With the upcoming by-law to license short-term stays, there’s been a lot of chatter about the topic online. Recently, the city even released a statement addressing some of the speculation about address-finding and explaining some of the rational behind the choice of a contract with a data-finding company. With both the city and Kingston Accommodation Partners selecting the same company and the city insisting that the company is only using public data, I wondered: Just how easy is it to find the address of an Airbnb location in Kingston?
Background and parameters
I’ve worked in tourism. I’ve stayed in Airbnbs friends have booked (I’ve yet to book one myself) and I’ve browsed the site multiple times. Airbnb, in its getting started pages for hosts says “Your listing’s address will only be shown to guests with a confirmed reservation. In public search results, we only show an approximate location for your listing.” With this in mind, I set up some basic guidelines:
- I’ve got 5 traveller profiles in mind and would try to find each of them a place to stay.
- My search would be done from a private browsing window to avoid my own history interfering with the search. No logging into Airbnb.
- I searched for rooms or units available for one night, March 11 (1 month in the future), for one guest and only browsed for superhosts. My search was simply for “Kingston, ON.”
- I browsed listing using map view, focusing on an area of the city that would match my traveller profile.
- I’d look over listings and pick something that sounded appealing or charming to me and matched the profile of the traveller I had in mind. I wasn’t picky about if it was a room or an entire residence, though I did ensure the primary photo wasn’t showing a house number – I didn’t want this to be too easy.
- Once I clicked on a listing, I’d have 15 minutes to read it, look through its photos and see if I could pinpoint an address. Once I picked my location, no turning back.
- I’m allowed to use any public resource, Google Maps, KMaps and Google street view included.
- In keeping with the statement from the city, my focus will be on the address, not the hosts.
Search #1: Heart of the City
When I travel, I like staying right next to a downtown. I love the hustle and bustle, being close to nightlife and popular attractions, but usually just far enough that it can be relaxing. This profile was, essentially, me — so I picked the first listing on my search that said ‘Limestone.’
My first pick gave me exactly what I was looking for: a picture of the front of the house, street number included. The Airbnb map showed the location as Johnson Street between King Street East and Wellington Street, I dropped my street view search at Wellington and William streets to give me more residential. I clicked around, ‘walking’ along Wellington, then Earl, then King Street before finding the matching location on William. Location found in 4 minutes, 23 seconds.
Search #2: Traveling professional
For my second search, I pictured a professional coming into the city for a few weeks. I wanted something close to downtown or Queen’s University, but not right in the core. My selections in my search area were a little more sparse, but the one that said “Presidential Suite” caught my eye and I clicked it before I realized how many bedrooms it had.
Once again, it didn’t take long to browse the pictures and find the front of the house and a street number. This time, the map showed a location on Oakridge Avenue, near Johnson Street. I dropped my street view search onto the location, moved towards Johnson and had planned to search down Gibson Avenue, thinking it was my more likely candidate. When I turned right onto Johnson, though, I spotted the bushes shown in the listing and matched the address. Location found in 3 minutes, 47 seconds.
Search #3: Frequent flyer
Next, I was wanting something near the airport. I thought someone flying in may not know the distance between the airport and other Kingston amenities and that a location near where they fly in might be preferred. I picked the one with a picture of a beachfront that promised me the choice between a peaceful sunrise or sunset.
Yet again, a house number was clearly visible in a photo, though the house was far more zoomed in this time, so it was tougher to see distinct features. AirBNB showed me a location on Barnsley Crescent near Acadia Drive, and I was, admittedly, nervous that suburbia would prove a bigger challenge than downtown. I didn’t even have to leave Barnsley Cresent to match the number from the photo, though. Location found in 2 minutes, 16 seconds.
Search #4: East-side traveller
I envisioned a traveller here for RMC or CFB Kingston who might find it convenient not needing to cross the Causeway on a regular basis. With my map focused on Greenwood Park, a listing with a unique front door and brickwork stood out first. While tempting to keep my streak up, the primary photo broke my guidelines, so I chose one that offered breakfast instead.
Airbnb offered me the corner of Greenwood Park Drive and Rainbow Crescent. I did get a photo of the exterior, but I had no street number for the first time. I dropped my pin on Google Maps and went walking. As my timer hit 8 minutes, and with no sign of my house, I thought I’d try a different approach and took to Google’s satellite view to try to match the distinct shape of the front windows. With the history that the Airbnb suggested location has not been far, I kept my search focused north of Rose Abbey Drive and Martha Street. While I found lots of houses that were clearly the same builder model, none of the brickwork, windows or front steps matched my photo before my timer went off. Location not found within 15 minutes.
Search #5: Highway rest
For my last search, I wanted to stay close to the highway and pictured someone pulling off for a rest on a long trip. I took my search to the north-west of the city, both to balance my eastern search and also to include the Invista Centre in case, maybe, my traveller was someone interested in sports and hockey. I chose the one closest to Highway 401.
I was hopeful after seeing two exterior photos that I’d have a chance — but with no street address visible and the Airbnb pin located not on a street but between Sierra Avenue and Birchwood Drive, I was nervous. I started this search at Sierra Avenue and Pearl Road, looping east along Sierra before looking along parts of Frank Street, Birchwood Drive, Pearl Road and Grace Avenue. In keeping with the thought that the suggested pin location from AirBNB shouldn’t be too far, I kept north of the greenspace below Sierra Avenue. I was working my way along Hanover Drive when my timer beeped. Location not found within 15 minutes.
Using only public Airbnb photos, Google Maps, and a little time, I was easily able to identify the exact address of 3 properties — each in under 5 minutes — and while two properties went unlocated, I think I could have found them, too, if I hadn’t restricted my time. A machine learning algorithm would have made my search far easier, too. A list of “houses within 500m of the Airbnb pin made of dark red brick” could easily be generated with some AI and street view imagery, for example, and would have dramatically lowered the locations I was checking manually.
My searches were done without looking at (or caring about) who the hosts were, visiting the locations, making bookings or even signing into an Airbnb account — just as the city expects its proposed data partner to do. They were done without the use of AI to narrow down my field. They were done with public data, street view images, and a couple of maps.
So, just how easy is it to find the address of an Airbnb location in Kingston using only public data? Turns out, it’s not hard at all.