Hotel KPIs: Occupancy, ADR, RevPAR and What Each Booking Channel Really Costs
By PlainSight — Insightful Actions · Updated October 2026 · ~5 min read
A hotel room that goes unsold tonight can never be sold again, and one sold through an online travel agency arrives with a commission attached. So a small hotel's year is decided by three things: how many rooms it fills, at what rate, and how much of each booking it keeps after the channel and the cost of turning the room. These are the numbers to read every month.
Where your numbers live
Your property management system can export reservations or the daily revenue report to CSV or Excel: one row per stay or per night with the date, the property, the stream (rooms, food and beverage, parking, fees), the channel (direct, Booking.com, Expedia, a wholesaler), the nights and the amount.
Guest names aren't needed.
The numbers that decide what each room earns
1. Occupancy, ADR and RevPAR
What good looks like: RevPAR growing, from the mix of occupancy and rate that suits your market.
Occupancy is rooms sold divided by rooms available. ADR (average daily rate) is room revenue divided by rooms sold. RevPAR (revenue per available room) is room revenue divided by rooms available, which is occupancy times ADR, and it settles the argument between filling rooms and holding rate. For context, CoStar and Tourism Economics raised their 2026 US forecast in August: RevPAR up 4.4% on the year, occupancy at 63.1%, ADR up 3.1%.
2. Channel mix, and what each channel keeps
What good looks like: each channel's share known, with its full cost per booking.
Online travel agencies commonly charge 15% to 25% of room revenue, more with their visibility programs, and by one industry estimate independent hotels took about 63% of their bookings through OTAs in 2025. Divide each channel's commission by its room revenue, then compare what a room sold there leaves you with what the same room leaves you sold direct.
3. Direct booking share
What good looks like: rising, especially among repeat guests.
Direct reservations (your website, phone, walk-ins) divided by all reservations. A guest who found you on an OTA can book direct next time if you give them a reason: a better rate for returning guests, a perk at check-in, an email after the stay.
4. Labor cost per occupied room
What good looks like: steady as occupancy moves, with schedules that follow the forecast.
Housekeeping and front-desk wages divided by rooms sold. Labor is the biggest cost in most hotels, and it doesn't fall by itself when occupancy does. Hotel wages are rising in some cities: in Los Angeles the minimum for hotels with 60 or more rooms reached $25.00 an hour on July 1, 2026, on its way to $29.00 by 2029.
5. Other revenue per occupied room, and fees
What good looks like: parking, food and other services adding to every stay, with every mandatory fee in the advertised price.
Non-room revenue divided by rooms sold. A bar, breakfast, parking or event space raises what each guest is worth. Under the FTC's rule on unfair or deceptive fees, in effect since May 12, 2025, mandatory resort or destination fees have to be included in the price you advertise, not added at checkout.
6. Booking window and cancellations
What good looks like: known by channel, so rates move at the right time.
Days between booking and arrival, and the share of reservations cancelled, channel by channel. A shrinking window or a cancellation spike on one channel tells you when to adjust rates, and a channel with generous free cancellation brings bookings that are less certain than they look.
Warning signs worth acting on
- Occupancy rising while RevPAR falls — rate given away to fill rooms.
- Repeat guests still booking through an OTA.
- Labor per occupied room rising in a slow month.
- Mandatory fees added at checkout.
- One channel's cancellations far above the rest.
Make it a monthly rhythm
Monthly: occupancy, ADR and RevPAR by property, revenue and commission by channel, and labor per occupied room. Compare each with the same month last year rather than the month before; a hotel's months are seasonal.
Let PlainSight read your PMS export
Upload a reservation or revenue export and PlainSight breaks revenue down by property, stream and channel, shows the seasonal shape and how concentrated it is, and writes the next steps in plain English. Everything runs in your browser — your data never leaves your device.
See it on a live example →
Opening or buying a hotel? Read the free handbook for opening and running a small hotel — the market, the stages ahead, how each kind of work pays, and what to do first. No account needed.
Frequently asked questions
- What file does PlainSight need from a hotel?
- A reservation or daily revenue export with a date, the stream, the channel and the amount; nights or rooms sold add occupancy and rate. One property or several.
- What is the difference between ADR and RevPAR?
- ADR is the average rate of the rooms you sold. RevPAR spreads room revenue over every room you had, sold or not, so it rewards filling rooms and holding rate together.
- Should I leave the OTAs?
- Rarely all at once. OTAs reach guests you wouldn't reach on your own; the aim is to pay their commission on first stays and win the repeat stays direct.
- Is my data safe if I use PlainSight?
- Yes. Files are processed entirely in your browser and never uploaded. Optional AI features send only anonymized summary totals, never names or raw rows.
Sources
- CoStar, “U.S. Hotel Forecast Assumptions,” August 2026, and Hotel Dive, “CoStar, Tourism Economics ‘significantly’ boost US hotel outlook”.
- Asian Hospitality, “Report: OTAs top direct bookings for independent hotels”.
- Preno, “How Much Commission Does Airbnb, Expedia and Booking.com Take”: typical OTA commission ranges.
- City of Los Angeles, Office of Wage Standards: the Citywide Hotel Worker Minimum Wage.
- Federal Trade Commission, Rule on Unfair or Deceptive Fees takes effect May 12, 2025.
This guide is general information for hotel owners and managers, not financial, tax, or legal advice. Figures described as “typical” or “common” are survey results and rules of thumb, not standards — always read your own numbers in context.