Once in a while I use AirBnB. There are a couple of features that I (intuitively) use to judge if an apartment is save to book; ratings, images of the flat and the user avatar. Apparently, these avatars play an important part in the overall service and usage of AirBnB.
A recent study finds that “Attractive Airbnb hosts are more likely to get bookings, even with bad reviews”.
With the easy availability of image recognition services, even the everyday researcher can do a small analysis.
A friend provided a small sample of 200 AirBnB avatars which I ran through Microsoft’s face recognition API.
Of the 200 images, 117 were labeled as faces, for these faces the API provides a couple of features:
e.g. age, gender, smile detection, mustache detection and the size of the face.
So let’s have a look at the aggregate demographics of our (probably unrepresentative) sample of AirBnB hosts. 62% are male, with an average age of 37.5.
The age distribution is shifted upwards for the male population.
The difference is substantial and significant. Male AirBnB hosts are almost 7 years older than female hosts.
Let’s have a look at the emotion expressed in the avatars. Are female avatars more likely to smile?
Contrary to my base assumption, this is not the case.
Another feature that might be worth exploring is the size of the face rectangle. It is basically a measure of the face to image size ratio.
It seems that guys are more likely to use close-ups as avatars. Even though this study is very limited in the features and sample size used, I can already see the potential of image recognition as an implicit data source to improve e-commerce services.
I wonder if and how AirBnB uses face detection.