How skewed is AirDNA???

Question

Hello,

I am having a hard time running my numbers for STR investments, because I don't currently trust AirDNA. Depending on the neighborhood or city, I am finding some significant revenue pockets that make the number work easily. A lot of the top producing comps have 90% occupancy, or no reviews, or terrible listing pictures etc. All of which make me believe the info is wrong.

What is your best advice to sort through anything that is skewed on AirDNA? Things to look for in the comps, people to talk to, etc.

I would love to hear some people projections vs. actual revenue comparisons.

Sarah Kensinger · Answer

Reach out to some of the coaches in the STR Secrets Mastermind. They have been using it for years alongside their properties. I know there's some lenders on here that use Airdna for DSCR loans, so maybe they can give you a peace of mind with their long track record.

For some pointers....

--Airdna is proven to be the most accurate real time data software and the algorithm is constantly refining and making sure they pull the right data. If you want the most accurate numbers, do your research on a late Tuesday or Wednesday since data is updated every Tuesday.

--The following comps would not be the best to use...those with no reviews, brand-new listings, a listing that has two comps (you can thank the hosts that won't use a PMS for that :)), and listings with Airbnb are the most accurate since Airdna can pull data off them much easier then VRBO. I usually avoid the comps that only have a VRBO listing.

--Depending on the location, seasonality of the area, and amount of STR listings for that specific zip code, it can skew the numbers a bit. For example, people get confused in highly regulated areas saying the numbers are just so good. Yes, they usually are, but it's because there's a demand for STR but the supply has been capped due to regulations or an outright ban. Seasonality can really mess things up, because most people shut down in the off-season. So naturally that would make the projected numbers to high. Also, if there's is a limited number of STR in a popular tourist area, they get booked no matter how bad the listing looks! I love traveling all over the west, but the hosts out there are very guilty of this and since it's a limited supply it doesn't matter, they make plenty during the summer vacation months.

--There can be pockets of REALLY good areas in a large metro area or even near national parks. Sometimes all it takes is crossing a major 4 lane road or moving a couple miles in a different direction. Proximity is huge right now for guests and the numbers are proving that.

--Cleaning fees are included in the projected numbers of a property. The OTA drop all the money into one bank account, then the host disperses it out to cleaners, etc. So, when Airdna is pulling those numbers, they can't separate the cleaning fee from the income.

That's about all I can think of right now, hopefully that helps. If you want to learn more about the innerworkings of Airdna here is their Youtube Channel. I have sat through many live webinars, some have even been with a founder, were they explained how they get the numbers they do and we were able to ask questions. Unfortunately, I can't share the replays of those, but this not a fly by night software that is pulling numbers out of thin air. If you learn how to use Airdna, just like any other software it can be a great asset.

Collin Hays · Answer

I look at Airdna like Car Fax.  It's a cute data point, but really doesn't tell you much.  The best way to evaluate the financial viability of a STR is the old fashioned way:  Comparing rates, calendars, and amenities of similar properties in the same/similar location.  You can do this easily on VRBO.  It probably would take you half an hour, and you would feel much more confident of the data versus some software program.

Andrew Steffens · Answer

I compare it often to Zillow Zestimates - Sometimes they can be pretty spot on if there are a lot of very similar comps in a tight area.  If you look at a cookie cutter subdivision all similar homes, amenities, etc Zillow is typically pretty accurate.  Go to a neighborhood where homes homes vary and the Zestimate becomes less accurate.  Similar with AirDNA.  Also are you using the premium version which allows you to see the comps or the free version where you just get a number?  We use the premium version and then analyze the comps to see if the number is valid or we manually adjust it for our client projections.  I hope that helps

Michael Baum · Answer

Hey @Jordan Campbell, so AirDNA is a real mixed bag. I used it a bit very early on and found it way off for our market.

When you see something as anomalous as you describe, there is usually a reason. Maybe the demand is very high and the supply is very low. Then people can get away with less than stellar listings. I think that is not the reason.

You have to check and see how many rentals there are in the that area. If there are a ton, then you need to dig deeper. If there are only a few, then AirDNA gets very sketchy.

Our area has only a few STRs compared to somewhere like the Smokey's. Thus our data was skewed very high compared to realistic nightly rates.

Your best bet is to use AirDNA or other algorithms as a base but verify with direct looks at VRBO and AirBNB listings. What are similar properties nightly rates, amenities and occupancy?

There is no way to guarantee success, but with some research, you can get pretty close.

Michael Klein · Answer

Hi @Jordan Campbell

AirDNA is directionally right, but it's not accurate enough to use directly for underwriting.

If it seems off, it probably is. If it seems right, it's probably directionally correct but can still be off +/- 10% or more. With STRs you're actually trying to find the outliers so it's just a starting point for research - are homes doing well, what is 90th/75th/50th percentile performance, etc.

Before making an investment, you should have a point of view on the market and occupancy, which doesn't vary drastically e.g. in my market 5-6 BRs normally have a ~50% occupancy.

Then find ADRs for similar homes as your target and create a range around low and high performance.

At the end of the day, you are buying real estate. You can't just take data, put it in a model and make a decision. You have to look at it holistically along qualitative factors, risks and opportunities.

Guests don't just put sleeps 11, 7 beds, hot tub, and pick the cheapest one. And the same is true about making STR investments.

Happy hunting!

Ryan Moyer · Answer

I've compared some of my properties to the numbers that AirDNA, Rabbu, and STR Insights claim they're making.  AirDNA is the closest, though still typically off by about 10%.  Small sample, of course.

Property 1Actual Rev: 155kAirdna: 140.4kRabbu: 89.5kSTR Insights: 138k

Property 2Actual Rev: 115kAirdna: 105kRabbu: 78kSTR Insights: 91k

This is not the projections for each of these if I put in the address (those are even more inaccurate).  These are what each of those platforms believe the TRUE REVENUE of my homes is.  IE the amount they list them as making when using them as comps for nearby properties.

Zachary Cain Humphrey · Answer

What I would do is look at comparable bed / bath count properties on air bnb itself, see what their average nightly rate is per month by filtering through their calendar each month for a year out. Then figure out average occupancy rate percentages for each of those months. Then multiply your ADR you figure up for each month by the occupancy rate (ex. 239$ adr x 15 nights booked or 50% occupancy= 3,585$ revenue for that month). By doing this you get real time results that are based on averages. If you feel an average you are using is high (example, the data tool says the median listings are doing 75% occupancy for the month of october) maybe consider projecting in your revenue calculations a 65% occupancy number to be safe.

Ultimately, be sure your looking at comparable properties to your property that have similar design, amenities, etc so that you get accurate averages for your ADR and occupancy rate projections.

I have found this to be a great way to project yearly revenue and monthly revenue without using or depending too much on data tools.

Sean Reichert · Answer

How does everyone analyze using the enemy method?  I've played around with AirBNB and VRBO.  I've found VRBO to be most helpful as I can use the calendar to understand bookings & pricings, and also use reviews which lists when and for how long the guest stayed at the property.  But I'm concerned I'm missing important data because it seems as though AirBNB is the more often used booking site for the Lake Harmony/Split Rock area.  Any insights?

Alyson Gordon · Answer

AirDNA is a great data source but double check the comps in the area. You may want to remove any outliers and also go on Airbnb to double check any rates for the next month, then 3 and 6 months out to get a feel for seasonality.

In my experience those listed on VRBO only it always skews high so I usually through that out as well.

Once you've controlled for these things you'll have a great comp set. This is assuming you're using the paid version, the free version isn't specific enough to inform a good purchasing decision IMO.

Zach Edelman · Answer

Could be a good idea to seek another research tool and compare and contrasts if the costs allow for it. Some suggestions are Rabbu, STR Insights, and Revedy.

How skewed is AirDNA???

Most Popular Reply