STR Occupancy: Why Benchmarks Mislead and What to Check

TL;DR: There is no occupancy benchmark worth targeting, and the market-type ranges circulating in STR coverage are modeled estimates presented with more confidence than the method supports. Nightlydata does not publish occupancy benchmark numbers, for reasons this article sets out. What is worth doing instead: compare your measured occupancy against comparable listings in your own neighbourhood, and when a property underperforms, work through gap nights, price position, ranking signals, and calendar blocks in that order.

Why the “average occupancy rate” stat is misleading

Industry-wide occupancy averages circulate constantly in STR coverage. They are close to useless for operational decisions, for a reason that is structural rather than a matter of the average being imprecise.

A single national figure blends unit types, seasons, and market structures that have nothing to do with one another: a dense urban studio with flat year-round demand, a ski cabin that earns most of its year in four months, a coastal property whose August and February are different businesses. Averaging them produces a number that describes none of them. Moving from a national average to a market-type average narrows the blend but does not fix the problem, because the variance inside “urban” or “beach” stays larger than the gap between them.

The right comparison is narrower than any published benchmark can be: how does your listing perform against comparable properties in the same specific neighbourhood, with similar bedroom counts, price range, and listing maturity? That comparison you can actually make, from your own booking records and your platform dashboard.

Why we do not publish occupancy benchmarks

This is a deliberate editorial position, not a gap we intend to fill later.

Occupancy figures for a market are almost never measured. They are inferred from what is publicly visible, and the inference carries assumptions that move the output substantially.

Inside Airbnb documents its method openly, which makes it a useful worked example of the general problem. It compiles public listing data including each listing’s 365-day availability calendar and its reviews, then estimates bookings from review counts using what it calls the San Francisco Model: a 50% review rate, chosen as a middle ground between a 72% rate it judges unreliable and a 30.5% rate derived from New York Attorney General data, multiplied by an average length of stay, with the resulting occupancy capped at 70% (source: Inside Airbnb, retrieved July 2026).

That cap matters more than it first appears. A market where competent operators genuinely run at 80% cannot read above 70 in that dataset. The ceiling compresses precisely the part of the distribution an operator most wants to see, and it is a deliberate conservatism rather than a measurement.

Paid market-data tools model differently, and their outputs move on choices that are rarely visible to the buyer: whether a night that vanished from a calendar was sold or blocked by the host, whether inactive listings sit in the denominator, whether the same unit listed on several platforms is counted once or three times, and where the boundary of “the market” was drawn. Two tools can report materially different occupancy for the same city without either being wrong.

Publishing a tidy “urban markets run at 67%” on top of that would lend false precision to a modeled estimate. How to read any aggregated STR figure is covered in our piece on short-term rental data.

If you want a number to steer by, use break-even occupancy instead. It comes from your own cost structure, it is measured rather than modeled, and it answers a question that changes decisions: how full does this unit need to run before it earns anything.

The diagnostic framework

When a property underperforms the comparable listings you can see in your own neighbourhood, the cause is usually one of four things. Work through them in this order, because each one invalidates the diagnosis of the next.

The checks below are starting points for investigation, not measured findings. Calibrate them against your own portfolio: what matters is which properties sit far outside your normal range, not whether they cross a number quoted in an article.

graph TD
    A[Occupancy below your neighbourhood comps?] --> B{Comparing measured against measured?}
    B -->|Your data vs a modeled estimate| Z[Invalid comparison, re-baseline first]
    B -->|Comparison is sound| C{Gap nights high for your portfolio?}
    C -->|Yes| D[Minimum stay settings are the first suspect]
    C -->|No| E{Price above your comp set median?}
    E -->|Yes| F[Pricing or repricing strategy]
    E -->|No| G{Weak ranking signals: ratings, reviews, responsiveness?}
    G -->|Yes| H[Operational fix, not a pricing fix]
    G -->|No| I{Large share of calendar blocked?}
    I -->|Yes| J[Real availability lower than the data suggests]
    I -->|No| K[Look at demand for this specific area]

Gap nights. Bookings leave gaps on either side that can be too short to sell under your own minimum stay rules. When an unusual share of your available nights sits unbookable for this reason, the minimum stay settings are costing more than they protect. Our minimum stay optimization guide works through the tradeoff.

Pricing position. A listing priced well above the median for genuinely comparable units will appear in search results without converting. Occupancy responds to a price change with a lag, since it moves future bookings rather than tonight’s, so give a change time to show up before judging it and do not adjust daily.

Listing quality and ranking signals. Airbnb states that search ranking considers listing quality, including ratings and reviews, alongside price, location, popularity, and host responsiveness (source: Airbnb Help Center, retrieved July 2026). Airbnb publishes no numerical thresholds for any of these, so treat a specific star-rating cutoff quoted anywhere as someone’s inference rather than a platform rule. The direction is what is actionable: a young listing with few reviews competes at a structural disadvantage against an established one, and the fix is operational rather than a price cut. See our listing ranking guide.

Calendar blocking. Nights blocked for owner use or maintenance were never for sale, but a third-party tool reading the public calendar cannot know that and will count them against you. A unit available 200 nights that books 140 is running at 70% of what it actually offered, while a market tool working from the full year sees 38%. Know your real availability before diagnosing anything.

Platform analytics vs third-party market data

The two sources answer different questions, and the most common analytical error is treating them as interchangeable.

Platform analytics (Airbnb’s host dashboard, Vrbo’s owner dashboard) show how your listing performs against similar listings on that platform. Your occupancy, earnings, and search impressions on that channel, measured. What they cannot show: total demand in the market, how units at other price points or amenity levels are doing, or what comparable properties earn on competing channels.

Third-party market tools aggregate across platforms and give market-level estimates. They see the comp set more broadly than any single dashboard, at the cost of everything being modeled rather than measured. Our comparison of STR market research tools covers which is worth paying for at which portfolio size.

Use them in sequence rather than in parallel:

1. Check your platform dashboard against similar listings on that platform. Below the platform benchmark means the problem is on-platform: pricing, reviews, responsiveness, listing quality.
2. If on-platform performance looks normal but total occupancy is low, the gap is probably distribution. A property leaning on one channel in a multi-channel market will structurally underperform a well-distributed competitor, which is the subject of OTA distribution beyond Airbnb.

Seasonality: why an annual figure hides the decision

Annual occupancy conceals the information that actually drives revenue: whether you extracted the available value in peak weeks and managed the shoulder acceptably.

Track monthly targets separately. A coastal unit has a different target in August than in February, and holding both to one annual number tells you nothing about which one you got wrong.

The useful diagnostic for a seasonal market: work out what occupancy your top three months would need to hit your annual revenue target, then check whether you are getting it. Hitting peak while missing the annual target locates the problem in the off-season, which points at pricing, minimum stays, or a genuine demand gap rather than listing quality. For operators running dynamic pricing, shoulder-season occupancy gaps are often a tool configured to defend rate rather than fill nights.

Key facts

• Market occupancy figures are modeled estimates, not measured bookings, and the modeling assumptions move the output materially.
• Inside Airbnb’s San Francisco Model infers bookings from reviews at a 50% review rate and caps estimated occupancy at 70%, so genuinely higher-performing markets cannot read above that ceiling (source: Inside Airbnb, retrieved July 2026).
• Your measured occupancy and a market tool’s modeled figure are different quantities; comparing them directly is the most common diagnostic error.
• Nightlydata does not publish occupancy benchmark numbers by market type, because doing so would lend false precision to modeled estimates.
• Airbnb states that ranking considers quality, ratings and reviews, price, location, popularity, and host responsiveness, and publishes no numerical thresholds (source: Airbnb Help Center, retrieved July 2026).
• Break-even occupancy, derived from your own costs, is more decision-relevant than any market benchmark.

Methodology note

This article publishes no occupancy benchmark figures, by design. The only quantitative claims it makes about third-party data are descriptions of published methodology, each linked to its primary source and dated. Where a check appears in the diagnostic framework it is flagged as a starting point for investigation, calibrated against your own portfolio, not as a measured finding.

Inside Airbnb publishes open datasets for major cities under a CC BY 4.0 licence and documents its assumptions publicly, which is why it is used here as the worked example of how occupancy estimation actually works.

Information current as of July 2026.