Instagram is testing new entry points for its Your Algorithm controls, moving a once-buried recommendation setting closer to the moments when people are actually scrolling Feed and Reels.
The latest tests were described by Instagram head Adam Mosseri in a recent post and first reported by TechCrunch on June 27. The examples include pulling down in the main feed to bring up the Your Algorithm menu, swiping up from a Reel to reach a similar customization prompt, and buttons below Reels that let users tell Instagram whether they want more videos like the one they just watched.
Mosseri framed the work as an attempt to make the tool feel central to Instagram rather than hidden in settings. He also cautioned that the examples are not all final: “Some of this is testing now, some is coming soon, some might not work.”
What Your Algorithm Already Does
Your Algorithm lets Instagram users review the topics the app thinks they are interested in and adjust those topics to influence recommendations. Instagram expanded the feature to the main feed earlier in June after introducing it in other recommendation surfaces such as Reels and Explore.
In practice, the feature is less like a full ranking-control panel and more like a topic-tuning screen. A user may be able to add or remove interests such as cooking, travel, sports, parenting humor, or dog videos. Instagram then uses those signals to reshape recommended posts and videos across major parts of the app.
That matters because Instagram’s home feed is no longer just a stream of posts from followed accounts. Like TikTok, YouTube Shorts, Threads, and other recommendation-heavy products, Instagram increasingly treats the feed as a discovery engine. The new controls are Meta’s attempt to make that system feel less one-sided without abandoning algorithmic recommendations.
Why Instagram Can Expose More of the Algorithm Now
The technical shift behind the feature is language. As Engadget noted when the main-feed expansion rolled out, Mosseri has said large language models help Instagram describe inferred interests in plain words rather than leaving them as opaque recommendation clusters.
That does not mean Instagram is handing users the actual ranking model. It means the app can translate behavior signals into editable labels. If someone watches a string of climbing videos, recipe Reels, and AI art clips, Instagram can show those as understandable topics and let the user correct the list. The resulting control is still mediated by Instagram’s systems, but it gives people a more direct vocabulary for steering them.
The new tests are important because they bring that vocabulary into the browsing flow. A buried setting is easy to forget. A prompt that appears while pulling down the feed, swiping through Reels, or reacting to a specific video turns algorithm tuning into an everyday interaction.
The Big Limitation: Followed Accounts
The feature still does not give users the control many of them ask for most often: a main feed that prioritizes people they already follow. TechCrunch noted that popular comments under Mosseri’s post focused on that request, and Engadget reported that asking the tool for posts from followed accounts returned no useful result during its testing.
Instagram does offer a dedicated Following feed, but Your Algorithm is built around interests, not relationships. That distinction is the heart of the tension. For ordinary users, the new controls may reduce the feeling that the app is drifting into irrelevant recommendations. For creators, they do not restore the older bargain where building a follower base reliably meant reaching those followers in the primary feed.
Mosseri has acknowledged that the follow graph used to be a more meaningful way for people to shape Instagram. The company’s current answer, though, is not to reverse the recommendation shift. It is to make recommendation systems more legible and more adjustable.
What It Means for Creators
For creators and brands, Your Algorithm points toward a more topic-explicit version of Instagram discovery. Content that clearly fits a recognizable category may be easier for Instagram to match with users who have chosen that topic. Content that jumps between unrelated themes may become harder for the system to classify.
That does not mean every creator needs to flatten their work into one narrow niche. It does mean the first few seconds of a Reel, the caption, the visual subject, and the account’s broader pattern all matter more when users are actively telling Instagram which topics they want more or less of. A fitness creator who also posts cooking clips can still make both work, but Instagram needs enough consistent signals to understand when each post belongs in a user’s chosen interest bucket.
The current tests also suggest that Reels feedback may become more granular. A simple “more like this” button below a Reel would be a stronger explicit signal than a watch, like, or skip, because it tells Instagram that the category of the content matters, not just the individual post.
A Control Panel, Not a Power Shift
Instagram’s algorithm controls are becoming more visible, but they are still controls inside Meta’s recommendation business. The company remains heavily invested in discovery feeds because they increase time spent, surface more creator inventory, and compete directly with TikTok-style entertainment loops.
The useful change is that Instagram is starting to expose more of the recommendation system in language ordinary users can understand. The unresolved problem is that people may want different kinds of control than the platform is ready to provide. Topic tuning helps when the feed is wrong about interests. It does much less when the feed is wrong about relationships.
If the tests ship broadly, Your Algorithm could become one of Instagram’s most important product surfaces: not because it replaces the feed algorithm, but because it gives users a visible way to negotiate with it.
