Sometimes you open an app and have the feeling that it knows more about you than you do. Algorithmic recommendation systems that sort, weight and prioritize content according to certain criteria are responsible for this. They determine which content is preferentially displayed to users, including children and young people. It is important for parents to understand how these systems work and what effects they can have.
Algorithms decide how content is displayed in the personal feed and also suggest further content in the search area, for example. In social networks, such areas have names such as “For You”, “Feed” or “Discover”. The principle is always the same: content is not displayed randomly, but selected individually. Such recommendation mechanisms are not only found in social networks, but also in streaming services, video portals and online stores. The basis for this is primarily usage data. This includes viewed content, likes, comments, length of stay, search queries, device information and, in some cases, location data.
Algorithms prefer to show content that matches previous interests or triggers particularly strong reactions, for example through surprise or provocation. Platforms want to display relevant content and keep users there for as long as possible. For children, this means that if they like something or stay with it for a long time, they will see more of it. Other topics fade into the background.
Instagram ‘s standard feed has been sorted algorithmically since 2016. A learning system evaluates various factors and decides which posts appear relevant. Among other things, previous interactions, the relationship to other accounts and the type of content are taken into account. Since 2022, users have been able to switch between a chronological view (“Follow me” or “Favorites”) and the standard algorithmic feed via the menu at the top of the app. “Follow me” shows posts from all subscribed accounts in chronological order, “Favorites” only shows selected accounts chronologically. The selection only applies to the current session and must be reactivated if required.
The system is constantly adapting. The more your child – or you yourself – interacts with certain content, the more it shapes the feed. This creates a very personalized stream of messages that often feels right for your child, but can also become one-sided.
Spotify also works with personalized recommendations. Spotify creates a taste profile based on listening behavior. Songs listened to, playlists, repetitions, times of day and duration of use are taken into account. On this basis, individual suggestions such as personalized mixes or weekly recommendations are created.
The system also compares listening habits with those of other users. Your child will be recommended music that matches similar profiles. This can open up new things, but also reinforces existing preferences.
TikTok places particular emphasis on viewing time and interactions. After just a short time, the system recognizes which videos your child watches longer or repeatedly. These signals have more influence than likes or the number of followers.
One effect of this is that content from unknown accounts can also very quickly gain a wide reach. At the same time, the video stream can narrow considerably if a topic is viewed particularly often. Sensitive or problematic content can also appear more frequently.
According to the Digital Services Act (DSA), very large platforms in the EU must explain more transparently how their recommendation systems work and, in some cases, offer non-personalized content display. These options are available, but are often not easy to find.
Your child should know that recommendations are technically controlled and not random. Discussions about algorithms help to better categorize content:
It is often difficult for children – and often for adults too – to understand why an app seems to understand them so well. This makes it all the more important to promote media literacy. This also includes talking about the fascination of social media and how technology draws attention. Trying things out together, such as resetting histories or adjusting settings, can lead to helpful “aha” moments.
