Company presentation: Tinder, a data company?
Every Saturday we present you a company and its way of using data. Today, a company that is making a lot of noise: Tinder.
Quick overview of Tinder
Tinder is an app launched in 2012. It is a dating app with an estimated 2019 revenue of around $810 million. Tinder describes itself as the best free dating site. The app also has a paid offering. The app claims more than 50 million active users and 55 billion matches since the app's inception. The app has 3 paid offers: Tinder Plus, Tinder Gold and Tinder Platinum.
The use of this application is simple. People are presented to us. We swipe right if we want to meet them, left if we don't want to meet them. If both people swiped right, there is a match and they can talk.
A big part of its success is due to its use of data! Let's see how Tinder uses data.
Data usage at Tinder
Like any platform, one of Tinder's goals is to get its users to stay on the platform.
What data is collected by Tinder?
To better understand what Tinder can do with data, here is the data Tinder has on you. If you have a Tinder account, thanks to the GDPR you have the right to ask any site or app for the data they have on you. Here is the list:
- Your current profile pictures and videos. So far it clearly makes sense, just to be able to display them.
- Gender, Name, Sexual Orientation, Biography, City, IP address, Account creation date
- For each day: number of times you swiped and in which direction, number of matches obtained, number of times the application was opened
- The history of all the messages you have exchanged
They have a lot of elements, but ultimately nothing surprising. If you want to analyze these results in depth, Tinder insights gives you an analysis. We learn for example that 33% of women who swipe right have a match while this figure is only 2.5% for men.
What does Tinder do with this data?
It will suggest the people who correspond to you the most. For this, it will use NLP (natural language processing) to analyze your biography. With computer vision, it will analyze your photos, to understand if you smile, if you do this or that activity or for example if you wear sunglasses.
Tinder will then use this information to suggest the most relevant profiles to you. Your profile will also be proposed to others based on these criteria. A very important basic criterion will be geography and age proximity. Note that your profile may not be proposed very much if :
- Your account is not very active
- You always swipe on the same side. Swiping right all the time will not maximize your chances.
Tinder may suggest profiles that resemble, for example, the previous matches you've exchanged with the most.
Tinder is a data-driven application where everything is consciously calculated to maximize the chances of meeting.
In its early years, the Tinder application worked on an ELO-like system, like in chess. If you matched with people with a high score, then your score would go up, and you would be introduced to people with higher and higher scores. On the other hand, if you did not match with people with lower scores, your score would decrease more and more. The application now attests that it no longer uses this type of scoring.