Overtime Cost Calculator
Enter a typical week of overtime to see the premium you are paying above straight time — and what the same hours would cost on part-time cover instead.
What an overtime hour really costs
Overtime is the only labour cost that arrives at a price you never agreed to. You set base rates when you hire and you set the schedule on a Thursday afternoon. Then service runs long, someone calls in sick on a Saturday, a delivery lands late and the prep list has to be finished after close. The same hour of work that cost you one figure on Tuesday costs half as much again on Sunday, and because it all lands inside a single payroll total, most operators never see it as a separate number at all.
The figure worth measuring is not what you paid for overtime hours. It is the premium — the difference between what those hours cost and what the identical hours would have cost at straight time. That premium is additional expense with no additional output attached to it. The work would have been done either way. You simply paid more to have the same person do it later in the week.
Take a kitchen paying an average base rate of $18 an hour that runs 40 overtime hours across the team in a typical week, at a 1.5× multiplier, with an 18% payroll burden for employer taxes and insurance. The premium works out at $10.62 per overtime hour, $424.80 a week, and $22,089.60 a year. Not one dollar of that bought an extra hour of coverage. It bought the right to schedule the coverage badly.
Payroll burden matters more here than people expect. Employer taxes and insurance are almost always charged as a percentage of gross wages, so they scale with the overtime rate rather than the base rate. A 1.5× hour does not simply cost fifty per cent more in wages — it costs fifty per cent more in burden on top of that. Leaving burden out of the calculation understates the premium by whatever your burden rate happens to be, which for most operations is not a rounding error.
Read the premium, not the total
It is tempting to look at the full overtime cost line and conclude that eliminating overtime would save all of it. It would not. Most of that money pays for hours you genuinely needed worked, and if the overtime disappeared you would still have to pay someone to work them. The recoverable amount is the premium alone, which is typically a third of the headline overtime figure at a 1.5× multiplier.
That distinction changes what you do about it. If your premium is $22,000 a year, the question is not "how do I stop paying for these hours" but "how do I get these hours covered at straight time instead". Those are completely different projects. The first one cuts service. The second one is a scheduling exercise.
The premium also belongs in your wider labour picture. Once you know the annual figure, run it through the labour cost percentage calculator and the prime cost calculator to see how many points of margin it is actually consuming. A premium that looks small in isolation often turns out to be a meaningful share of the gap between your current labour percentage and your target.
Overtime versus hiring part-time cover
The standard alternative to persistent overtime is adding a part-time person to cover the hours at straight time. Whether that is cheaper depends almost entirely on the rate you would have to pay them, and the arithmetic is unforgiving in one direction: an overtime hour at 1.5× a $18 base is $27, so any new hire at less than $27 an hour covers those hours for less money.
| Option | Weekly cost for 40 hours | Annual cost |
|---|---|---|
| Overtime at $27/hr (1.5 × $18) + 18% burden | $1,274.40 | $66,268.80 |
| Part-time cover at $16/hr + 18% burden | $755.20 | $39,270.40 |
| Difference | $519.20 | $26,998.40 |
The break-even point is simply the overtime rate itself. Any cover rate below it wins on wages, any rate above it loses, and the gap between the two is what you have available to spend on recruiting, onboarding and the inefficiency of a new person for their first month.
That comparison is not the whole answer, and treating it as one is how operators end up hiring a person they cannot keep busy. A new hire brings costs the wage rate does not show: recruitment time, training hours that produce nothing sellable, uniform and onboarding admin, and in many cases a minimum shift length that forces you to schedule four hours where you only needed two. There is also the risk that the person leaves within a quarter, at which point you have paid the full cost of hiring twice — run the numbers through the employee turnover cost calculator before assuming a hire is the cheap option.
The honest rule is that overtime wins when the need is genuinely intermittent and unpredictable, and hiring wins when the same hours appear in the same places on the schedule week after week. Persistent, predictable overtime is a headcount gap that you have chosen to finance at a premium.
How much overtime is normal
| Overtime as % of total scheduled hours | What it usually indicates |
|---|---|
| Under 2% | Normal absorption of no-shows and late finishes. Leave it alone. |
| 2–5% | A scheduling pattern rather than an accident. Worth an hour with the rota. |
| 5–8% | Specific shifts are understaffed, or the schedule is built with no slack at all. |
| Over 8% | Structural. A headcount gap being financed at a premium every single week. |
Treat these as orientation rather than targets. A site with heavy seasonal swings and a small core team will run higher than a steady-trade operation with a deep casual pool, and comparing the two tells you nothing. What matters is whether your own figure is drifting upward, and whether the overtime lands in the same shifts every week or scatters randomly across the schedule. Random overtime is absorption. Repeating overtime is a rota you have not fixed.
Where the hours actually come from
Overtime very rarely originates in a single dramatic event. It accumulates from a handful of ordinary causes, and each one has a different fix:
- Schedules written with no slack. If every person is rostered to their full contracted hours before the week starts, any absence at all has to be covered at a premium.
- Absence and no-shows. The most common trigger, and the one operators feel is out of their control. It is not — it responds to holding a small pool of trained cover.
- Late finishes that nobody planned for. Fifteen minutes past the scheduled end, five nights a week, across four people, is five hours of overtime that never appeared on any rota.
- Skill concentration. When only two people can run the grill or close the bar, those two people work every gap. Cross-training is a labour cost control, not just a development exercise.
- Prep volumes that do not match forecast. Kitchens that prep to habit rather than to a forecast routinely run long on quiet weeks and short on busy ones.
- Managers approving their own hours. Where nobody has to authorise overtime in advance, it stops being a decision and becomes a default.
The useful exercise is to tag every overtime hour for two weeks with one of those six causes. The distribution is almost always lopsided, and the largest bar tells you where to spend your effort. Guessing at the cause and fixing the wrong one is how overtime reduction programmes quietly fail.
Cutting it without losing coverage
- Build the schedule against a forecast, not against last week. Rostering to a demand forecast rather than to habit removes the single largest source of avoidable premium hours.
- Hold deliberate slack. Leaving a few contracted hours unrostered at the start of the week gives you somewhere to absorb an absence at straight time instead of at 1.5×.
- Require advance approval. Overtime that has to be authorised before it is worked falls immediately, purely because someone has to think about it first.
- Stagger start and finish times. Most late finishes come from everyone starting and ending together. Offsetting shifts by thirty minutes covers the close without a premium.
- Cross-train against your bottlenecks. Every additional person who can cover a critical station is one fewer reason to call the same person back in.
- Watch output, not just hours. Use sales per labour hour to check whether the overtime is producing revenue or simply extending a shift that was already unproductive.
- Review the trend monthly, not annually. Overtime creeps back the moment attention moves elsewhere, and a monthly number keeps it visible.
One caution worth stating plainly: cutting overtime by cutting hours is not a saving, it is a service reduction with a payroll benefit attached. If you pull the premium out by sending people home early and the food goes out slower, you have converted a labour cost problem into a guest experience problem, and the second one costs considerably more to repair. The objective is to buy the same hours at a lower rate, not to buy fewer hours.
Frequently asked questions
Do I enter the base rate or the overtime rate?
The base rate. The calculator applies the multiplier for you, which is what allows it to separate the premium from the straight-time portion. Entering the overtime rate would inflate the result by the multiplier a second time.
What multiplier should I use?
Enter whatever your payroll actually applies, since it varies by jurisdiction, contract and day of the week. A 1.5× rate is the most common default, but some operations pay a higher multiplier on public holidays or beyond a second hours threshold. If your week mixes rates, run the calculator separately for each block of hours.
Should payroll burden be included?
Yes. Employer taxes and insurance are usually a percentage of gross wages, so they rise with the overtime rate as well. Excluding burden understates the premium by exactly your burden percentage, which for most operations is enough to change the conclusion about whether hiring is cheaper.
Is hiring always cheaper than paying overtime?
No. Hiring wins when the hours are predictable and repeat weekly, because you are buying them at straight time instead of at a premium. Overtime wins when the need is genuinely intermittent, because a new hire carries recruitment, training and minimum-shift costs that a few unpredictable hours will never cover.
Why does the annual figure assume 52 weeks when our overtime is seasonal?
It assumes the week you entered is typical. If your overtime concentrates in a season, run the calculator twice — once with a peak week and once with a quiet one — and weight the two annual figures by how many weeks of each you actually have. A peak week annualised across the year will overstate the cost substantially.
Where does overtime show up in prime cost?
In the labour half. Overtime raises total labour cost without raising sales, so it pushes both your labour percentage and your prime cost percentage up at the same time. It is one of the more common explanations for a labour percentage that drifts upward while headcount and menu prices stay flat.
Does salaried manager overtime belong in this calculation?
Only if you actually pay a premium for it. Where managers are salaried and unpaid for extra hours, the cost is real but it does not appear in payroll — it surfaces later as burnout and turnover, which is a different and usually larger number.