Instagram Shares How Stories, Reels, and Feed Algorithms Work

Just like how Mr. Krabs from SpongeBob guards the secret formula for the Krabby Patty, tech companies aren't really known for sharing how their algorithms work — until today.

Recently, Adam Mosseri explained how Instagram's algorithm worked in a blog post.

And we're going to break down not just one but all four formulas of how the platform ranks content.


Instagram (IG) shared its algorithms for how Stories, Reels, Explore, and content shared on the Feed work, so here's the gist for each of them.

Feed: Posts you've liked, shared, saved, or commented on helps IG personalize your Feed to find a balance between showing you content from people you follow and accounts you don't and ads too.

The predictive algorithm uses 1000s of signals to figure out how popular/recent a post is, how quickly people are interacting with it, and other info like when it was posted, and where.

IG looks at your history of how often you engage with an account's posts to determine whether it's relevant enough to exist on your Feed.

Those signals also consider the format of posts you prefer; for example, if you like photos, IG will show you more of that, and if you like watching videos more, then it's the same story. Speaking of stories…

Stories: The algorithm works similarly to the Feed, but IG ranks Stories by how often you view an account's videos and then prioritizes them.

The more you like or DM a person you follow after you watch their Story, IG will show more posts from them.

Another Signal IG uses for Stories is "Closeness." They're looking at your relationship with the poster and how connected you are with their friends and family or followers.

IG uses signals to predict what content might be relevant and interesting to you based on how often you like, reply, or skip a Story. The posts you engaged with the most are shown higher up in your Story feed.

Explore: This grid of recommended new content uses an algorithm that takes into account photos/videos you've liked, saved, shared, or commented on in the past.

IG is using signals to measure how fast and how many other people are liking, commenting, sharing, and saving posts to then show that content on the Explore feed.

So if you interact with a specific type of post, IG will show similar content from a wide array of people.

Reels: Like the Explore feed, Reels use the same signals to show videos you might like.

IG looks at how often you engage with Reels and your history of interacting with the person who posted; they have signals that can determine what the video is about visually, and if the song being used is popular or not is also a ranking factor.

When it comes to Reels, here's what IG doesn't like: Videos that are low-resolution, watermarked videos (remove the TikTok watermark!), muted or contain borders, are mostly text or focus on political issues, or Reels that have already been posted.

Check out the link above to learn more about how the app ranks/shows content and how you can curate your Feed to be more relevant to you.

But IG wants to set the record straight and said that they don't prioritize paid ads over organic content.

Want to read this in Spanish? Spanish Version >>