Blog

Comparing labor cost between two hotels you own

· portfolio, labor cost, reporting

Put two of your own properties side by side and one of them will look lazy. At the invented pair below — 48 rooms and 44 rooms, twenty-five miles apart, same pay scale, same owner — one pays 1.00 hours per occupied room-night and the other 1.28. On the smaller property’s volume that is a 58-hour-a-week gap, about $54,000 a year at an $18 loaded rate. Every one of those 58 hours is accounted for by three facts about the two buildings, and none of the three is anybody’s pace. The useful version of this comparison is not the ranking; it is the subtraction you do before the ranking means anything.

Comparing labor cost between two hotels, side by side

Ridgeline Inn and Cedar Court are invented, and identical everywhere I am not trying to prove a point: same wages, same 16-hour front desk, same maintenance, breakfast and cart-stocking hours. One week:

Ridgeline InnCedar Court
Rooms4844
Occupancy72%66%
Occupied room-nights242203
Average length of stay2.4 nights1.4 nights
Departures / stayovers101 / 141145 / 58
LaundryOff-site contractIn-house
Hours paid241260
Hours per occupied room-night1.001.28

Hours per occupied room-night, not hours or dollars, because raw totals rank the biggest property as the worst one — settled ground, and what the per-room metrics measure if the vocabulary is new.

So: Cedar Court is 44 rooms, sells 39 fewer room-nights than Ridgeline, and pays 19 more hours. Bring it to Ridgeline’s ratio — 0.996 before rounding, not a flat 1.00 — and it would pay 202 hours instead of 260. Fifty-eight hours a week, $1,044, and the number that gets somebody in a car.

Where all 58 hours actually went

What differsHours of the gap
Laundry done in-house instead of contracted out18
More checkouts per room-night sold14
The same fixed coverage spread over fewer room-nights26
Total58

Those three lines are not a sample of the causes. They sum to the whole 58 hours, because I built the two properties to differ in exactly three ways and held everything else identical. A real pair will not be that tidy, but the order of the exercise is the same.

Laundry: 18 hours. Cedar washes its own linen and Ridgeline sends it out. Those 18 hours are labor at one property and an invoice at the other, and in a labor-hours comparison they read as pure inefficiency. In-house versus outsourced laundry has a defensible answer either way; what it is not is a staffing problem. Subtract the departments only one property runs before anything else.

Checkout mix: 14 hours. A checkout clean and a stayover refresh are different jobs, and the checkout runs about twice as long. Say both properties have measured their own and got 28 minutes and 14. Ridgeline, at a 2.4-night average stay, turns 101 rooms and refreshes 141 — 80 hours of cleaning. Cedar, at 1.4 nights, turns 145 and refreshes 58 — 81 hours of cleaning to service 39 fewer room-nights. Per room-night sold that is 0.40 hours against 0.33, and across Cedar’s 203 nights it is 14 hours of the gap. Length of stay is a demand fact, not a housekeeping one: a two-night average stay is a structural discount no amount of supervision buys at a one-night property.

Fixed coverage: 26 hours. The desk is open 112 hours a week at both properties whether 44 rooms or 22 are occupied, and the same goes for maintenance (20 hours), breakfast (21) and cart-stocking (8) — 161 hours apiece, unmoved by occupancy. Divided by Ridgeline’s 242 room-nights that block is 0.67 hours per room-night; divided by Cedar’s 203 it is 0.79. Same hours, same work, 26 hours of apparent gap. Four fewer rooms and six points of occupancy, not a single scheduling decision. Hold Cedar at Ridgeline’s 72% and the line halves to 13 hours, so about half of it is just the smaller building and will not move at any occupancy. This is the one that catches people out, because it moves in the direction that feels like a verdict: the quieter property always looks worse per room, and the quieter property is the one you were already worried about.

If you want yesterday’s hours at each property on that property’s own clock rather than assembled by hand on a Monday, that is what ProfitClock does for a portfolio.

The adjustment nobody makes: hours that are not on the payroll

Suppose the owner lives at Ridgeline and covers 14 desk hours a week himself, unpaid. Ridgeline’s comparable figure is not 1.00, it is 1.05 — 255 hours over 242 room-nights, 1.054 unrounded — and the gap you were chasing was never 58 hours. It was 46.

This correction is the one most likely to be missing and the hardest to reconstruct afterwards, because unpaid hours are the only category of labor that leaves no record anywhere. It also runs opposite to instinct: the property that looks most efficient is often the one where somebody is working for free, and that figure has its own uses besides this one. Any property where an owner or a family member covers shifts needs those hours written down before it goes in a table with a property where nobody does.

What would actually show a pace difference

At these two properties, nothing does: both were given the same measured 28 and 14 minutes, so the residual after all three subtractions is zero. Say that plainly, because it is also where a pace difference hides — not as a leftover at the end, but as a larger cleaning line in the middle.

So compare like work: checkout cleans against checkout cleans, never blended minutes against blended minutes. Suppose Cedar’s attendants genuinely take 31 minutes on a checkout where Ridgeline’s take 28. That is 145 checkouts × 3 minutes, 7.25 hours a week. Cedar’s cleaning would be 88 hours rather than 81, its total 267, its ratio 1.32, and the gap 65 hours rather than 58 — of which the cleaning line is 21 instead of 14. The seven hours of pace sits inside that line. It is about a ninth of the 65-hour headline the same assumption produces, and it is the only ninth anybody at either property can do anything about.

Doing this at all requires that housekeeping hours are recorded per room rather than per shift. A payroll export cannot tell you which minutes were a checkout and which were a refresh; it has one number per person per day, and that number is the average this whole post is arguing against.

Where this comparison stays unreliable

Three things do not come out in the wash.

  1. Building shape. Three floors and one service lift is not the same job as a single-storey exterior corridor, and the difference is minutes of walking and lift-waiting that nobody logs. It is usually smaller than it feels — the saving from consolidating floors at an invented 66-room motel runs to about 12 minutes a shift — but it is not zero, and it appears in none of the three lines.
  2. Quality. Two properties can hit the same minutes per checkout and hand back rooms of different standards. Hours measure effort, never outcome; an inspection pass rate is a separate number, collected separately.
  3. Occupancy itself. The fixed-coverage line means the ranking moves when occupancy does: compare the same two properties in a week where Cedar runs at 80% and much of the gap closes with nobody having done anything differently. One week is not a measurement; four, subtractions applied each time, is the start of one.

Which gives the order to work in, and it is deliberately boring: take out the departments only one property runs, then divide the fixed coverage block by each property’s own room-nights, then compare checkouts to checkouts. Whatever survives all three is worth the drive. At these two invented properties nothing does — the honest answer, and not a disappointing one: a $54,000 headline sent somebody hunting a staffing problem that was a laundry contract, a length of stay and four fewer rooms. A real pace gap is the size of the seven hours above. It is never the size of the number that started the search.

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