In this day and age, where social media algorithms dictate a big part of our lives, it’s no surprise that industry experts and people have started to question how the algorithm works. Now, in an effort to clarify how its recommendation system functions and levy off the misconceptions about its algorithm, Instagram executive Adam Mosseri, in a recent blog post, explained how they rank content across different parts of the app.
Mosseri explained that rather than relying on a single algorithm, which many users speculated, the content rankings for different Instagram sections like Stories, Reels and Search are influenced by a complex web of factors, with a significant portion of them stemming from user-generated data.
Factors in consideration for stories and reels
Starting with the stories, multiple factors influence the story rankings, including the frequency of a user’s engagement with an account’s updates, as well as their interactions with others through direct messages and Story interactions such as likes. Additionally, Instagram also evaluates the user’s relationship with an account, such as whether they are friends or family.
When it comes to reels, the influencing factors differ slightly, as instead of relying on interactions with a specific account, Instagram takes into account a user’s previous actions, such as likes, saves, and shares, depending on the type of video. Moreover, the platform also calculates the predictive value of indicators like video resharing, completion rate, likes, and engagement with audio.
Tackling Shadowbanning
Shadowbanning generally refers to the suppression of an account or content without a clear explanation. And after widespread speculation, Instagram has finally acknowledged this concern and announced that they are actively working to enhance transparency through the introduction of an “account status” feature. This feature will not only alert users if Instagram deems their posts “ineligible” for recommendations but also allow them to appeal the decision.
While Instagram’s transparency regarding its recommendation system is commendable, it is important to understand the intricate nature of such algorithms, as they rely on countless data points and machine learning models. Therefore, providing a simple definition is not possible. Nonetheless, gaining insight into the underlying factors that affect recommendations will empower users to navigate the platform more effectively.