What is occupancy rate?
Occupancy is a percentage describing how full a property — or a whole portfolio — is over a period. One rental can be 50% occupied for the 4th of July; a 45-property portfolio can be 15% occupied in December. It’s a handy demand signal for a unit, a portfolio, or a market. Comparing against the industry is a fine secondary check, but your own historical data — this period versus the same period last year — tells you far more about what’s actually changing.
How to calculate it (and the blocked-days trap)
The math is simple: booked nights ÷ available nights. A property available 31 days in December with 20 booked nights is at 64.5%. The vacation-rental complication is owner stays, which show up as blocked dates — neither booked nor vacant — and shouldn’t count as available inventory. Removing blocked days is the accurate method, even though it nudges the number up; leave them in and occupancy is understated, which has a real cost downstream.
Why understated occupancy breaks pricing algorithms
When occupancy is understated, demand forecasts are wrong — and the damage is amplified because most pricing systems lean on data scraped from Airbnb and Vrbo, which can’t tell a blocked night from an unsold one. This is a core reason scraped data is dangerous for setting prices. The most accurate models bias strongly toward host data — your actual reservation records — for a far more reliable forecast.
What moves occupancy: shocks and trends
Demand shocks like COVID-19 or an Olympics produce dramatic swings — one nearly impossible to forecast, the other predictable in timing but not location. Other changes arrive as slow trends of a few points a year, like the multi-year shift from urban toward rural and drive-to markets. The job is recognizing which kind of change you’re looking at, because the right response differs.
Occupancy bias vs yield bias
Every strategy leans one way: occupancy-biased (take more reservations) or yield-biased (hold higher rates). An experienced revenue manager holds neither — the only goal is maximizing RevPAR. There are rational exceptions: a brand-new listing with no forecast justifies an occupancy bias, since bookings generate the data the model needs. And last-minute discounting is the telltale symptom of a missing forecast — if the discounted rate was acceptable close-in, why wasn’t it offered earlier?
Read occupancy as a trend, not a number
A static occupancy number tells you where you ended up; the trend tells you what to do next, while there’s still time to act on price and minimum-stay rules. Plot the current month against a benchmark — the same month last year is excellent under normal conditions — and investigate when you’re pacing behind early: did you just raise prices, or did competitors drop theirs?
Occupancy only matters because of RevPAR
RevPAR = occupancy × ADR. In isolation, occupancy is just a demand gauge — maximizing it alone is trivial (cut rates until you sell out), and maximizing ADR alone is just as easy if you don’t mind empty nights. The hard, valuable problem is maximizing their product, and the winning combination changes constantly.