Instagram Has Answered Your Questions About How Its Algorithm Works

It’s been nearly two years since Instagram switched up its algorithm from showing all posts in reverse chronological order to a more complex, machine-learning driven approach. As soon as the platform announced this change, uproar ensued. (A group of users even spearheaded a Change.org petition imploring Instagram not to move away from the reverse chronological format.)

When Instagram started prioritizing one account’s content over another, users and businesses started complaining about fairness, especially for those trying to build brands and bolster marketing on the platform. It was a mystery to many users: What would cause Instagram to rank posts from one account over another in users’ feeds? And the fact that different users would see posts from the same account at different intervals made it all the more opaque and hard to crack.

To demystify the change, the company recently hosted an information session at its San Francisco offices to set the record straight to a group of reporters about how the algorithm decides what an individual user sees.

As TechCrunch reports, prior to the flip to algorithmically generated feeds, the average Instagram user would never see 70 percent of all posts by brands and friends they followed — and 50 percent of friends’ posts alone. Today, the company says that users see 90 percent of their friends’ posts and spend more time on the app than they did in early 2016. Yet, despite there being half a billion daily users, the platform is more saturated than ever with individuals and brands vying for eyeballs.

The company reportedly has no plans to give users the option to revert to a reverse chronological feed, despite demand. Instead, it’s decided to educate users about how the algorithm works. Here are the factors that go into how Instagram prioritizes what an individual user will see as they scroll, according to the company:

Related: 10 Instagram Tools to Help You Build Your Following in 2018

Interest

Instagram uses machine learning to surface up the posts a user is most likely to care about based on past behavior. If a user has liked, commented on or lingered over similar content in the past, they’ll likely see it again. For example, if you always hover over photos of slices of cheesy pizza, Instagram will show you similar images…