A user might decide to stop pressing the like button to reduce interaction with social networks or to distance themselves from content that appears on their screen daily. At first glance, it seems that by removing this reaction, one of the most important ways of signalling preference to the platform is also lost; but recommender systems are far more complex than relying solely on likes.
Social platforms typically consider a set of behavioral signals to determine users' interests. The type of content the user views, the amount of time they spend on a post or video, searches they perform, accounts they follow, and content they share with others can all play a role in shaping their recommendation list.
What change does stopping likes create?
When a user no longer presses the like button, a direct signal about the content they like is reduced. A like is usually an explicit reaction that shows the platform the user liked a particular piece of content and can be used in the content recommendation system.
However, removing this reaction does not mean the user's behavior becomes anonymous or that all activity-related data is deleted. For example, watching a video to the end, returning to the same video again, a long pause on a post, opening the comments section, or visiting the publisher's page can all provide information about the user's level of interest or attention.
As a result, even if the number of likes recorded by a user drops to zero, the recommendation system can still form a picture of their interests based on their other behaviors.
Even the amount of time spent on the screen matters
One important factor in today's content recommendation systems is the amount of time a user spends with a piece of content. A user who pauses a few extra seconds when viewing a particular topic may send a different signal than a user who immediately scrolls past the same content.
Thus, even the simple action of scrolling can contain information. Content that a user passes without stopping can be different, from the recommender system's perspective, from content they stop to read, view comments, or open its page.
Of course, the importance of each of these signals depends on the platform and the design of its algorithm, and recommender systems are continually changing.
Searches and following accounts also reveal your tastes
Likes are not the only way users show interest in a topic. Repeated searches for a topic can be a sign of interest, and the accounts a user follows provide another clue about the subjects that attract them.
Sharing content and repeatedly watching a particular type of video or post can also influence future recommendations. Therefore, a user may never like a video, but if they regularly watch similar videos, they will still receive more content from that category.
For this reason, simply not liking cannot be considered a definitive way to prevent content personalization.
How can the influence of algorithms be reduced?
If a user's goal is to change the content suggested to them on a social network, there are other methods. Hiding unwanted posts, unfollowing accounts that are no longer of interest, and using options like "not interested" on platforms that offer such a feature can help refine recommendations.
Checking privacy and personalization settings on the account is also important. Some services allow users to delete search history or part of their activity or to change how their data is used for personalizing content. Of course, the available options in this area vary from one platform to another.
Ultimately, stopping liking is more like removing one piece from a large puzzle than severing the connection between user behavior and recommendation algorithms entirely. How one views, searches, scrolls, follows, and interacts with content can still affect the experience the user receives on a social network.
Therefore, the important question is not only what happens when you don't press the like button; it is also necessary to ask which everyday behaviors-those we don't even consider "interactions"-platforms use to learn our interests and adjust recommended content.