on your lowest-paid staff. Everything else is a clean multiple that holds at every classification level. For permanent staff the weekend gaps are wider again: Saturday costs 25% more than a weekday, Sunday 50% more, a public holiday 125% more.

A Sunday hour has to earn 40% more than a Tuesday lunch hour just to hold the same labour percentage. Most rosters are built as though it doesn't.

Two rosters, 200 hours each, three points apart

Take a 70-seat bistro in Newtown turning over $28,000 a week excluding GST, with 200 rostered floor hours and everyone on casual Level 2. Roster A spreads the trade across the week. Roster B is weekend-led, the way most Sydney venues actually run. Identical hours, identical people, identical sales.

Three percentage points. $838.94 a week. About $43,600 a year, and nobody worked an extra minute. For context on how much room sits underneath that, the ATO's own data has restaurants turning over more than $2 million reporting total expenses at 88% to 94% of turnover.

Now the honest part. In real life the weekend-led venue takes more money on the weekend, which is exactly why the hours are there. The question was never whether to roster Sundays, it's whether Sunday earns its premium. Roster B needs about $30,800 in sales, roughly 10% more than Roster A, to land back on 29.7%. If your Sunday isn't 10% busier per hour than your Tuesday, the roster is telling you something your P&L won't for another month.

The order to cut in, before you cut hours

Cutting hours is the first thing most owners reach for and it should be close to last, because every hour you cut also cuts service, and service is the top line. There's a cheaper sequence.

1. Fix the finish times before you touch the start times. The last half hour of a shift is usually the least productive and, on a Sunday close, the dearest minutes in the building. Thirty minutes off four Sunday closes is $94.78 a week, about $4,900 a year, and no customer notices.
2. Kill the rounding. Eight minutes rounded up per shift is pure leakage rather than a service cut, and the Ghost Hour maths runs to five figures for a ten-person team.
3. Audit your classifications both ways. A Level 3 doing Level 2 work costs 89 cents an hour more than it needs to before penalties multiply it. A Level 2 doing genuine Level 3 work is an underpayment waiting to be found. Get the levels right and both problems disappear.
4. Move everything that isn't service off the weekend. Stocktake, ordering, deep clean, prep, admin, writing next week's roster. Each of those hours done on a Sunday costs 40% more than the identical hour on a Wednesday, and no customer is paying for any of them.
5. Convert the hours you'd roster in any weather. A part-time Level 2 Sunday hour is $40.62 against $47.39 for a casual, so $6.77 an hour cheaper. The catch is real: you take on annual leave with its 17.5% loading, personal leave and paid public holidays, and you lose the ability to flex down. Convert the base, keep casuals for the hours that move with trade.
6. Only then cut hours, starting where sales per labour dollar is worst. That's almost never the peak. It's the open, the changeover between lunch and dinner, and the hour after the kitchen closes.

One more honest note. If your roster is genuinely tight and you're still sitting at 36%, the problem is the top line, not the labour line. A venue doing $22,000 a week on a roster built for $28,000 has a sales problem wearing a labour costume, and no amount of trimming Sunday closes will fix it.

Where Shiftly fits

Two things wreck a labour percentage: building a roster without seeing what each hour costs as you place it, and paying a premium every time a hole opens up. A spreadsheet does neither, which is why free roster templates get expensive.

Shiftly is free workforce management for Australian venues: rostering, plus award-aware timesheets that help you estimate what a shift should cost before you commit to it (a calculation tool, not payroll, and not legal advice). The part that matters here is what's wired in behind it. It's the roster that fills itself, with an on-demand network of local hospitality workers built into the same roster you already run, so a dropped Saturday shift doesn't turn into overtime at 150% or a night spent running short. Get started with Shiftly.

Frequently asked questions

What is a good labour cost percentage for a restaurant in Australia?

There isn't one number, and anyone quoting a single figure is guessing. The ATO's benchmarks, built from lodged tax returns, put labour at 18% to 30% of turnover for restaurants under $500,000, 23% to 32% between $500,000 and $2 million, and 27% to 34% above that. Pubs and coffee shops sit on their own ranges. Use the band that matches your turnover, and remember the ATO figure excludes the owner's own labour, so add a market wage for your hours before you compare.

How do I calculate my labour cost percentage properly?

Total labour cost divided by sales excluding GST, times 100. Labour cost means wages, the 12% super guarantee, workers compensation, and if your Australian wages top $1.2 million a year, NSW payroll tax at 5.45% on the excess. The most common mistake is leaving GST in the sales figure, which flatters the number by roughly a tenth: on the worked example above, that's the difference between reporting 27.0% and the true 29.7%.

Why is my Sydney venue running higher than the benchmark?

Usually roster shape, then rent. A weekend-led trading pattern in Newtown or Bondi puts a bigger share of your hours on Saturday and Sunday rates, which cost casuals 20% and 40% more than a weekday day hour. Rent does the rest: the ATO has restaurants over $2 million reporting rent at 6% to 9% of turnover nationally, and a Sydney lease rarely sits at the bottom of that. Neither is a reason to accept 40%, but both explain why 28% is harder here than in a regional town.

Did award rates go up again on 1 July 2026?

Yes. Modern award minimum wages rose 4.75% from the first full pay period on or after 1 July 2026, and the National Minimum Wage went to $26.44 an hour. The Hospitality Award's two lowest classifications rose about 6% because of a structural change the Fair Work Commission started this year, which we covered in why hospitality's lowest rates went up 6%, not 4.75%. Check the Fair Work Ombudsman pay guide for MA000009 for your exact rates.