Tinder Analysis: User Experience, Algorithm and Psychology (2)

How does Tinder's algorithm work?























Have you ever wondered who's gonna be shown to you?



Tinder's Objective: increase the matches as many as possible




















In the first article, we talked about the user experience of tinder. One of the unique design is the card-swiping game: focus on one project at a time and match is a powerful concept. Which makes people feel excitements and addicted to it. That's why I would like to discuss more about the algorithm behind Tinder, or for my personal curiosity.

First of all, you have to understand what's Tinder's objective that people will be willing to use this app either make friends or get into relationships. Before that, we have to get as many matches as possible, right? Then, we could continue to talk about the further steps.



Tinder users' pool






























Imagine, there are tons of users are waiting for matches. No matter, they're cute, good-looking, talented or hot.
Just pretend they are in a "Tinder's pool", ready to line up.

The icons represent they like you or not.



1. First basic rule is that people they like you "must" be shown to you.

For example, if you like Simpson, you are expected you're matched. Therefore, Tinder has to make sure people who like us must move up to the top of our list. However, the one who like you does not mean you like them too. What if you swipe left to all the people who like you. We could not get any matched at all, that will ruin the user experience.

It's not a simple math problem. We have to take into account lots of situations.

Who is or Who should be prior to whom?

It should be fairly assigned, you do not want to see the people you're not interested for a long time; the people make you feel attracted to should scatter in to your queue. On the contrary, we do not want every time we swipe right, we got matched. It might be too boring, not challenging. The interesting part is that you do not know if you're matched or not. This key user experience create the a-ha moment significantly.



Tinder Algorithm: internal rating based on suggested factors 






























After using Tinder and reading several online articles, I try to conclude in this way:

There are three main factors composed of the Total Tinder score which could decide the order
1. Attractiveness
2. Activeness
3. Pickiness

The size of circle might be the significance in the formula.
Each factor are affected by several subfactors shown in the diagram


*Reference:
1. Quora: How does Tinder Algorithm work? (Highly recommend)



Tinder algorithm: Attractiveness























Frankly speaking, it's important to show attractive people between undesirable people that make this card-swiping system be fun. Therefore, it is critical to provide an "attractiveness" score based on your profile(mainly the first photo) to every user. In my view, there might be several subfactors to determine the score shown in the picture. (feel free to discuss with me, I'm happy to hear about your opinion)


The app seems to be able to show you attractive women/men at a much higher rate then normal dating apps which makes me think that there is a team working at Tinder whose only job is to rate profiles by their attractiveness (or lack of) so you spend more time using it.

Why?

1. Fairly assigned people, Excitement Feature

2. Focus on New Users' UX

It was mentioned in several articles that in order to let new users hooked on the app, Tinder show new users more attractive people; also, the algorithm tries to find matches right away in order to give new users positive reinforcement.


Tinder algorithm: Activeness
























It basically tracks how active you are using the app, and pushes your profile closer to the top of the pile. Tinder needs to make sure it's a real person or it's an active person. Or the matches could not get further steps. 

Maybe you have noticed that if you don't reply to your messages, you seem to not have any more matches. The algorithm has to discourage you when it comes to using the app and only keep users who are active and recurrent, keeping the ones who don't act upon their matches away.



Newness of account: 

The longer you stay on Tinder, generally the less matches you will get. This is because you are old news according to Tinder. They like fresh faces and new users, who get max exposure when they first sign up.


Besides, deleting your Tinder account and then immediately logging back in to your Tinder account will refresh the pool of matches that you can potentially view – just like starting fresh.



Tinder algorithm: Pickiness























Tinder also do not want people who are too picky that users could not get any matches, even though the people have a high "attractiveness score". On the other hand, you should not always swipe right, then you will be top of everyone's lists.

To prevent this kind of situation, Tinder limits the number of likes a user can make in a consecutive 12 hour period in Sep 2015. Tinder has already seen a dramatic increase in the quality of matches.

After introducing this feature in a week, Tinder has seen a 25% increase in the number of matches per right swipe, a 25% increase in the number of messages per match, and a 52% decrease in spam bots (another of our biggest user complaints).



Therefore, here are the visualization of the algorithm:

Tinder algorithm visualization (1)























Before adding up by the three factors(Attractiveness, activeness and pickiness),
the data is dis-ordered, scattered and unorganized.



Tinder algorithm visualization (2)
























After calculating all the scores, there is an order of the pile.
Then, we could continue the matching system.

It's interesting to know that tinder now makes 30-50% of your cards as matches until you run out. If have no matches in your waiting in a queue this does not apply. In addition, Men swipe right on Tinder 3 times as much as women.

I use as many as fun photos I can to explain the algorithm in Tinder. Hope you enjoy this article!
The last article we're going to talk about the psychology inside Tinder.
Why people want to use it? why people think it's fun? why people get addicted to it?






Reference:
1. What's an algorithm? - David J. Malan (Basic insight for algorithm)
https://www.youtube.com/watch?v=6hfOvs8pY1k
2. Quora: How does Tinder Algorithm work? (Highly recommend)
6. Tinder matches made easy with new intelligent algorithm
7. Is There an Algorithm Behind Tinder?
8. What is Tinder no matches bug?


Image Resources: 
1. Elmo
4. Jon Snow
5. Jonny Depp
6. Simpson
7. Jimmy Fallon
8. Adam Levine
9. Trump
10. A woman reject a man

Comments

  1. "I use as many fun photos as I can... " ?
    滿有趣的,想到之前上演算法的課也學過類似這種配對問題(不過是很理論的角度XD)
    能夠每周寫一篇blog還滿猛的,
    期待你的更新~~

    ReplyDelete
  2. IT's very informative blog and useful article thank you for sharing with us , keep posting learn more about Product engineering services | Product engineering solutions.

    ReplyDelete

Post a Comment