How accurate is AirDNA for STR revenue projection?

Question

Hello BP family,

I wanted to get some insight from other STR hosts that have operating revenue numbers. I currently have 1 STR in Palm Bay FL and I am trying to build a system for accurately predicting what revenue we could be generating for the next property we are searching for.

I think the AirDNA rentalizer tool is awesome, but I'm curious as to its accuracy.

I've compared our actual numbers to what it projects for our property, and I found that AirDNA's projected occupancy is 67% compared to 76% that we are actually seeing. In addition, the projected average daily rate (ADR) is $244 compared to $259 that we are actually seeing (a difference in monthly revenue of about 20% underestimating.

Now this is still pretty close, and I know that it is predicting the "average" property performance based on historical data. I am just curious if other hosts have found similar accuracy or if it is way off in their scenario. Please post your properties AirDNA numbers as well as your actual numbers if you have them and are willing to share!

Zach Edelman · Answer

So much comes down to the host, but I will note that we, a lender, compared twelve months of actuals against Air DNA, and do so whenever we underwrite actuals, and have found it to be at times scarily accurate.

Trent Reeve · Answer

i dont use 1 tool, ever. also i dont only get 1 estimate when buying something. I also use Rabbu, Awning, and enemy method to evaluate. also if you didnt know, AirDNA also includes cleaning fees into "revenue" projection, so keep that in mind.

Samuel Boyd · Answer

Hey Trent. I totally agree with getting multiple estimates. I'm going to look into Rabbu and Awning for sure! Are those similar services to AirDNA? and good catch but yes, I know AirDNA's numbers are inclusive of cleaning fees.

Samuel Boyd · Answer

That's the answer I was hoping for! Obviously yes, every valuation method has its limitations as everyone has mentioned, but as a first step quick and dirty analysis I feel that it is close enough. I'm glad to hear you are finding the same in your research! Very valuable insight Zach much appreciated man!

Sarah Kensinger · Answer

Airdna is the most accurate real time data software, but it's nice to use a couple different data softwares. Not only are they built specifically for pulling STR data (and not just an add on bonus), but the algorithm is constantly changing which helps to make the data more accurate. Also, you do have to make sure your following a few guidelines regardless of what software you are using. I have found the rentalizer to be pretty close but yes it does run a bit low, better then to high that's for sure! And of course, if a property is on more the Airbnb and VRBO that revenue probably won't be showing on the data sites.

Michael Baum · Answer

Hey @Samuel Boyd, I think you are getting the gist of AirDNA. It is just an algorithm that pulls the data it can in order to come up with the "GUESS".

It is just an estimate. You are never going to get perfectly accurate on your estimates. I guess that is why they are called estimates. :)

Combine the AirDNA data with the Enemy Method. Look at all the properties nearby that offer similar size, location and amenities. See what they are doing.

Combining these data points will get you closer, but it will never be exact.

To me it looks like you beat the estimates on your current property so I think you can feel confident that you can do it again!

Jon Martin · Answer

I find it is reasonably good at pulling past performance data. Predicting that forward is something that no platform or algorithm can do. Far too many factors.

Even with a rentalizer estimate, that estimate is only as good as the comps it pulls in to make an average from them. If the comps are relatively similar to the quality and location of your listing, then it is reasonable to assume that yours would have performed about as well. However if it is pulling in comps that are clearly in a better location, better view etc then you have to take that into account, and same goes on the other end if it's pulling in dumps.

I like to run estimates based on one less bed/bath than the property I am considering. If those numbers work, then it's reasonable to assume that it should work at least as good with an extra bed/bath, nice decor/furnishings, service etc

Alex Scattareggia · Answer

This question gets asked a lot and there is no 100% right answer.

As a lot of people have already stated try to get as many data points as possible which is correct.  Depending on how far back you are going I would actually just throw out 2020 and 2021 as reliable indicators because of how they will positively (and innacurately) skew the data.  Those years are probably closer to anomolies than predictive of the future.

Take all your data points, shave 5% off the top and create an average and you will have a pretty safe estimate.

Andrew Steffens · Answer

We use AirDNA but it is important that you do not take the generated number at face value.  We have a premium account which pulls the comps and allows me to adjust the projection based on the comps in the area.

Think of it like a Zillow estimate.  Sometimes it is very accurate.  Sometimes it is very inaccurate.

Samuel Boyd · Answer

Hey Michael, thanks for the advice man! Looking at comps via the enemy method is ultimately the deciding factor what how well can we actually do, so totally agree. Still trying to figure out a way to tell what other hosts base price is. If they are dynamically priced, simply looking at a month of numbers might not mean much about the average on a yearly time scale because of seasonality. But even this I know you can account for with other tools. I think Pricelabs actually has a pretty good "comps" set up where you can monitor performance of others but I need to dive into that further to be sure.

Thanks again!

How accurate is AirDNA for STR revenue projection?

Most Popular Reply