Does restoring the social feed really change anything? Where did the interesting content go

For some time now, many social media users – especially younger ones – have been aware that the real purpose of recommendation systems on these platforms is not so much to suggest interesting content, but rather to keep people “glued to the screen” for as long as possible. In short, social media platforms compete daily for users’ attention, just like news outlets, streaming services, and many other sources of entertainment. The key difference, however, is that social media platforms do this regardless of the type of content: from cooking recipes to dance videos, tutorials to breaking news – all mixed together in the feed without apparent order, in a continuous stream designed to encourage so-called doom scrolling, the passive consumption of large quantities of content. Yet, this constant exposure to wildly different stimuli can lead to a kind of numbness. It’s a mechanism that often triggers a vicious cycle, where users, more or less unknowingly, are exposed to content they find engaging, but which can sometimes be harmful (even if that’s not immediately clear). In some cases, recommendations can even have negative effects, especially in terms of mental health, particularly for young and vulnerable people, but not only. Another consequence of the recommendation model used by major social networks is the tendency to standardize post formats, making them look increasingly alike even when covering very different topics, with all the resulting, sometimes indirect, consequences.

Social media algorithms recommend content based on various factors, including a more or less accurate estimate of the user’s interests, as well as each post’s ability to attract attention, generate engagement, and keep people on the platform longer. To decide what to show, platforms analyze users’ past behavior and continuously update their suggestions, taking into account likes, time spent on specific content, saved posts, followed accounts, and similar elements. Moreover, suggested content almost always comes from accounts that the user does not follow or interact with regularly, but from third-party profiles, sometimes designed specifically to appear in the feeds of platforms like Instagram and TikTok. The fact remains that more and more users are realizing how these recommendation systems work, and many are starting to find them not just annoying, but problematic. Unsurprisingly, techniques to take a break from doom scrolling are spreading on the very same platforms. Tech companies have been aware of the issue for some time, and some have introduced the option to “reset” the system that determines what is shown in the feed.

This function has been interpreted as, at least partially, a response to users’ frequent requests for more control over the content they encounter, in an attempt to achieve a healthier browsing experience. In practice, it’s a setting often available within the individual social media apps that allows users to delete the data that the algorithm has been using to select content. Resetting the algorithm is a bit like “retraining” the system. But even those who enjoy suggested content can benefit from this feature: for instance, it can help reduce the influence of outdated preferences, which algorithms tend to keep considering, and allow for the discovery of new areas of interest. One immediate effect of the reset, which, on Instagram for example, is irreversible, is that, at first, the suggested content may seem almost completely random. However, users shouldn’t be alarmed; it’s as if the system is starting from a “blank page”: it has to relearn what the user likes, which is why, at the beginning, it often suggests random content.

For some users, making their feed less appealing is also a way to limit the time spent on social media, avoiding the more drastic step of deleting their account. “Resetting suggestions is just the latest in a series of Instagram features aimed at personalizing the user experience. Among them is the option to flag content you’re interested in within the Explore section,” notes Wired Italia. Still, resetting the algorithm does not mean erasing user data: this information is still used by platforms, for example, to personalize advertising. Moreover, a reset does not automatically improve the feed: it may help break out of certain “bubbles,” but the risk of falling into new ones (which may also be controversial) remains, because that’s how social media works today.