By Ryan Drew | Superhost and Vacation Rental Expert | Last Updated April 23, 2026
AirDNA claims its data is 97.5% accurate. That number comes from a CBRE report and it’s real — but it doesn’t mean what most hosts and investors assume it means. That accuracy figure applies to market-level supply data, not to the revenue projections for your specific property.
After 9 years hosting in Boise and using AirDNA regularly, here’s the honest breakdown of where the data is genuinely reliable and where you need to use it more carefully.
Check your own market’s data for free — no credit card required. See occupancy rates, competitor pricing, and the Rentalizer before you pay anything.
Try AirDNA Free →Quick Answer: AirDNA’s market-level data — aggregate occupancy rates, average daily rates, seasonal demand patterns — is reliably accurate for most established markets. Individual property revenue projections, especially in small or rural markets, are directional estimates, not financial forecasts. Use them to inform decisions, not to build spreadsheets.
What the 97.5% Accuracy Claim Actually Means
AirDNA’s published accuracy figures come from a CBRE study that measured two specific things: the accuracy of AirDNA’s active supply count (97.5% accurate) and the accuracy of its revenue data (96.2% accurate). Both are market-level measurements — how well AirDNA tracks the total number of active listings in a market and how close its aggregate revenue estimates are to actual Airbnb-reported revenue.
This is meaningful and genuinely impressive. It means if you’re looking at a market and asking “how many active listings are there?” or “what does the average host in this market earn?” — AirDNA’s answer is very close to reality.
What it doesn’t measure: how accurately AirDNA estimates the revenue for a specific property at a specific address. That’s the Rentalizer — and it’s a fundamentally different calculation using comparables, not verified transaction data.
Ryan’s Take: The 97.5% accuracy figure gets quoted constantly, but it’s measuring the wrong thing for most people asking the question. Hosts aren’t asking “how many listings are in my market?” — they’re asking “how much will my property earn?” Those are two very different questions, and the accuracy of the first doesn’t guarantee the accuracy of the second.
How AirDNA Collects Its Data
Understanding where AirDNA’s data comes from is the fastest way to understand where it’s reliable and where it isn’t.
AirDNA collects data through two primary channels. The first is scraping — daily automated collection of publicly visible listing information from Airbnb and Vrbo. This gives AirDNA visibility into listed prices, calendar availability, property details, and review counts across over 10 million listings globally.
The second is direct partner data — actual reservation data from over 1.1 million properties connected to AirDNA through channel managers, property management systems, and individual hosts who have linked their accounts. This data comes directly from bookings and is 100% verified. It’s used as a calibration layer to validate and improve the accuracy of the scraped estimates.
The key challenge AirDNA had to solve: Airbnb stopped showing whether calendar dates were booked vs. host-blocked in 2015. AirDNA trained a machine learning model on 18 months of pre-obscuration data to distinguish between the two — using 16 different booking signals including stay length, lead time, and booking patterns. This booked-vs-blocked methodology is the technical foundation of AirDNA’s occupancy estimates.
Where AirDNA Data Is Most Reliable
AirDNA performs best when there’s a lot of data to work with. The more active listings in a market, the better its machine learning models can calibrate against real outcomes. In dense, established markets the aggregate metrics — occupancy rates, ADR trends, seasonal demand patterns, revenue per available night — are consistently close to reality.
| Data Type | Reliability | Why |
|---|---|---|
| Market-level occupancy rates | ✅ High | Aggregate across many listings smooths out individual errors |
| Average daily rate (ADR) trends | ✅ High | Listed prices are scraped directly and updated daily |
| Seasonal demand patterns | ✅ High | Multi-year historical data makes seasonality very reliable |
| Competitor calendar pricing | ✅ High | Listed future rates are directly visible and scraped daily |
| Rentalizer in large markets | ✅ Solid | Many comps available to build a reliable estimate |
| Rentalizer in small/rural markets | ⚠️ Use carefully | Fewer comps = wider variance in estimates |
| Revenue estimates for premium properties | ⚠️ Often low | Averages based on all comps, not top-performers; amenities like hot tubs and pools aren’t weighted |
| Individual competitor occupancy | ⚠️ Estimated | Booked vs. blocked is modeled, not observed — occasionally misclassified |
Where AirDNA Is Less Reliable
Small and Emerging Markets
This is the most consistent criticism from real-world users. In markets with fewer than 100–200 active listings, AirDNA has less data to build reliable averages from. The Rentalizer estimates become wider-range guesses based on fewer comps, and individual occupancy estimates can swing significantly. Hosts in rural markets, small towns, or newly developing STR destinations should treat AirDNA data as a starting point and validate it with local knowledge.
Premium Properties with Standout Amenities
AirDNA estimates revenue based on comparable listings — and “comparable” defaults to bedroom count and location without fully weighting premium amenities. A property with a private hot tub, pool, or high-end finishes in a market where most comps are standard rentals will typically see AirDNA underestimate its revenue potential. The platform compares you to the average, not to the top performers your property actually competes with.
The fix: when using the Rentalizer, manually refine your comp set to listings that actually match your property’s amenity level. The paid plan lets you do this — and it meaningfully improves estimate accuracy for non-average properties.
Properties Listed on Multiple Platforms
If a property is cross-listed on both Airbnb and Vrbo, AirDNA uses a deduplication algorithm to avoid double-counting occupancy. This is generally accurate but occasionally misses duplicates, which can slightly overstate market supply in markets with high cross-listing rates.
Very New Markets or Rapidly Changing Conditions
AirDNA’s market-level data is updated weekly, not in real time. In markets experiencing rapid change — a new short-term rental regulation, a major employer relocating, a viral social media moment — AirDNA’s data may lag several weeks behind actual market conditions. For forward-looking pricing decisions in volatile markets, supplement AirDNA with direct calendar observation of competitors.
Ryan’s Take: I use AirDNA’s Competitor Calendar every week and I trust it for pricing decisions. What I don’t do is take a Rentalizer estimate and put it in a financial model as a hard number. I treat it as one data point — usually directionally correct, occasionally off by 15–20% in either direction. For my Boise market specifically, the market-level data is genuinely accurate. If I were buying in a market with 50 total listings, I’d cross-reference with a local property manager before relying on it.
How to Get the Most Accurate Results from AirDNA
AirDNA’s accuracy isn’t fixed — you can meaningfully improve the quality of estimates you get by using the platform correctly.
- Refine your comp set manually. The Rentalizer defaults to all nearby listings of the same bedroom count. Narrow it to properties that actually match your amenities, quality level, and guest profile. This is the single biggest improvement you can make to estimate accuracy for non-average properties.
- Look at trends, not snapshots. A single month’s occupancy figure means less than the 12-month trend line. Use AirDNA’s historical data to understand seasonality patterns rather than relying on any single data point.
- Cross-reference the Rentalizer with AirDNA’s market data. If the Rentalizer says $48,000/year but the market average for your bedroom count is $32,000, dig into why — it usually means your comp set needs adjustment or your property has atypical characteristics.
- Use the Confidence Score. AirDNA’s Rentalizer shows a confidence score based on how many strong comps are available in your area. Low confidence = fewer comps = wider variance. Pay attention to this number before drawing conclusions.
- Verify with local knowledge. Talk to a local property manager or look at actual booking calendars of comparable listings for cross-validation before making major investment decisions based on AirDNA data alone.
The Bottom Line
AirDNA is genuinely accurate for the things most hosts and investors need: understanding how a market performs, what seasonality looks like, how competitors are pricing, and whether a market is worth pursuing. At the market level, the data is reliable enough to make real decisions from.
For property-specific revenue projections — especially in smaller markets or for non-typical properties — treat AirDNA as a strong directional tool rather than a financial forecast. The data gets more reliable the more you refine your comp set and the more active listings exist in your market.
If you haven’t used AirDNA before, the free plan is genuinely the right starting point. You can run the Rentalizer on your property, check the Competitor Calendar, and look at your market’s occupancy data — all without paying anything — and evaluate from there whether the paid tier adds enough value for your situation. See our full breakdown of whether AirDNA is worth it for different host types and market sizes.
Frequently Asked Questions
How accurate is the AirDNA Rentalizer specifically?
The Rentalizer uses comparable active listings to estimate a property’s annual revenue. In large, data-rich markets with strong comps it produces estimates that are generally directionally reliable — often within 10–15% of actual performance for a well-operated property. In small markets or for premium properties without many true comparables, estimates can vary more widely. Always check the confidence score AirDNA displays and refine your comp set manually for the best results.
Does AirDNA use real booking data?
Partly. AirDNA scrapes publicly visible listing data from Airbnb and Vrbo daily and also collects verified booking data from over 1.1 million partner properties connected through channel managers and property management systems. The partner data is 100% accurate based on real reservations. The scraped data uses a machine learning model to estimate whether calendar dates are booked or blocked — which is where some variance comes from.
Is AirDNA accurate in small markets?
Less so than in large markets. With fewer active listings to pull comps from, AirDNA’s estimates carry more variance. Market-level data like seasonality patterns and ADR trends is still useful, but individual property revenue projections in markets with under 100–200 active listings should be cross-referenced with local knowledge and property manager estimates before being used in investment decisions.
Does AirDNA overestimate or underestimate revenue?
The direction of the error depends on the property. Premium properties with standout amenities tend to be underestimated because AirDNA averages across all comps rather than top performers. Properties in markets with a wide quality range may see estimates that reflect average performance rather than what a well-run, well-photographed listing would actually earn. Hosts who outperform their market typically find AirDNA’s estimates are conservative relative to their actual results.
Can I trust AirDNA data for investment decisions?
For market-level research — whether a market has strong STR demand, what seasonal patterns look like, how competitive the supply is — yes, AirDNA is a reliable tool. For property-specific acquisition underwriting, treat AirDNA revenue projections as one data point among several. Cross-reference with local property manager estimates, actual comparable listing performance, and your own market research before committing capital.
Related Articles
- In-Depth AirDNA Review — full breakdown of every feature, pricing, and who it’s for
- Is AirDNA Worth It? — when it pays for itself and when it doesn’t
- AirDNA Promo Code — the only real discount and how to get it
- Best AirDNA Alternatives — free and paid options worth knowing
- AirDNA vs Mashvisor — which data tool is right for your situation
See Your Market’s Data Before You Pay
AirDNA’s free plan gives you real occupancy rates, the Rentalizer, and competitor pricing for your market. No credit card required — check the data quality yourself before deciding whether to upgrade.
Try AirDNA Free →
