The Instagram Algorithm
Do I know WHY the instagram algorithm is... like it is? No - and I don't think anybody does. But I have some thoughts.
If you’re on any social platform at all, you’ve heard - or participated in - the collective frustration with instagram and its algorithm. I am NOT here to tell you the algorithm isn’t changing, hasn’t changed, or won’t change. I have no idea about any of that - because instagram/meta does not publicize that. To me, and my overly-logical brain, that makes it a moot point. I do not care if the algorithm has changed, because I cannot effect something that I don’t understand (or, cannot even confirm exists).
THEREFORE: what can we do? Something is going on. Things have changed drastically. Huge accounts and small ones alike have seen a drastic decrease in reach for the last six months minimum, and engagement in general has nosedived. What I’ll do here is explain what we know for sure, and what I suspect is going on outside of the algorithm - because it gives us a starting point for action.
(of course, if you aren’t interested in improving instagram reach or engagement right now, which is absolutely your right, this will be useless to you! as it’s a long read, I’ll tell you now that the prescription here is more work. If you’re burnt out, don’t have enough time for writing, or any number of completely valid reasons to be disengaged from instagram yourself right now, I free you from the need to read this post lol)
What does instagram say?
Instagram says (side eye) that the algo doesn’t change. They do have a wealth of information on their website that explains how the algo works, and what they “promote”, and I’ll summarize this all here for you (before getting to my less concrete suspicions about what else is going on).
How it works
The algorithm is a series of rules (or toggles, check marks, whatever way makes the most sense for you to visualize) that determine what an instagram user sees when they open the app. The engine works a little differently between stories, feed posts, search/explore, and reels, but the general concept is the same throughout. It considers:
Activity - posts the user has liked, shared, saved, or commented on. The more they do this with the posts of a given account, the more of that account’s posts they see. Ditto to keywords, hashtags, etc - the algo knows what the user likes based on actions they’ve taken, and works to show them the same.
Information - on a post basis, they look at how many people interacted and how quickly after it was posted, the tagged location of the post, when it was posted, etc. This adds to the activity “bucket” described above for future recommendations.
The Poster - the account that MADE the post they interacted with. The poster is ranked against others the user interacts with, and the algo measures similarities to help bolster its understanding of what the user wants to see.
History - the relationship, over time, that the user has with a given poster or post category. If they always like your posts, they’ll always see them. If showing the user Hello Kitty is a 10/10 slam dunk for the algo every time, they’re going to put those posts on every corner of the app for you.
I want to quickly sidebar here to note a really important framing for the back half of this post - notice the POV of all the above decisions. It’s not what you’re doing in these rankings - it’s the actions taken by other users.