PTO accrual for hourly hotel staff who work unequal weeks
· payroll, operations
A housekeeper who works 24 hours one week and 46 the next earns a different amount of paid time off depending on which of two ordinary methods your handbook uses — and at 30 hours a week the gap between them is about 20 hours of leave a year, per person. Neither method is wrong. What goes wrong is running one of them in the handbook and the other in the spreadsheet, and never noticing because the two only agree for somebody who works exactly forty hours every week. Almost nobody at a small hotel does.
Two methods, and they are not close
Take a placeholder policy of 80 hours of PTO a year for a full-time person — substitute your own figure; the shape of the answer does not change. There are two common ways to hand it out.
Flat accrual by pay period. Divide the annual grant by the number of pay periods. On a biweekly payroll that is 80 ÷ 26 = 3.08 hours every period, for everybody, whatever they worked. PTO is earned by being on the payroll. Plenty of handbooks prorate the flat grant by contracted hours instead, which is a third option and sits between the two below — if yours does, the part-timer row in the table further down will not look like your numbers.
Accrual per hour worked. Divide the annual grant by a full-time year. Using 2,080 hours (40 × 52), that is 80 ÷ 2,080 = 0.0385 hours of PTO per hour worked — or, more usefully for mental arithmetic, one hour of PTO per 26 hours worked. PTO is earned by working.
Both are standard. The first is easier to administer; the second ties the benefit to hours actually given. That choice is a policy decision rather than an arithmetic one, and it deserves to be a decision somebody made rather than whatever the spreadsheet inherited.
One invented housekeeper, four unequal weeks
One invented property throughout: a 62-room exterior-corridor motel, biweekly payroll, $16 an hour as a placeholder. Nothing below is drawn from a real property or a real payroll.
| Week | Hours worked | Why |
|---|---|---|
| 1 | 24 | Three days, low midweek occupancy |
| 2 | 38 | Normal week |
| 3 | 46 | Covered a two-day call-out |
| 4 | 32 | Four days |
| Month | 140 |
Two biweekly periods: 62 hours, then 78 hours.
- Flat: 3.08 + 3.08 = 6.16 hours accrued.
- Per hour worked: 140 ÷ 26 = 5.38 hours accrued.
Three-quarters of an hour apart over a month — small enough to ignore, which is why it survives. Annualise it and it stops being small.
What the gap is worth over a year
Because the per-hour rate here is one PTO hour per 26 worked and a year is 52 weeks, this placeholder policy works out to a tidy rule: your PTO hours for the year are roughly twice your average weekly hours. Thirty hours a week earns sixty hours of leave. Flat accrual earns eighty regardless. The factor is the annual grant divided by 40, so recompute it if your policy is not 80 hours — on a 40-hour grant the multiplier is one, not two.
| Average hours a week | PTO earned, per hour worked | PTO earned, flat | Gap |
|---|---|---|---|
| 20 | 40.0 | 80.0 | 40.0 hours |
| 30 | 60.0 | 80.0 | 20.0 hours |
| 35 | 70.0 | 80.0 | 10.0 hours |
| 40 | 80.0 | 80.0 | none |
A part-timer at 20 hours accrues, under flat accrual, four weeks of their own schedule in paid leave. Across six part-time staff averaging 30 hours — a plausible crew at this size — flat accrual grants about 120 more hours of leave a year, roughly $1,900 at the placeholder rate.
That is not an argument for either method. A deliberately generous flat policy at a property that struggles to keep housekeepers may be retention spending worth making — but price it against your own turnover rather than against an assumption. Worked through on an invented 48-room motel at a $16.50 placeholder rate, one housekeeping departure runs about $1,130, most of it vacancy overtime and the replacement’s first six weeks, and that figure is a floor. On those inputs, a year of the more generous accrual costs less than two departures. The argument is only that you should be able to say which number you are paying.
PTO accrual for hourly employees runs on hours worked — so where does that number come from
Pick the per-hour method and every accrual now multiplies a count of hours worked. There are only two places that count can come from, and they do not agree.
The schedule is what you planned. It is wrong by every early start, late finish, covered call-out and unrecorded swap. In the month above, week 3 was scheduled as 40 and worked as 46 — accrue on the schedule and that person is short 6 ÷ 26 = 0.23 hours of leave. A year of covering call-outs can easily be 100 unscheduled hours, nearly four hours of PTO never granted to the most willing employee on the property. It runs the other way too: a scheduled 40 that came in at 34 over-grants.
The clock is what happened. It is the right base, on one condition: it has to be settled before anyone multiplies it.
ProfitClock does not accrue PTO and holds no leave balance — that lives in your payroll system and your handbook, and we would not pretend otherwise. What it produces is the number the accrual runs on: hours from punches rather than from a schedule, approved by a manager before they are used for anything, punch edits kept as an append-only history so a correction is visible instead of silent, and every hour filed on the property’s own calendar day rather than a server’s. If your accrual base is currently a schedule, a fortnight of real punches will tell you how far apart the two numbers are — start a free trial.
Three places the base is quietly wrong
1. PTO hours inside the base. If the column you multiply is total paid hours rather than hours worked, paid leave accrues paid leave. Somebody who takes the whole 80-hour bank accrues an extra 80 ÷ 26 = 3.08 hours they did not work for, and it compounds year over year. Nobody decides this; it arrives when a payroll report’s “total hours” column is reused as an accrual base.
2. Overtime hours in the base. Week 3 was 46 hours worked. Feed all 46 in and the people covering call-outs accrue fastest, which is arguably correct. Cap the base at 40 and they do not, which is also a defensible policy choice. What is not defensible is not knowing which your sheet does — recomputing one person’s last period by hand answers it in about four minutes.
3. A punch corrected after the period closed. The accrual ran Monday on 38 hours; on Wednesday somebody fixed a missed clock-out and the week became 41. The leave bank and the timesheet now disagree permanently, because nothing recomputes. One a month is a rounding error; a habit of them is a bank drifting away from the record it was derived from, and the drift only ever runs one way.
A fourth looks like a date problem and is really a money problem: which calendar day a punch lands on decides which pay period it falls in, so an overnight shift, or a system filing hours in the wrong timezone, can move hours — and the accrual they generate — into the wrong period entirely. Where that boundary falls is its own decision worth making on purpose.
The question this post will not answer
Three things an owner reasonably wants to know, and this is not the page for any of them: whether PTO hours count toward the threshold at which overtime starts; whether unused PTO must be paid out when somebody leaves; and whether your state sets a minimum accrual rate, a carryover floor or an accrual cap.
All three are set by federal and state rules, they differ by jurisdiction, and they change. A wage-and-hour rule is worth printing only if it has been read at its source and confirmed current, so we are declining rather than guessing — and you should not take any of the three off a blog post, this one included. Put the question to your payroll provider or an employment attorney in writing, and keep the answer with the policy. A state or local paid-sick-leave mandate, where one applies, is a second bank with its own rules.
Where this doesn’t apply
- Salaried exempt staff. Flat per-period accrual is the norm, and the unequal-weeks problem does not arise because the weeks are not recorded unequally.
- No PTO policy at all, still common at this size. Nothing above applies, and whether you should have one is a retention question rather than a payroll one.
- Unlimited PTO. No rate, no bank, no accrual — and a different set of problems, mostly about coverage.
- A crew that genuinely works forty every week. The two methods agree exactly, so pick whichever is easier to run.
Everywhere else the two methods disagree by an amount nobody has priced, on a base nobody has checked. The arithmetic takes an evening; the policy decision it supports lasts years.
Start seeing what labor actually costs
Set up your property, put staff on the clock, and read tomorrow morning’s labor number.
14-day free trial · No credit card · Cancel anytime