Northeastern University researchers just published findings that shake up everything creators thought they knew about TikTok's feedback tools. The "Not Interested" button outperforms swiping — but its effects can fade within minutes of stopping. Here's what that means for your growth strategy right now.
1B+
TikTok daily active users whose feeds are shaped by the algorithm
Source: Northeastern University, 2026
Minutes
How fast filtered topics can reappear after you stop pressing Not Interested
Source: Northeastern University Research
72%
Year-over-year jump in TikTok video content volume — more competition for every slot
Source: Metricool 2026 TikTok Study
In June 2026, Northeastern University published a study examining how much real control TikTok's billion-plus daily users actually have over their own For You Page. The headline finding: the "Not Interested" button is a genuinely stronger signal than simply swiping past content you don't want to see. That's good news on the surface.
The catch, explained by Northeastern professor Piotr Sapiezynski of the Khoury College of Computer Sciences, is that the suppression effect may not stick. Topics users mark as "not interested" can begin reappearing in their feed within minutes of them stopping — suggesting the algorithm treats the button as a short-term dampener rather than a permanent exclusion rule.
As Knowridge Science Report summarized the findings: "users may need to continually click on that button because video topics they specified as 'not interested in' may begin popping up again within minutes after stopping." The implication is blunt — users are ultimately at the mercy of TikTok's company decisions about what surfaces on their feeds.
The button isn't labeled the way you might expect. As AEANET's guide to the feature explains, TikTok surfaces it through a long-press gesture on any video in your feed. Hold down on the post and a menu appears with the "Not Interested" option alongside other actions. There's no standalone button in the app's main navigation — it's a contextual tool buried inside an interaction menu.
TikTok officially describes this as a way for users to tell the algorithm they no longer want to see specific content. The platform's For You Page algorithm analyzes interactions, video completion rates, and time spent on content — and negative signals like Not Interested are supposed to recalibrate the mix. The Northeastern research suggests this recalibration is real but temporary.
| Action | Immediate Effect | Durability | Effort |
|---|---|---|---|
| Not Interested button | Stronger signal | Fades within minutes | Long-press required |
| Swiping past | Weaker signal | Inconsistent | Effortless |
Based on Northeastern University 2026 research
TikTok has never published documentation explaining why suppression effects decay, but the research and platform behavior point to a few likely mechanisms. The algorithm is designed to keep users engaged above all else — and it's built to re-test content categories periodically to see if preferences have shifted. This is the same mechanic that can surface something you watched once six months ago and decided you hated.
According to the Northeastern study, the platform's algorithm is essentially a black box — highly personalized but operating on signals users can't fully audit or control. Professor Sapiezynski's team found that the Not Interested feature "doesn't always work as advertised," and that users are ultimately dependent on TikTok's own decisions about what gets surfaced.
For creators, this has a direct implication: the algorithm's tendency to re-probe suppressed categories means your content can get re-surfaced to audiences who already told TikTok they didn't want it — creating a fresh round of negative engagement signals that compound over time.
According to Metricool's 2026 TikTok Study (which analyzed over 2.3 million posts), content volume jumped 72% for videos and 140% for image posts in a single year. More content in the feed means the algorithm has more options to re-fill any slot left by suppressed content — which may be why suppression effects disappear so fast. There's simply always something to replace what you filtered out.
The research doesn't just affect what viewers experience — it reveals something important about how fragile algorithmic feedback loops are. As a creator, the lesson isn't that the Not Interested button is irrelevant; it's that you can't rely on audience self-filtering to protect your distribution. You need to think proactively about avoiding negative engagement signals in the first place.
If viewers skip or press Not Interested in the first few seconds, that's your worst outcome. The algorithm weights early engagement heavily — an immediate drop-off tells it your content was mis-targeted.
According to Metricool's data, watch time is TikTok's core ranking signal. Videos with strong completion rates get pushed to wider audiences regardless of Not Interested presses on similar content.
If you're getting consistent negative signals on a topic, don't just post it more — the algorithm's re-probing behavior will keep surfacing your content to audiences who rejected it.
Since users can't reliably control their feeds, and the algorithm re-probes suppressed content, creators with email lists and community touchpoints outside TikTok maintain growth that's algorithm-proof.
The broader takeaway from the 2026 TikTok marketing landscape is that TikTok's recommendation engine rewards engagement quality over quantity. With content volume up 72% year-over-year, the creators winning aren't necessarily posting more — they're getting stronger signals from each post. And according to Metricool's analysis, videos still dominate: they get 5.6 times more views and 7.8 times more interactions than image posts.
The Northeastern research raises a deeper question that goes beyond creator tactics: whether TikTok's stated commitment to user control is backed by engineering reality. The platform says the Not Interested button is a meaningful tool for managing your feed — but when its effects fade within minutes, the practical value for users is much lower than advertised.
This isn't unique to TikTok. Recommendation algorithms across social platforms are optimized for engagement metrics first, and user preferences are one input among many. But TikTok's particular reliance on its For You Page — which surfaces content from accounts users don't follow — makes the quality of its feedback tools unusually consequential.
For creators, the practical reading of the Northeastern findings is this: you can't control what your audience does with the Not Interested button, but you can reduce the likelihood they reach for it. That means sharper content-audience fit, stronger hooks, and consistent niche positioning — the same fundamentals that drive organic growth regardless of platform changes.
The research is clear: users can't fully control TikTok's algorithm, and neither can creators who post and pray. SocialScale Hub gives you the automation and analytics to build consistent growth that doesn't depend on algorithmic luck.