Social media algorithms are always a thing — and with Meta’s Threads in the mix, many are curious about how the Threads algorithm works.
In other words: How does Threads decide which content to show a user? How does it seem to “know” what they will be interested in?
Wonder no more. Meta has given us insight into the Threads Feed AI system — and it’s worth taking a peek!
First, the system makes predictions based on several factors: what and whom you’ve followed, liked or engaged with recently. Here’s how it works:
Inventory
First, the Threads algorithm looks at all the platform’s content it could show a user. So, when you post content to the platform, it doesn’t automatically go to all of your followers. Instead, the algorithm takes inventory and evaluates it before sending out to engaged users.
Signals
Next, Threads considers signals — or signs a user has either liked or disliked content. For example, a user might share or comment on a post. That’s a signal they enjoyed that type of content. On the flip side, if someone scrolls past something, the algorithm knows they didn’t enjoy the specific content.
Rank content
Finally, the system ranks content based on signals. If the system thinks certain content will provide more value for someone, that user will see it higher in their feed. This results in the platform making content recommendations that more closely match users’ preferences.
So, just how does the AI deliver content to users? That’s where ‘prediction models’ come into play, and those could be anything from whether or not a user will like a post, how likely they are to read replies on a post, and more.
Of course, users can customize what they see, too. Whether they want to unfollow an account, mute content, or hide it — there’s always more options.
Have questions or want to learn more about Threads, in general? SND is always ready to discuss strategy! Drop us a note: support@socialnewsdesk.com.
