Instagram is one of the most popular social media platforms out there right now. There are a total of 1 billion monthly users on the platform according to Omnicore Agency’s Instagram stats post. In June 2016 Instagram got rid of their chronological feed and started using an algorithm to display posts it thinks users would like to see. This is some type of recommendation algorithm and uses various criteria to determine what to show you. It also uses image recognition technology to see what type of photos you usually interact with.
It’s not a secret that a lot of people dislike the feed changes, but this was meant to help users have a better experience. After that change a lot of posts started popping up about how to “beat this algorithm” or “Instagram hacks for fast growth”. I’m sure you’ve read one or two or even tried some of these methods. I think there is a big misunderstanding on what Instagram was meant for and a lot of the advice on these posts are actually making things worse for normal users. Instagram themselves has explained how their algorithm works, yet people keep trying to cheat their way to the top, which ultimately clutters feeds and throws off the recommendation system.
Let’s get a little technical
Some of you might know, I am a computer geek, especially when it comes to data and data analytics. The more I study and work with data science and machine learning, the more I understand why these tactics being passed around about beating the algorithm aren’t going to work.
What is the Instagram algorithm?
So what is this “algorithm” everyone talks about? Basically an algorithm is a recipe to do something that a computer follows. Machine learning algorithms take in a set of parameters and apply them to formulas used within the algorithm to get results, in this case it’s the rankings or weights of photos/accounts or however they classify their data*.
The learning part means the system is able to tweak these parameters by itself or with the approval of someone at Instagram. The algorithm feeds itself back the data it has, re-evaluates results and makes the appropriate changes to the parameters, this is considered AI.
So what does all the technical stuff mean?
It means that your own data is what changes the algorithm along with the ML supervisor † at Instagram (if they choose to tweak the model ‡).
This also means that the way the algorithm works is different for each of us. What works for you to get a photo bumped up on feeds may not work for me since our accounts and followings are different.
Users that have similar content and activities are more likely to have similar results. Also you will notice that people you interact with more show up on your feed more and vice versa.
Now because of all the bots and people trying to swing the algorithm in their direction, these rankings are most likely getting skewed. This is probably the reason for those clean ups that Instagram has been doing. They are trying to eliminate the noise being fed to the algorithm.
What does this mean for normal users
Well it means as long as there are people trying to cheat the system by using autolikers, bots and engagement groups to boost their numbers, people trying to grow legitimately will ultimately suffer because volume reduces reach. So accounts that artificially boost their engagement will be shown to more people, but this means trying to do it legitimately will be harder. This is why now the likes on a photo matter less for your reach than they used to, because of people artificially getting likes to boost popularity…. so next time you think of joining mass comment or like groups, or even using a bot (which btw is against the ToC ) think about the affect you are having on Instagram and your own (and even your followers) recommendations.
The truth
I hope this post was able to bring you some clarity on the truth about all these articles floating around on trying to beat the algorithm. The reality is you cannot beat it, it changes all the time depending on what you do. What you should do is use Instagram honestly and the way it was meant to be used, as a way to share content with friends and get discovered by people with similar interests. If we all start doing this, I believe Instagram will eventually work better for everyone in the end.
* I’m simplifying the explanation of a machine learning algorithm, but the details are not relevant here.
† I made this title up but I’m certain. Instagram has a team dedicated to maintaining and checking the ML algorithm
‡ A machine learning model can be a mathematical representation of a real-world process. To generate a machine learning model you will need to provide training data to a machine learning algorithm to learn from.
