Substack gives readers the dial: what the platform's new feed controls mean for creators

Substack has launched a set of content preferences that let users adjust how much video, audio, longform or community content appears in their feed, mute specific keywords and phrases, fine-tune the topics they follow, and automatically label AI-generated material. The controls live in the content preferences section of user settings.

The move follows a wave of similar feed-customization tools rolled out this year by Meta and TikTok. But where those platforms started with topic sliders and interest tags, Substack's suite is broader and more granular from the start. Users can hide up to 50 words or phrases, each up to 100 characters, and they can dial format types up or down with sliders covering video, audio, chat and longer posts.

The timing matters. In the past six months, the Substack app has driven nearly 800,000 paid subscriptions and more than 80 million free subscriptions to writers and creators, with more than half of paid subscriptions from the app coming directly from the Notes feed. Feed controls arriving now signal that Substack sees its discovery surface as critical infrastructure, not just a nice-to-have.

The pattern across platforms

Substack is joining a trend that accelerated sharply in the first half of 2026. Instagram rolled out its Your Algorithm tool to all US users in December 2025, letting them manually control which Reels are algorithmically suggested. By June 2026, Your Algorithm was live across all of Instagram, including Feed, Reels and Explore, giving people ways to control their experience globally.

Threads launched Your Algo on 16 June 2026, building on the platform's Dear Algo tool that debuted in February; the new feature makes content preferences private, without requiring a public post. Users can tell Threads they want to see more or less of certain topics and choose how long the request lasts: one, three, or seven days.

TikTok's Manage Topics feature lets people use sliders to see more or less of big categories like sports, travel, humor, creative arts and others in their For You feed. TikTok has also rolled out smart keyword filters that use AI to block not just exact words but related terms in captions and hashtags, plus a slider to control AI generated content.

Meta is now bringing Your Algorithm to Facebook after testing it on Instagram, which means users can tell the app in their own words what they'd like to see more or less of in Feed and Reels; the feature started on Instagram Reels in December and spread to Feed and Explore in June.

What separates this wave from earlier "not interested" buttons is the degree of transparency and specificity. Platforms are surfacing what they think you care about and letting you correct or amplify it in plain language. That's a shift from inferring preferences passively to asking users to state them directly.

Why platforms are doing this now

Substack says the level of control serves both creators and readers, and that both sides want less noise and more help connecting readers with work that they'll find meaningful. That framing is deliberate. The company positions the Substack feed as built not to maximize time spent scrolling but to get users to the work they'll care about, work they'll stick with and ideally pay for.

The broader driver is user frustration with feeds that feel hijacked. For years, social media giants controlled what users saw in their feeds; while people could follow accounts, like posts or hide content they didn't enjoy, recommendation algorithms controlled what was shown to them, and now social platforms are handing over some of their power by allowing users to personalize their algorithms with the help of AI.

There's also a competitive element. Almost every major platform has been heading this way in 2026; Threads has Dear Algo, TikTok users have been able to adjust their Manage Topics sliders for some time, and over in streaming, Spotify's Prompted Playlist and Amazon Music's Alexa+ are teaching listeners to simply ask for what they want.

Algorithmic transparency is becoming table stakes. Offering none puts a platform at a disadvantage when audiences can get it elsewhere. That applies even to subscription platforms like Substack, where the feed is competing for attention against email inboxes, podcasts, YouTube and everything else a reader could open instead.

What it means for creators and marketers

The immediate implication is that reach is no longer a question of what the algorithm infers. It's now a question of whether your content type and topic survive an active filtering decision.

Substack writers are already responding: one creator noted that readers getting mute controls means a newsletter that covers five topics for no one in particular is easy to pass over. Specificity is defensive. If a reader can't quickly identify what you're about, they can now remove you from their feed in seconds rather than just scrolling past.

The format sliders matter too. A creator who leans heavily on video or community features is now competing against the user's explicit preference for longform or audio. That doesn't kill video, but it does mean video needs to justify itself to an audience segment that has opted down. The barrier to ignoring you just got lower.

Keyword muting is the sharper edge. Posts and notes that contain a muted word or phrase are hidden from the feed entirely, with users able to hide up to 50 words or phrases, each up to 100 characters. A social media marketer covering a polarizing topic, a saturated trend or even an overused buzzword is one keyword away from zero reach with that user. The "ignore this once" behaviour has been upgraded to "never show me this again, and make it automatic."

For brand social teams, the pattern is more important than any single platform's implementation. Our breakdown of the algorithm change on Threads in June showed the same dynamic: audiences want control, and platforms are giving it to them. That control includes the ability to dial brands down or out entirely.

The shift also affects content mix decisions. Teams that defaulted to video because the algorithm rewarded it must now consider that a portion of their audience has explicitly dialled video down. Longform written content, which many platforms deprioritized for years, now has a lever working in its favour if readers choose to use it.

Feed control is not the same as feed ownership

It's worth noting what these tools don't do. Your Algo requests on Threads deliberately expire after a week, so Meta keeps the final say over the feed; users can nudge the algorithm, but they cannot take the wheel, unlike on Bluesky, which has leaned into handing users full control.

Substack's sliders are similar. They influence what appears, but the underlying recommendation system is still running. Users are adjusting inputs, not rewriting the algorithm. The platform still decides how much weight each preference carries and how it balances competing signals.

That's a critical difference from the kind of control Instagram's Mosseri has publicly resisted, where users wanted a purely chronological feed of accounts they follow with no algorithmic suggestions at all. These new tools don't offer that. They offer a more responsive, transparent algorithm, but still an algorithm.

The result is that creators are not competing in a user-controlled environment. They're competing in an environment where users can express dissatisfaction more clearly and the platform will act on it faster. That's a different challenge. It rewards consistency, clarity and staying within the bounds of what an audience has signalled they want, rather than trying to break through with something unexpected.

What to watch next

The obvious question is whether other platforms follow Substack's specific feature set or whether the trend plateaus here. YouTube, LinkedIn, Reddit and X have not rolled out comparable tools at this scale. If they do, that would complete the shift across the social ecosystem.

The less obvious question is how much users actually use these controls and whether the behaviour sticks. Early adoption is easy to tout, but sustained behaviour change is harder. If a majority of users never touch the sliders, the competitive pressure eases. If they do, and if feed preferences become as routine as adjusting notification settings, then every platform without them will face louder demands.

For now, the pattern is clear. Platforms are competing on transparency and user agency, and that competition is moving faster than most social teams have adjusted their strategies. As Pinterest demonstrated earlier this year when it launched AI content controls after user backlash, waiting until your audience revolts is not a winning move. The smarter play is to assume feed controls are the new normal and build content strategies that survive them.

That means being specific about what you cover, consistent in format and topic, and honest about whether you're still showing up for the audience you think you have or the one that's already quietly filtered you out.