A UX approach to Instagram Algorithms
Updated: Apr 8, 2020
My responsibilities: Research and analyzing data.
About the project: This school project was done to analyze the feedback on Twitter on the Instagram algorithm change. Instagram changed their algorithms of showing us posts from chronological to a more personalized approach. I was a part of a team of 4 with various academic backgrounds to complete this semester long project.
The Platform: Instagram is a popular social media platform designed for easy and free photo and video sharing.
What is being researched?
Instagram posts used to appear in chronological order on your feed. In 2016 they changed the algorithm to be more personalized with each user.
The algorithm determined what posts to be first on feed based on what the user preferred to see.
Who does it affect?
Marketing users It affects the most because they are unable to reach their entire network in an efficient fashion.
What the Factors are of the new algorithm?
Engagement: how popular the post is.
Relevancy: the genres of content.
Relationships: accounts regularly interacted with.
Timeliness: how recent the posts are.
Profile searches: the accounts you check.
Direct shares: whose posts you share.
Time spent: the duration spent viewing.
Methodology
Synthesio Social Media Intelligence Suite & suite Insights.
Analyzing tweets on Twitter prior to 2016 after the algorithm was implemented to 2019.
Sampling Conditions:
We focused our sampling around the month Instagram implemented the new timeline -- January 2018.
In order to analyze the public’s reactions, we pulled our data from the 1st of December, 2017 to the 28th of February, 2018.
Query Approach:
(“Insta” OR “Instagram” OR “IG”) AND (“algorithm” OR “timeline”)
We chose this query because people typically refer to Instagram in a variety of ways. We chose to include the verbiage “algorithm” and “timeline” because people talk about their timeline being in a different order without knowing or using the word algorithm. As a result, our query is all-encompassing.
Demographics of the Sample:
The demographics of the Tweets analyzed on Synthesio.
Sentiment:
Sentiment Analysis was performed for the tweets analyzed along with hand coding. Sentiment values are how a particular tweet emotion is analyzed by the software. This is important because it helps understand how the Instagram change was affecting users.
As you can see from the data gathered the change has received some positive feedback on Twitter. Which benefits Instagram's UX team when implementing the algorithm.
But also what is noticed is seeing how the negative sentiment value is increasing. Which means users could be noticing the change and reacting negatively to it.
Conclusion
This research is very beneficial when it comes to the User Experience of the Instagram. This research would help see how users react to the change on Tweeter and whether its positive or negative before making it permanent.
Also its important looking at it from the point of view of Marketers and companies. This change may not be ideal for those using Instagram to generate profits.