Instagram’s Algorithm Uncovered

Updated: Feb 27, 2021

Author: Hanzade Durmusoglu

From: Istanbul, Turkey

We can all agree upon the fact that algorithms of mainstream social media applications are quite literally controlling our way of living, our thoughts and even our career paths. So naturally, I thought it was Instagram’s turn to be exposed for its new algorithm that allegedly helps content creators more than ever. What really piqued my interest was because of my poor internet: I was logged into Instagram on my computer, and due to my slow/nonexistent internet connection, one post wasn’t able to load. Instead, there was a small image icon on the top left with a text saying something along the lines of “girl, portrait, sitting”. That’s when I realized Instagram was, in fact, analyzing every picture to place it in its algorithm accordingly, whether it was from a private account or a public one.

photo by: mpsocial

With the reverse-chronological feed that graced us until July 2016, Instagram users were missing 70 percent of all posts and 50 percent of their friends’ posts. In order to fix this, Instagram decided to ditch this feed order to introduce the brand new algorithm based on machine learning. Despite the backlash about losing the chronological order, Instagram now says relevancy sorting has led to its 800 million-plus users seeing 90 percent of their friends’ posts and spending more time on the app.


As mentioned, Instagram relies heavily on machine learning based on your past behavior to create a unique, personalized feed. Meaning, even if you follow the exact same accounts as someone else, your feed will differentiate according to how you interact with said accounts. From a conference that took place in Instagram’s new San Francisco office, three main factors as well as three additional signs determine what you’ll see on your feed:


Interest

Instagram determines how much you’ll care about or interact with a post based on your past activity and using image recognition. You can see how this affects your explore page by simply looking at the bottom of the posts that usually have either “Suggested by Instagram” or “Based on the posts you liked/saved” or “Based on the accounts you’ve interacted with” to further understand why and how Instagram decided to put it on your explore page. For example, I’ve recently taken a great interest in food and cat videos and Instagram didn’t disappoint with one of the quickest changes I’ve ever seen on my explore page. That doesn’t mean that I’ve completely stopped seeing my other interests, but the gradual increase in cat & food-related posts is very satisfying.

Recency

How recently the post was shared, with prioritization for most recent posts over weeks-old ones. However, this also differs according to the time a user spends on the app. If the user prefers short time intervals of usage and averages 30 minutes per day, which is not considered much, Instagram will most likely try to show the newest and most relevant posts in the hour the user was logged in. If the consumer is an avid user like me, who spends most of their leisure time on the app, the range of posts will increase and older posts will appear on their explore page, contrary to the previous user. In the example above, you can see that the post on my explore feed was posted 5 days ago, an old post in Instagram’s book but since I’m already interested in any kind of food content, recency is not the key factor for that post to be recommended.


Relationship

How close you are to the person who shared the post, with higher ranking for people you’ve interacted with the most. Basically, Instagram’s solution for you to not miss your best friend’s post in a sea of posts from all the accounts you follow. This interaction may be through commenting, regularly liking their posts, being tagged to their posts or replying to their stories. In conclusion, if your friend’s spam account is all you see in your home page, it is probably because of this factor.


Frequency

How often you open Instagram, results in the app trying to show you the “best” posts since your last visit. This method likely increases the competition of reaching more people since it has become a race to be the center of the attention of a user who spends very little time on the app. Frequency isn’t the most liked function of the new algorithm, especially for creators but in my opinion, it was made to encourage creators to create the best content they can while enhancing the consumer experience.


Following

If you follow a lot of people, Instagram will naturally show you a wider selection of content, resulting in seeing less of every individual’s content. The downside of a user following too many people for creators is once again the competition to get the desired attention and interaction.


Usage

How much time you spend on Instagram changes your content selection: the less time you spend, the more you’ll see the “best” posts. The more you spend, the more Instagram will show you a wide range of your preferred content.


Commonly Known Wrongs

In light of the newest algorithm, it is only appropriate to bust some strongly believed myths about how Instagram works.

  • Instagram is not considering going back to the old reverse chronological feed order.

  • There are no hidden posts in your feed, you will see all of them if you scroll enough.

  • The algorithm does not favor any form of content, but feeds are customized according to which type of content the user consumes the most, meaning if you watch a lot of videos, you’ll see a lot of videos.

  • Stories, Lives or other special features doesn’t affect visibility, although it may increase follower interaction which can result in more exposure.

  • Users posting too regularly won’t be downranked.

  • The type of account you have doesn’t affect your visibility unless it’s private of course.

  • Shadowbanning is not a real thing, and it is not the same thing with being restricted by Instagram either. If your post isn’t doing well, it isn’t doing well.


Conclusion

Instagram and various articles claim that the new algorithm is much more beneficial for both the creators and the users but judging from the key factors, it seems like it was created to benefit the user experience more than the creator experience. Nevertheless, I don’t think it’s a valid reason to complain since this algorithm allows the most personalized feed ever with the slight pressure on creators to put out the best content they can. Although it increases competition a lot, there’s nothing more natural for an app with 800 million+ users, it was bound to happen.


If you’re a creator and would like to learn how to navigate through Instagram in the most efficient way, I suggest you check out @creators on Instagram, especially their highlights “FAQ” and “Creator Tips”. I hope this blog was helpful enough to understand how Instagram works and how you can use it for your own benefit!

 

Author: Hanzade Durmusoglu


Hanzade is a rising sophomore from Turkey who is mainly interested in computer science, politics and history. She has some knowledge in JavaScript, Java, HTML, CSS and Swift and looking forward to learning more.

 

References

Saraco, Fran. (2020, December 18). Inside the Instagram Algorithm for 2021: How It Works and Where to Shift Your Strategy. Shopify.

https://www.shopify.com/blog/instagram-algorithm#:~:text=The%20new%20Instagram%20algorithm%20dictates,up%20being%20placed%20further%20down


Warren, Jillian. (2020, November 17). This is How the Instagram Algorithm Works in 2021. Later. https://later.com/blog/how-instagram-algorithm-works/


Cooper, Paige. (2020, April 20). How the Instagram Algorithm Works in 2020 (And How to Work With It). Hootsuite. https://blog.hootsuite.com/instagram-algorithm/


Lua, Alfred. How the Instagram Algorithm Works in 2019: Everything You Need to Know. Buffer. https://buffer.com/library/instagram-feed-algorithm/


Dobeedo. Instagram Photo Recognition ?. MP Social. https://mpsocial.com/t/instagram-photo-recognition/77594


Ayres, Scott. (2020, June 1). The Science of the Instagram Algorithm. Agorapulse. https://www.agorapulse.com/social-media-lab/science-instagram-algorithm/


Bloemsma, Megan. (2017, September 15). Image recognition in Instagram analytics. Medium. https://medium.com/@meganbloemsma/image-recognition-in-instagram-analytics-c0bae7755168








960 views2 comments

Recent Posts

See All