Feeld Algorithm
Optimize interaction + Engagement
Just like any other algorithm, the Feeld algorithm can be influenced through optimization. The single most important factor is getting positive engagement signals from other users [likes + messages]
Read to learn all the visbility factors.
“I asked Tinder for my data. It sent me 800 pages of my deepest, darkest secrets”
Dating apps, including Feeld are incredibly good at collecting data from you. You therefore want to make sure you send the best possible signals, in your profile and in your messages.
Key takeaways
Our findings:
The Feeld algorithm isn’t as nearly as developed as Hinge or Tinders, and relies heavily on positive user signals + matching desires + activity + location.
Positive User signals: This is simple, the more people write to you/respond to you/like your profile, the more frequent the algorithm will show your profile.
Desires: Obviously, people have very different “kinks” & “desires”. Some people would like to experiment with couples others aren’t interested in that at all. The algorithm tries to match you with people with similar desires, so make sure you get those right.
Activity: Feeld is still developing and has a fraction of the amount of users Tinder does, it therefore needs to prioritize active users to create the best possible experience.
Trust is a major issue on dating apps in general, any suspicious signals the algorithm picks up and your visibility can potentially be affected, worst case a so called “shadow ban”. [It stops showing your profile entirely].
Location: This speaks for itself, but from what we’ve seen, London has the most active users. If you switch your position to London and still don’t get likes, some of the factors above need to be fixed. To become successful on feeld you’ll most likely have to update your profile and up your messaging skills.
The foundation of dating algorithms
Dating algorithms rely on machine learning based on scanning profiles, user behavior, user activity, and user engagement. What does this actually mean? It can be broken down into positive and negative signals, and also personalized signals.
Examples of positive signals:
A lot of people swipe yes on your profile = A boost in visibility.
A user is active daily on the platform = a boost in visibility [no app wants inactive users]
A user seems to get a lot of responses when chatting with people = a boost in visibility.
I’m in “top picks” on Tinder and “standouts” on Hinge, where only the people with the highest visibility end up, learn how I do it.
Examples of negative signals:
People leave the chat/ unmatch after you have interacted with them = a decline in visibility.
People don’t reply to your messages = a decline in visibility.
Your match rate [matches per 100 swipes] is really low = a decline in visibility.
Personalized signals
Dating algorithms use machine learning to figure out “your type”. It tries to understand based on your swiping preferences who could be a good match for you. It will also scan the information you’ve inputted in the app to use as support.
The Goal of the algorithm
The best way to understand an algorithm is to reflect on the app’s goals. Obviously, Feeld wants to make money like any other tech company, but in order to do so they need to create a good experience for people to convince them to stay on the app.
So, how does Feeld create a good experience for their users?
Through relevance. If you go on the app and feel like the people appearing are far from what you find attractive, the algorithm has failed its job. Taking your “desires” into account is therefore crucial. And quickly picking up on your “type” based on machine learning is also key. There are, however, one major thing that sometimes makes the “fish pond” seem very small with limited people to choose from, and that is obviously location.
Location influences The feeld algorithm
The app gives the possibility to “explore” different cities. London and Barcelona seem to have the most active users from my personal experience. Give those a try to evaluate the quality of your profile!
so, how does the field algorithm work?
The Feeld algorithm relies heavily on user signals + matching desires + activity + location.
