Three numbers describe how a hotel is performing. Two of them can be improved by damaging the business, and understanding why is most of revenue management.
The definitions, stated precisely
Occupancy = room-nights sold ÷ room-nights available, over a period.
Note room-nights, not rooms. A twenty-room hotel over a thirty-day month has 600 room-nights available. Occupancy is not "how many rooms are full right now" — that is a live snapshot and tells you nothing about the month.
ADR (Average Daily Rate) = room revenue ÷ room-nights sold.
Room revenue only. Not food and beverage, not spa, not the conference room. And divided by room-nights sold, not by number of bookings — averaging per booking weights a one-night stay the same as a fourteen-night one, which is a different and much less useful number.
RevPAR (Revenue per Available Room) = room revenue ÷ room-nights available.
Equivalently, ADR × occupancy. The difference from ADR is the denominator, and that difference is the entire point: RevPAR counts your empty rooms against you.
Why RevPAR is the one that matters
Each of the other two can be improved by doing something stupid.
Push occupancy by cutting rates and you can fill the hotel while earning less. Occupancy rises, the business gets worse.
Push ADR by only accepting expensive bookings and you can post a superb rate on a half-empty hotel. ADR rises, the business gets worse.
RevPAR moves only when you are genuinely selling more room-nights, or the same room-nights at a better rate. It cannot be improved by trading one against the other, which is exactly why it is the number owners and lenders benchmark on.
Cancelled bookings are not revenue
An error worth stating on its own, because it is common and it corrupts every number above it: a cancelled or no-show booking must not count toward room revenue or room-nights sold.
If it does, your ADR is wrong, your occupancy is wrong, your RevPAR is wrong, and — worse — the same hotel shows different revenue on different screens depending on which report happened to filter correctly. Count them separately, as counts, because the cancellation rate is itself worth watching.
What to do with the numbers
Compare like with like. This month against the same month last year, not against last month. Hospitality is seasonal, and month-on-month movement is mostly the calendar.
Segment before you conclude. Overall RevPAR hides everything. Split it by day of week, by channel, by rate type, and by room type. A hotel with a strong overall number can be losing money every Sunday, or on one room type that is priced wrong, and the aggregate will never show it.
Watch the mix, not just the total. Rising RevPAR driven entirely by a high-commission channel is worth less than the same number driven by direct bookings, because the net is different. Track net RevPAR after distribution cost if you can.
Look at pace, not just performance. Bookings on the books for a future date compared to the same point last year tells you what is about to happen, while RevPAR tells you what already did. Pace is the number you can still act on.
The other one worth knowing
TRevPAR (Total Revenue per Available Room) uses total revenue — rooms, food and beverage, spa, everything — over available room-nights.
It matters when a meaningful share of your income is not the room. A resort or a hotel with a genuine restaurant business can have unremarkable RevPAR and excellent TRevPAR, and managing purely on RevPAR would push you to reject exactly the guests who spend most on site.
GOPPAR (Gross Operating Profit per Available Room) goes one further and accounts for cost. It is the honest number and the hardest to produce, because it requires allocated costs.
A practical starting routine
If you are not measuring any of this:
- Get RevPAR by month for the last two years. One line. It will immediately show you your real seasons rather than the ones you assume.
- Split the current month by day of week. Almost every hotel finds one or two systematically weak nights.
- Split by channel, net of commission. Frequently changes which channel you thought was your best.
- Look at pace for the next ninety days against the same point last year.
Those four views, produced monthly, are more useful than any amount of daily rate-fiddling — and they are what tell you whether last month's decisions actually worked.