Airbnb Revenue Estimator Actor: short-term rental market rollup with seasonality

This Airbnb revenue estimator actor returns one row per market with a short-term rental market rollup: nightly rate, occupancy days and revenue as medians and means with their denominators named, a room and bedroom mix, a neighbourhood mix, and a month-by-month booked-share curve. Give it a market name for the published snapshot rollup, and add a US address for the live estimate alongside it. The two indexes sit side by side with their own dates and are never averaged. Aggregates, attribution and the snapshot date only.

What you get

• One dataset row per market, so sizing forty cities is billable only for the markets you actually received.
• Nightly rate as a full distribution: count, median, mean, 25th and 75th percentile. Not one number pretending to be the market.
• Occupancy days over the last year and availability over the next, each with its own distribution.
• The revenue block, which is where most tools go wrong. Median as the headline, and all three ways of averaging the column reported beside the count each was taken over: the mean over rows carrying a revenue figure, the mean over rows with a revenue figure above zero, and the mean over every listing in the file with missing counted as zero. The filter is pinned in the code and stated on the row, so no figure can be quoted without its denominator.
• Room mix (entire home, private room, shared room, hotel room), bedroom mix (studio to 7 plus) and the ten largest neighbourhoods by listing count.
• A month-by-month curve: listed days and booked share per month, aggregated over every listing's calendar.
• The live estimate for a US address: average daily rate, occupancy percentage, monthly revenue, revpan, the 25th and 75th percentile bands for each, the seasonalised annual revenue with its bands, a twelve-month revenue curve, and the number of comparables behind it.
• snapshot_as_of on the row: the date the published snapshot was taken, not the date you ran the Actor. A six-week-old snapshot is a fact you should see.
• Required attribution on the row, as the data licence asks.
• A data_coverage count block saying which of the three blocks returned figures, and indexes_are_never_averaged set to true so no consumer silently blends a June snapshot with a live estimate.

Why scrape short-term rental market data

An investor about to sign a lease on an arbitrage deal needs to know what the market pays before the lease exists. A lender or an insurer pricing short-let risk needs the same numbers, at market grain rather than listing grain. A platform sizing a new city needs the distribution, not the average.

The problem with the tools that already answer this is the shape of the answer. A free calculator returns one estimate and an email funnel. A subscription product returns the same numbers behind a login and a plan. A listing scraper returns ten thousand rows and no rollup, leaving the buyer to work out the average themselves, which is where the trouble starts.

It starts because the revenue column has three honest averages. On a real market snapshot of just over ten thousand listings, the mean over rows that carry a revenue figure is 18,297. The mean over all listings with missing counted as zero is 11,253. The mean over nonzero rows only is 23,350. Those are three different questions. A tool that reports one of them without naming its denominator hands a lender a number that will not survive scrutiny, and an investor a number that will not survive a spreadsheet.

So this Actor reports the median as the headline and puts every mean beside the count it was taken over. That single decision is the difference between a figure you can put in a committee paper and a figure you cannot.

Input

Field	Required	What it does
markets	Yes	Markets to roll up. City name, or city, country to disambiguate. Up to 20 per run.
address	No	A US street address for the live tier. The live tier is asked once per run regardless of how many markets you list.
include_snapshot	No	Aggregate the published snapshot. On by default.
include_seasonality	No	Month-by-month booked share. On by default.
include_live	No	Live estimate for the address. On by default, skipped when no address is given.
min_bedrooms	No	Narrow the live tier's comparables by bedroom count. Default 1.
max_bedrooms	No	Upper bound on bedrooms for the live tier. Default 5.
max_markets	No	Upper bound on markets per run, 1 to 20. Default 5.

Output

{
  "market": "Amsterdam",
  "as_of": "2026-09-25",
  "snapshot_as_of": "2026-06-15",
  "snapshot_index": {
    "grain": "market",
    "listing_count": 10369,
    "price": { "count": 6377, "median": 291.0, "mean": 346.62, "p25": 206.28, "p75": 411.0 },
    "occupancy_days_l365d": { "count": 10369, "median": 16.0, "mean": 48.31 },
    "revenue_l365d": {
      "count": 6377,
      "median": 9496.0,
      "mean_with_figure": 18296.92,
      "mean_with_figure_count": 6377,
      "mean_nonzero": 23349.9,
      "mean_nonzero_count": 4997,
      "mean_all_listings_missing_as_zero": 11252.72,
      "mean_all_listings_count": 10369,
      "filter": "median over listings carrying a revenue figure; every mean named with its denominator"
    },
    "room_mix": { "Entire home/apt": 8489, "Private room": 1833, "Hotel room": 26, "Shared room": 21 },
    "bedroom_mix": { "1": 5250, "2": 2540, "studio_or_0": 1154, "3": 1057 },
    "neighbourhood_mix": { "De Baarsjes - Oud-West": 1818, "De Pijp - Rivierenbuurt": 1195 }
  },
  "seasonality": {
    "grain": "market",
    "day_count": 3819725,
    "months": {
      "2026-06": { "days_listed": 128962, "available_share": 0.2072, "booked_share": 0.7928 },
      "2026-07": { "days_listed": 324415, "available_share": 0.2508, "booked_share": 0.7492 }
    }
  },
  "live_index": {
    "grain": "market",
    "status": "ok",
    "address": "200 E 6th St, Austin, TX 78701",
    "comparable_count": 97,
    "adr_average": 410.39,
    "adr_p25": 257.0,
    "adr_p75": 531.0,
    "occupancy_pct_average": 49,
    "monthly_revenue_average": 5749.98,
    "seasonalized_annual_revenue_average": 66262.1,
    "seasonalized_revenue_by_month": []
  },
  "indexes_are_never_averaged": true,
  "field_grain": {
    "snapshot_index": "market",
    "seasonality": "market",
    "live_index": "market"
  },
  "data_coverage": { "snapshot": 1, "seasonality": 1, "live": 1 },
  "missing_blocks": [],
  "attribution": {
    "snapshot": "Market snapshot data by Inside Airbnb, licensed CC BY 4.0",
    "live": "Live estimate figures as published by the calculator at rabbu.com"
  }
}

Figures shown are from a September 2026 run over the June 2026 Amsterdam snapshot and the Austin live estimate, and move as the markets move.

Use cases

Rental arbitrage before a lease signature. The operator has a target neighbourhood and a rent number. The rollup gives the nightly rate distribution and the revenue median with its denominator; the live tier gives the estimate for the specific address and bedroom band. Both figures are on one row with their dates, so the comparison is honest.

Lender or insurer pricing short-let risk. Market grain is what a credit paper needs: how many listings, what mix, what the revenue distribution looks like, and which months pay. The seasonality curve is the part a spreadsheet built from listing rows never produces cleanly.

Private equity market sizing. Twenty markets in one run, one row each. max_markets bounds the run and the per-market charge means a pilot of three markets costs a pilot's money. The neighbourhood mix says where the market actually concentrates.

A platform deciding which city to launch. Booked share by month answers whether a city is a year-round market or a summer one, and the room mix says whether the inventory is whole homes or rooms. Both come from the published snapshot rather than from a live scrape, which is the point when the decision is about the market and not about one listing.

How it compares

	This actor	xtracto/airbnb-occupancy	malikgen/airbnb-revenue-calculator	AirDNA MarketMinder
Output grain	Market rollup	Per listing / one estimate	One estimate	Market
Per 1k rows	$50.00	Free (compute only)	Free (compute only)	Subscription
Revenue denominators named	Yes, all three	No	No	No
Median headline	Yes	No	No	Yes
Seasonality curve	Yes	No	No	Yes
Live estimate in the same call	Yes	No	Yes	Behind a login
Aggregates only, attributed	Yes	No	No	Yes
Snapshot date on the row	Yes	No	No	Yes

The closest competitor by scope is xtracto/airbnb-occupancy at 43 users, free, returning occupancy figures rather than a rollup. The comparison a buyer in this category actually makes is against AirDNA, which publishes the same class of number behind a subscription and a signed-in session, and against the free calculators that return one estimate and an email funnel. What those have that we do not: a live per-listing rentaliser on demand, and AirDNA's longer historical series. What this Actor has that they do not: the three revenue means with their denominators on the row, a seasonality curve, and a data product rather than a lead form.

Pricing

All pricing is pay-per-event at $0.05 per market, one event per returned market rollup row. $50.00 per 1,000 markets. Rows that carry no figures at all are delivered but never charged. All pricing is pay-per-event, you only pay for results you receive. No actor-start fee, no per-compute-unit charges.

Limits and gotchas

• The published snapshot is taken on a schedule and carries its own date. The row reports it as snapshot_as_of. A market published in June will show June numbers in October. That is the data, not a bug.
• Aggregates only. The data licence allows derived statistics with attribution and the snapshot date and does not allow redistributing the listings, so no listing identifier, per-listing price or coordinate appears in the output. If you need listing rows, you need a different product.
• The revenue column's three means are three different questions. Quote the median and its denominator. Do not quote a bare mean and do not average the three together.
• The live tier covers US addresses only and is asked once per run, no matter how many markets you list. It answers one address at a time and a burst earns a refusal, so the Actor runs it sequentially and reports a refusal as status: "blocked" rather than retrying into a block.
• The live tier's comparables are filtered by min_bedrooms and max_bedrooms. The defaults are 1 and 5. A studio-only market needs them lowered.
• The two indexes are never averaged. indexes_are_never_averaged is true on every row, and the two blocks carry their own dates. Blending a snapshot median with a live average is a decision for you to make deliberately.
• A market pulls two compressed files of several megabytes each and aggregates a few million calendar rows. One market takes tens of seconds; twenty takes several minutes.

FAQ

How much can I make on an Airbnb in a given market? The rollup gives the market's revenue median with the count it was taken over, and the live tier gives an estimate for a specific US address with 25th and 75th percentile bands. Quote the median, not the mean: on a real market the three ways of averaging the revenue column differ by a factor of two.

What does the Airbnb revenue estimator return that a calculator does not? A calculator returns one number for one address. This Actor returns the market around it: the rate distribution, the occupancy days, the revenue denominators, the room and bedroom mix, and the months that pay. The live estimate is one block of that row, not the whole answer.

Is there a seasonality view? Yes. seasonality.months carries listed days and booked share per month, aggregated over every listing's calendar. It is the fastest way to see whether a market is year-round or seasonal.

Do I get per-listing rows? No. The output is aggregates with attribution and the snapshot date, which is what the data licence allows. Per-listing rows would be a different product and a different licence.

Can I price a short-term rental market outside the US? Yes, for the snapshot rollup. The live estimate tier is US addresses only. Amsterdam, Lisbon, Barcelona and 120 other markets have published snapshots.

Why are there three different revenue figures in the row? Because the column has three honest averages and they differ by a factor of two. Each is reported with the count it was taken over so a reader knows which question it answers. The headline is the median.

Related Actors

• Cost of Living Actor for the cost side of a market next to its rental revenue.
• Zillow Scraper for US rental listings and long-let pricing beside short-let revenue.
• Apartments Scraper for long-let inventory next to a short-let market rollup.