Airbnb Data for Short-Term Rental Investors: Pricing, Occupancy Signals & Competition
A short-term rental purchase rests on three guesses: what a night will sell for, how often it will sell, and how many similar listings compete for the same guests. Public listing pages answer part of each question. Airbnb data will not give you a bank statement for someone else's property, but it shows asking prices, review depth, ratings and supply mix across a whole city at a single moment.
To make this concrete we collected a free sample of Lisbon listings. It is available as a CSV on the Airbnb Lisbon dataset page, and every number below is computed from that file. The rest of this post covers what the data can tell an investor, what it cannot, and how to collect it reliably.
Why investors look at Airbnb data
Investors compare markets before they compare buildings. The questions are usually the same:
- What is the typical nightly rate for a one-bedroom, and how wide is the range?
- Which areas price above the city median, and is the premium consistent?
- How much of the supply is entire apartments versus private rooms?
- Do highly rated listings charge more, or do they simply have more reviews?
Short term rental data answers these at scale. A single search page shows only a handful of listings, while a structured dataset lets you filter, group and compare. If you are also looking at purchase prices, pair it with sale-side data such as the portal figures in our Zillow real estate data guide, and for destination demand context see our Tripadvisor travel insights post.
What our Airbnb sample shows
The sample holds 87 unique listings from a Lisbon search for a two-night stay for two adults, starting roughly 30 days after collection. Prices are in euros and include the nightly rate and the two-night total with taxes. Here is what the file contains:
- Nightly price: the average is €77.38 and the median is €72.90. The cheapest listing is €41.50 and the most expensive is €139.50.
- Total for two nights: the median is €158, which is above two times the nightly rate because taxes are added at checkout.
- Room type mix: 73 of 87 listings (83.9%) are apartments, 10 are private rooms, 2 are hotels, 1 is a home and 1 is a loft.
- Room type pricing: the median nightly rate is €75.00 for apartments and €57.00 for private rooms.
- Bedrooms: 58 listings state one bedroom, with a median of €74.40. The 8 two-bedroom listings have a median of €85.25. Twenty listings do not state a bedroom count, which is typical for studios and rooms.
- Ratings: across the 85 listings with a rating, the average is 4.77 and the median is 4.81. Half of them (43 of 85, 50.6%) sit at 4.8 or higher.
- Review depth: the median listing has 167 reviews, and only 3 of 87 have fewer than 10.
- Rating and price: listings rated 4.8 or higher have a median nightly price of €83.33, against €68.28 for those below 4.8.
Treat the rating and price gap as a correlation, not a cause. Better-rated listings may also be larger, better located or newer. You need the full set of attributes to separate those effects.
Sample rows
| Room type | Area | Bedrooms | Rating | Reviews | Nightly (EUR) | Total (EUR) |
|---|---|---|---|---|---|---|
| Apartment | Lisbon | 1 | 4.95 | 203 | 65.00 | 146.00 |
| Apartment | Lisbon | 1 | 4.71 | 694 | 60.30 | 136.60 |
| Apartment | São Paulo | 1 | 4.88 | 41 | 56.25 | 135.25 |
| Apartment | Lisbon | 2 | 4.84 | 180 | 67.50 | 151.00 |
| Apartment | Lisbon | 1 | 4.91 | 186 | 69.30 | 154.60 |
The full file has the listing ID, title, URL and Superhost flag as well.
Pricing by area
Area is where investors spend most of their time. In the sample, the biggest group is listings shown simply under "Lisbon" (26 listings, median €73.35). Among named areas with at least four listings:
- São Paulo (10 listings) has a median of €85.60, the highest in that group.
- Santa Maria Maior (7) has a median of €80.50.
- Graça (6) has a median of €78.40.
- Santo António (4) has a median of €68.25.
- Areeiro (4) has a median of €57.00, well below the city-wide median.
Estoril, outside the city proper, has three listings with a median of €47.50, the lowest in the file. The highest single rate among areas with one listing is Parque das Nações at €116.95.
Be careful with small counts. A median built from four listings can move a lot with one addition. The point of the sample is to show the shape of the data and the fields available. A production dataset with several hundred listings per area gives much more stable numbers.
The Superhost question
Investors often ask whether the Superhost badge supports a higher price. In our sample, only 4 of 87 listings (4.6%) carry the Superhost badge on the search result card, with a median nightly rate of €75.00 against €72.90 for the rest. Four listings is too few to draw a conclusion, so we would not claim a premium from this file.
This is a useful reminder about sampling: the first few result pages are ranked by Airbnb's own ordering, so they are not a random draw. A broader collection across more pages, price bands and dates is needed to test the Superhost effect properly. Note that the search results also show a separate "Guest favorite" badge that is not the same thing as Superhost, so check which one your field refers to.
Occupancy signals: what you can and cannot infer
Public pages do not show booking calendars in a way that confirms how many nights were sold. What you can do is use proxies:
- Review count: a high count suggests a listing that has been active for a long time. Use it as an age and activity proxy, not as occupancy.
- Availability for specific dates: if a search for given dates returns fewer listings than usual, supply is tightening. Running the same search for several dates over weeks shows this pattern.
- Price movement over time: repeated snapshots reveal when hosts raise or lower rates for the same dates.
Avoid presenting any of these as a measured occupancy rate. They are indicators, and a careful investor stress-tests the plan against lower occupancy than the optimistic case.
Fields that matter for competition analysis
If you build a competitor set for a property you are considering, these fields carry the most weight:
- Room type and bedrooms: compare like with like. A private room and a two-bedroom apartment are different products.
- Area: the neighborhood label is the cheapest way to group listings by location.
- Nightly price and total price: the gap between the two shows fees and taxes, which matter to the guest's decision.
- Rating and review count: read them together. A 4.95 with 12 reviews is weaker evidence than a 4.71 with 694.
- Listing ID and URL: stable identifiers let you track the same listing across snapshots.
- Scrape timestamp: prices belong to a date, so you need to know which one.
We deliberately leave out host names and other personal details. Pricing and competition analysis does not need them.
Common challenges with Airbnb scraping
Airbnb scraping looks easy on the first page and gets harder at scale. These are the problems we plan for:
- Anti-bot measures: requests from data-center addresses are often blocked or served an empty page. Managed infrastructure with escalation to residential addresses and browser rendering helps.
- Pagination: results are paged in blocks using cursors, and the same listing can appear twice, so you must remove duplicates by listing ID.
- Date-dependent prices: a price without check-in and check-out dates means nothing. Always store the dates searched next to the price.
- Layout and data changes: the structure of the embedded page data changes, so parsers need monitoring and regular fixes.
- Missing fields: not every card shows a bedroom count or a rating, especially for new listings. In our sample 2 of 87 had no rating, and we left those cells empty rather than guess.
- Rate shown vs. rate paid: discounts, cleaning fees and taxes change what a guest pays. Capture both the nightly and the total price.
Use cases beyond buying
- Acquisition screening: rank cities and areas by median nightly rate and supply mix before visiting any property.
- Pricing for existing hosts and managers: benchmark against nearby listings of the same type and size.
- Lenders and advisors: support a business plan with market-level price data rather than a single comparable.
- Researchers and city analysts: study the balance between whole-home and room supply.
How often to refresh
Short-term rental prices move with season, events and lead time, so a single snapshot goes stale quickly. For an investment decision, collect the same search for several future date windows (a weekday, a weekend, a peak holiday and a low-season week). For ongoing pricing work, weekly snapshots are common. Quarterly refreshes are usually enough for a market screening exercise. Whatever the frequency, keep every snapshot with its timestamp so you can compare changes.
FAQ
Is the sample representative of all Lisbon listings? No. It is 87 listings from the first result pages for one date window. It shows the data structure and gives indicative figures, not a full market census.
Can I see exact occupancy from this data? No. Public pages do not confirm nights booked. Review counts and availability for repeated date searches are proxies only.
Does the dataset include host names? No. We collect listing-level data only, such as price, room type, area, rating and review count.
Can you collect other cities or dates? Yes. Tell us the city, the stay dates and the fields you need, and we will scope it. Datasets start from $49 on our pricing page.
Get the data for your own market
Download the free Lisbon sample and test it in your spreadsheet or notebook. If you are screening a different city or want scheduled refreshes, request a free sample of your own target and we will return a file in the same format.
Author: DataHarbor Team