Short-term rental occupancy rate: the denominator that lies

Last winter two dashboards told me two different things about the same one-bedroom in the same month. One said 61% occupancy. The other said 44%. Same apartment, same 30 days, the same bookings sitting in the same calendar. Neither number was wrong. They were dividing guest nights by two different totals — and until you know which total a tool slid under the line, an occupancy rate tells you almost nothing.

Occupancy rate is the most quoted and least defined number in short-term rental. This post is about what actually goes in the denominator, why your two favourite tools will never agree, and which version of the number you should steer the business by.

Why the same month shows two occupancy rates

Occupancy rate is always guest nights divided by some total. The arithmetic is trivial — a fraction. The entire argument is over the number on the bottom.

Three tools, three denominators, three answers for one apartment:

• A pricing tool wants the number to look healthy, so it divides by the nights you actually listed as available. That denominator is small, so the percentage is high.
• A tax-minded spreadsheet divides by every night in the period — including the week you blocked the place to stay there yourself, and the ten days it was shut for a bathroom re-tile. That denominator is the biggest possible, so the percentage is the lowest.
• A channel manager sometimes divides by nights since the listing went live, which quietly punishes a listing that only launched in March.

None of these is cheating. They are answering different questions. The pricing tool is asking "of the nights you tried to sell, how many sold?" The tax spreadsheet is asking "of the whole calendar, how much was earning?" Those are not the same question, and a single percentage can't answer both.

So the first move is never "what's my occupancy?" It's "guest nights over what?"

The three numbers, defined

There are really only three denominators worth tracking. Name them and the confusion evaporates.

Gross occupancy (also called calendar occupancy) is guest nights divided by total nights in the period — 365 for a year, 30 or 31 for a month. It is the most honest measure of how hard your asset is working, because it counts every night the property existed whether you offered it or not. It is also the number that drops every time you take a week off, which is exactly why pricing tools hate it.

Available occupancy is guest nights divided by the nights you actually offered for sale. You strip out the nights you deliberately blocked: personal use, maintenance, and the buffer days you auto-block between stays. This is the number that tells you whether your pricing and listing are working, because the denominator is only the inventory you were genuinely trying to sell.

Paid occupancy is paid guest nights divided by available nights. You subtract the comped friends-and-family stays and any night you discounted to zero. Most hosts don't track this and don't need to — but if you run a lot of "stay for free, just review honestly" nights, your available occupancy is flattering you, and paid occupancy is the corrective.

The gap between gross and available is the whole story. For a host who blocks 10% of the year for personal use and maintenance, gross occupancy runs roughly six to eight points below available occupancy on the same bookings. Quote one to your accountant and the other to yourself.

One apartment, one year, three rates

Here is a real-shaped year for a single city one-bedroom. Every night of the 365 lands in exactly one bucket:

Run the three rates:

• Gross occupancy = (210 + 9) ÷ 365 = 60.0%. Every guest night over the whole calendar.
• Available occupancy = (210 + 9) ÷ (365 − 28 − 10 − 18) = 219 ÷ 309 = 70.9%. Guest nights over the nights you actually offered.
• Paid occupancy = 210 ÷ 309 = 68.0%. The nine comped nights drop off the top.

Same apartment. Same 219 nights with a guest in the bed. The number is 60%, 71%, or 68% depending entirely on what you divide by — an eleven-point swing with no booking changed. If you ever compare your occupancy to another host's and one of you isn't sure which denominator the other used, the comparison is noise.

This is also why occupancy alone is a trap. The only way occupancy turns into money is through your nightly rate — the two multiply into RevPAR, revenue per available night, which is the number that actually pays the mortgage. A host at 80% occupancy and a low rate can earn less per available night than a host at 55% and a strong rate.

What "good" occupancy looks like — and the asterisk on every benchmark

Every benchmark you have ever read — "the average urban short-term rental runs 55%", "aim for 70%+", "anything under 50% is failing" — has a denominator it didn't tell you about.

Market data providers mostly report available occupancy, because they only see the nights a listing was open on the platform. They literally cannot see the week you blocked your calendar in your own head and never listed. So when a report says the median for your city is 58%, it is almost always 58% available, and your honest comparison number is your own available figure, not your gross one. Compare gross-to-published-benchmark and you will think you are underperforming a market you are actually beating.

A few rough shapes, all on an available-occupancy basis, all heavily market-dependent:

• A mature urban one-bedroom in a year-round demand city: 55–70% available occupancy is healthy.
• A seasonal coastal or mountain unit: the annual number can sit at 40–55% and still be very profitable, because the booked nights cluster in a high-rate season.
• A brand-new listing in its first 90 days: expect 30–45% while the platform ranking warms up and you have no reviews. Judging a three-month-old listing against a stabilised benchmark is the most common false alarm hosts give themselves.

There is no universal "good". There is only your available occupancy, this month versus the same month last year, in the same market. Year-over-year on a fixed denominator is the only occupancy comparison that means anything.

Raising the number that actually pays you

Once you steer by available occupancy, the levers that move it are specific and unglamorous. None of them is "lower your price until it sells".

Kill orphan nights. A two-night gap between a Sunday checkout and a Wednesday check-in is nearly impossible to fill at most minimum stays — it sits empty and drags available occupancy down. Tightening minimum-stay rules around known gaps, or pricing those exact nights to move, recovers nights that would otherwise rot. The full math is in orphan nights and gap nights.

Right-size the minimum stay. A three-night minimum protects your turnover costs but quietly blocks every one- and two-night request, and those rejected nights never appear in your booked total. The minimum-night-stay math is the trade-off between cleaning cost and lost availability — and lost availability shows up directly as lower available occupancy.

Don't drown the denominator in buffer days. Every auto-blocked buffer night is a night removed from sale. One buffer night between stays is often worth it for the cleaner; two is usually you padding your own anxiety at the cost of 5–10% of your sellable inventory. I learned this the wrong way: I once removed all buffers to chase a 75% occupancy target, hit a same-day turnover my cleaner physically could not make, and ate a three-star "apartment wasn't ready" review that cost me more bookings than the extra nights earned. The buffer question is a real cost trade, not a vanity dial — cleaning buffer days walks through when each is right.

Sell the short-lead-time nights. A night that is still empty 48 hours out is almost certainly going to stay empty at your standard rate. A standing last-minute discount on the closing window converts nights that would otherwise be pure loss — and every one of them lands in your booked total, lifting available occupancy where it counts.

A free property manager that keeps your calendars in one place is the precondition for all of this: you cannot manage a denominator you cannot see, and a calendar split across three platform tabs hides exactly the orphan nights and buffer blocks that are quietly eating your number.

One opinionated take

Stop quoting a single occupancy number. The moment you say "I'm at 65%" without saying over what, you have told a colleague nothing and, worse, fooled yourself into a comparison you can't actually make. Track available occupancy as your steering number and gross occupancy as your reality check, watch each one year-over-year on a denominator that never changes, and treat every benchmark you read as available-occupancy-until-proven-otherwise. The hosts who plateau are the ones chasing a percentage; the hosts who compound are the ones who know exactly what they're dividing by.