Instagram chief Adam Mosseri has shared a range of videos on his Instagram profile which aim to demystify the algorithm, and explain why and how the platform ranks content.
The aim of the video series is both to address concerns about algorithmic influence on content trends, which the Australian government is looking to curb with new regulations, while also giving creators insight to help them improve their Instagram posting approach.
Most of the insights Mosseri has shared have been pretty generic, and not particularly insightful, especially for long-time social media managers. Though it is worth assessing the detail from a social media marketing perspective, and taking note of Mosseri’s insights, which could reveal a new perspective or consideration.
Based on the information shared in Mosseri’s algorithm overviews, the key points made the executive are:
- Instagram uses algorithmic sorting to help users see the most relevant content, and make the most of their time in the app.
- If Instagram doesn’t rank content, the feed gets overrun with content from accounts that post the most, “and business and brands post a lot more than creators and people,” leading to a worse user experience.
- Without content ranking, people see less content overall, and the average reach goes down for creators.
- Algorithm ranking is based on systematic predictions of how users will respond to a post, including whether they’re likely to watch it, share it with a friend, etc.
- Recency matters a lot in content ranking, which is why the feed is “roughly in chronological order.”
- Creators looking to increase their post reach should focus on engagement rates, as opposed to total views or likes.
Mosseri also noted that there are four principles that drive Instagram’s algorithm direction:
- Originality – Instagram tries to give more precedence to content that’s original, and from the original creator, as opposed to re-posters.
- Recency – When something is posted matters a lot in terms of ranking and display, with more recent content getting greater exposure.
- Breaking content – Mosseri said that Instagram will show content from small creators who post publicly to accounts that don’t follow that creator in order to get a better idea of how people respond, with a view to boosting posts from new creators.
- Model size – Instagram has expanded its AI models to take in more factors from a broader range of posts to improve its ranking.
Mosseri also noted that users can influence what they’re shown in their feed by using direct feedback tools, like the favorites list, the “Not interested” indicator on posts and Your Algorithm, which enables users to add or remove topics that the system has attributed to them as interests.
Mosseri also advised that creators should experiment with different content and approaches, because trends change, and what’s interesting one week may not resonate the next. Mosseri suggested that trial Reels can be beneficial on this element.
Per the executive, creators look to post as much as they can to increase their chances of having a hit, while also managing the time they have to avoid burnout.
