the head of InstagramAdam Mosseri, published yesterday, Wednesday (31), a new article to explain to users of the social network how its classification algorithms for displaying publications on the platform work and also spoke about “Shadowban”.
Initially, the head of the social network said that the platform does not have just onealgorithm that oversees what people do in the app, but there are multiple systems that are responsible for different areas of the app, like the feed, Explore, Reels, Stories and surveys.
Mosseri said that each algorithm uses a multitude of signals to determine how content is displayed to each user. For example, the posts shown in the feed are determined by past activity, as well as interactions and engagement.
The classification of Stories is determined by the frequency with which the user views the publications of a profile and also takes into account the engagement in these posts, such as sending a like or a DM, in addition to the probability that the user is a friend or relative of the person.
In turn, the recommended content in Explore follows a slightly different line, as they are based on posts that have been liked, saved, shared or commented on in the past, in addition to related posts that the user has never interacted with.
Reels are also shown based on user activity, interaction history and according to the popularity of who posted or the audio track and visual resources of the video, for example. All this seems simple enough.
At the end of the article, mosseri commented on the so-called “shadowbanning”, which is a term not oficial used by creators who feel their accounts or posts have suddenly had more limited exposure or are hidden on Instagram without any warning.
The Instagram Executive did not confirm that the so-called “Shadowban” exists and recommended that users keep an eye out for the account status in the application to know if your profile is considered ineligible for recommendations on the social network.
Finally, mosseri highlighted that Instagram is testing new notifications to help creators understand when the reach of their content may be limited due to a watermark (from TikTokfor example) or for not meeting community guidelines.
