Hotel labor cost by department: hours in the wrong bucket
· labor cost, labor planning, small hotels
Your department labor numbers are not measurements. They are attributions: anyone who crosses a department line during the week has their hours filed wherever the schedule put them, and at a small hotel several people cross one every week. The weekly total is right. The buckets are not. At the invented 58-room motel below, 33 of 268 payroll hours — an eighth of the payroll week — sat in the wrong department, and the two decisions those numbers produced were both wrong: a $196-a-week saving that was really a $72-a-week increase, and a second maintenance hire against a backlog that was not a staffing problem.
Why hotel labor cost by department is attribution, not measurement
A schedule files hours by the label on the shift. A time clock records a person and a span. Neither knows the housekeeping supervisor spent Tuesday afternoon at the front desk covering a call-out, or that the maintenance tech ran the morning wash because 19 rooms departed at once. So a report saying housekeeping cost $2,300 last week is really saying the people we call housekeeping were paid $2,300 — and only one of those is a labor metric.
Splitting by department is still the right instinct — a blended whole-property number hides where the drift is, which is why it is worth computing the per-occupied-room figure per department rather than once for the building. The inputs just need one correction first, and it costs about ten minutes a week.
The week three people moved 33 hours
An invented property: 58 rooms, exterior corridor, interstate exit, 62% occupancy that week, eight hourly staff, 268 paid hours, placeholder rates. Three people crossed a department line, which is what cross-trained staff are for.
| Person | Rate | Scheduled as | What the week actually was |
|---|---|---|---|
| Maintenance tech | $20 | 40 h maintenance | 24 h maintenance, 10 h laundry, 6 h desk |
| Housekeeping supervisor | $17 | 40 h housekeeping | 28 h housekeeping, 8 h desk, 4 h laundry |
| Evening desk clerk | $16 | 32 h front desk | 27 h desk, 5 h making up vacant-dirty rooms |
Everyone else worked where the schedule said. That is 33 hours — most of one person’s week — in the wrong bucket:
| Department | As scheduled | As worked | Change |
|---|---|---|---|
| Housekeeping | 136 h | 129 h | −5% |
| Front desk | 72 h | 81 h | +13% |
| Maintenance | 40 h | 24 h | −40% |
| Laundry | 20 h | 34 h | +70% |
| Total | 268 h | 268 h | — |
The smaller the department, the worse the error
The part worth keeping is not obvious: misattributed hours barely move a big bucket and wreck a small one. The same 33 hours moved housekeeping 5% and laundry 70%.
Housekeeping’s distortion is almost cosmetic. At 252 occupied room-nights, 136 hours is 32.4 minutes per occupied room and 129 is 30.7 — a 1.7-minute gap, inside the noise of a single week. It points the wrong way, so housekeeping looks slightly worse than it is, but nobody makes a bad decision over 1.7 minutes.
Laundry is a different story, because laundry here is other people’s spare hours and nothing else. On the schedule it is one part-timer at 20 hours and $14: $280 a week. As worked it is that part-timer plus 10 hours of a $20 tech and 4 of a $17 supervisor: $548 a week. The bucket is not 70% bigger, it is 96% more expensive, because the hours that moved in carry the two highest rates on the property. A department staffed out of other people’s slack is always costed low.
The two decisions it changed
The saving that was not one. Last quarter the laundry attendant’s hours were cut from 34 a week to 20. The laundry line fell from $476 to $280 and the schedule recorded a $196-a-week saving. The 14 hours did not go anywhere: they are the tech’s 10 and the supervisor’s 4 in the table above, and at $20 and $17 they cost $268. Laundry labor rose by $72 a week — about $3,700 a year — while the report showed a cut, because the hours left the bucket somebody was watching for two nobody was. Whether laundry belongs in-house at all is a separate question with more than labor in it, and that break-even deserves its own arithmetic — but it cannot be run against $280 when the real figure is $548.
The second maintenance tech. “Forty hours of maintenance a week and still a two-month punch list” is a staffing argument. Twenty-four hours and the same punch list is a scheduling problem with a known cause: 16 of those hours are committed to laundry and the desk. The fix is to decide whether the tech is a maintenance tech who covers, or a utility position doing three jobs — then staff the gap that choice leaves. Hiring a second $20 tech against the misread is a $41,600-a-year commitment before employer taxes, against a problem 16 reclaimed hours would have covered.
If you want the room-cleaning half of this measured rather than estimated, start a free trial and let a fortnight of real cleans accumulate before you split anything.
Three conventions, and what each one costs
There is no correct answer here, only three defensible ones.
- Attribute by the work record. Hours follow what the person did, because the work carries its own timestamps. Most accurate, and limited to work that gets recorded — a room clean does, “looked at the ice machine” does not.
- Attribute by shift block. When somebody’s assignment changes mid-shift, the shift becomes two blocks and the supervisor splits it when they approve the week. It catches the hours that moved as blocks — in the week above, the tech’s 16 and the supervisor’s 12, so 28 of the 33 — and misses the five the desk clerk picked up an hour at a time. Thirty seconds per change, and the only version most 40-to-80-room properties will actually sustain.
- Do not split at all. One labor number for the property: hours and cost per occupied room. You lose the ability to see which department drifted, and you stop publishing four numbers you know are wrong — at three or four employees, the honest trade.
The rule that matters more than the choice: pick one, write it down, keep it for a year. Consistent attribution error still shows a trend, and the trend is the only part of this that changes a schedule. Switching mid-year destroys the comparison — the same reason a daily labor report is read for its three moving numbers rather than its totals.
What you can split without buying a second time clock
One split comes without extra tooling. When room cleans are worked as tasks, each clean carries its own start and finish time and a photo at completion, so housekeeping minutes are measured rather than inferred from the schedule. The rest of the punch is the remainder, and a convention you choose — but with the biggest department measured rather than guessed, the exercise changes character.
Where this gets genuinely hard
- Some hours honestly belong to two departments. The tech who answered the phone twice while changing a ballast was doing both jobs, and no convention resolves that. Round it the same way every time: a consistent rounding is a rounding, an inconsistent one is a number nobody can read.
- Two pay rates for one person is a payroll question, not a reporting one. If you pay a different rate for desk work than for maintenance, how overtime is computed changes — there is a federal regulation covering employees who work at two or more rates in a week (29 C.F.R. § 778.115), it has exceptions, and state rules add their own. Settle it with your payroll provider or your state labor department before you set a second rate, not from a blog post.
- Attribution changes nobody’s pay. Overtime is computed on each person’s total hours for the week, under whatever threshold applies to you — several states add a daily one — whichever bucket the hours landed in. Moving an hour between departments is a reporting act; treating it as more than that is how a reporting habit becomes a wage problem.
- This is not a department P&L and we do not produce one. Allocating payroll to accounts belongs to your bookkeeper’s chart of accounts — a different list of things.
- Below about 25 rooms, skip it. With three employees who all do everything, departments are a fiction and the work buys nothing. Track hours per occupied room for the property instead.
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