comparing occupancy data sources

[Tiffany Shan]

https://docs.google.com/spreadsheets/d/1Fqrb6yrmTn...

I'm in the middle of comparing Everbooked, Mashvisor, and AirDNA (may add Airnalytics). Spreadsheet is brief now but I will add about a dozen homes this weekend so you can use your own judgment and decide whose estimates seem most reasonable to you! Already I can tell they're quite seriously deviating. 

They're all quite expensive (pricing in spreadsheet) so I think the greatest FREE benefit from each is this:

Mashvisor: Can see their estimated annual revenue for homes without paying

AirDNA: Can see estimated occupancy by performance percentile by city (i.e. what an average home's occupancy is, what a top 25% home's occupancy is)

Everbooked: Can see estimated occupancy by size of home, by month, by city (so pretty granular!)

If you're trying to evaluate a potential home, I think you can triangulate pretty well by using all 3 services. If Mashvisor & Everbooked's predictions indicate a home would have to be a top-performer according to AirDNA, I usually downgrade my occupancy or ADR projections. I want to assume all potential homes will be average performers, and that way if they do perform exceptionally (or rather, if I as a host perform exceptionally) it can be icing on top of the cake.  

To me, the next interesting step is coming up with a systematic grading system and running the contents of the MLS through it periodically, so I can spend my time evaluating the top 10% of opportunities.

[Tiffany Shan]

@Jon Crosby that's an awesome idea, I'll do that. 

What I was *told* by these platforms is that they get manual info from actual vacation rental owners who use VRBO, Homeaway, etc. on revenue and occupancy, then match up that info with their AirBnB scraping to figure out what the delta is and apply that to all homes. 

I like that method if you can indeed get a lot of that data.

Also, it's actually up in the air how much scraping of VRBO/Homeaway that these platforms perform, I didn't get a straight answer the first time I inquired (which is to be expected, this is fairly proprietary info) but I will dig further. 

@Jon Latorre I just take the algorithmically generated stuff and run it by local property managers to get their input, which isn't a very efficient process, but like you said, is very useful.