Occupancy rate is the most basic STR metric — and one of the most commonly miscalculated. The formula itself is straightforward, but the inputs matter. Count the wrong nights and you'll end up with an occupancy figure that looks good on paper while hiding real performance problems.
What's the Correct Occupancy Formula?
Example: A property had 22 confirmed booking nights in a 30-night month, with 2 nights blocked for maintenance. Available nights = 30 − 2 = 28. Occupancy = 22 ÷ 28 = 78.6%.
What Counts as a Booked Night?
Only nights covered by a confirmed reservation count. In practice, that means reservations with a status of new, modified, confirmed, or accepted — depending on what your PMS calls them.
The statuses that do NOT count:
- Cancelled bookings — the property was not occupied
- Inquiries — a guest asked about the dates, nothing more
- Pre-approvals — not yet a confirmed booking
- Expired holds — the guest didn't complete the booking
Counting inquiries as bookings is the most common mistake. A single popular listing can receive 5–10 inquiries for the same weekend — counting them all can push occupancy above 100%.
Owner stays: in or out?
Whether to count owner stays depends on what you're trying to measure. Most operators track two separate things:
| Metric | Owner stays | Use for |
|---|---|---|
| Revenue occupancy | Excluded | Financial performance — how often the property earns revenue |
| Physical occupancy | Included | Usage tracking — how often the property is in use |
For portfolio analytics and any metric connected to revenue (like RevPAR), use revenue occupancy. Owner stays don't generate income, so including them inflates the occupancy figure without reflecting actual business performance.
What Counts as Available Nights?
Available nights is the total number of nights in the period minus any nights intentionally removed from inventory. This includes:
- Owner-blocked dates (personal use, renovations, maintenance)
- Nights where listing was deactivated or set to unavailable
Nights that are open on the calendar but didn't get booked — those are NOT subtracted. An empty available night is still an available night. Removing it from the denominator would make your occupancy look artificially high.
Example of the wrong approach: A unit had 20 booked nights and 10 empty nights in a 30-night month. If you subtract the empty nights from available nights, you get 20 ÷ 20 = 100% occupancy. That's meaningless — the 10 empty nights represent real lost revenue.
Multi-unit portfolios
For a portfolio, occupancy is the pool of all units combined — not an average of individual rates. Sum the booked nights across all units, divide by the sum of all available nights.
The difference matters when units have different available night counts. A unit that was offline for two weeks has fewer available nights than a unit that was active all month — averaging the percentages would distort the portfolio figure.
Common questions
Should cancelled bookings count toward occupancy?
Do inquiries count as booked nights?
Should I calculate occupancy on a calendar month or a rolling 30-day basis?
Why does my PMS show a different occupancy than what I calculate manually?
Get accurate occupancy automatically
BNBinsights connects to your PMS and calculates occupancy correctly — confirmed bookings only, deposits excluded, portfolio pooled.
Join the waitlist