YouTube subscriptions feed: why creators should worry (and adapt)

Your audience is learning a new trick: skip the "Home" feed entirely.

And if you've been living off algorithmic luck (or Shorts drive-bys), that should make you a tiny bit nervous. The kind of nervous that gets you moving, not panicking.

What happened

Felix Kjellberg (PewDiePie) just posted a video showing a browser add-on that changes what happens when you open YouTube: instead of landing on the recommendation-heavy homepage, it pushes you straight into your Subscriptions feed.

The headline isn't "celebrity builds extension." The headline is the philosophy behind it: less passive scrolling, more deliberate viewing. In the same breath, he pointed people to an existing tool that already does a lot of this decluttering: Unhook, a long-running extension that can hide homepage feeds, sidebar recommendations, Shorts, comments, end screens, and more - and it's massive now (over 1,000,000 users, updated March 22, 2026). ([chromewebstore.google.com](https://chromewebstore.google.com/detail/unhook-remove-youtube-rec/khncfooichmfjbepaaaebmommgaepoid))

This lands in the middle of a bigger YouTube tension: YouTube keeps experimenting with the Subscriptions experience - especially on TVs - by inserting "most relevant" blocks, live sections, and Shorts before you reach the clean chronological flow people actually came for. ([androidauthority.com](https://www.androidauthority.com/youtube-subscription-section-test-3642927/))

Also worth remembering: since 2023, YouTube has explicitly tied recommendations to Watch History. Turn history off (and don't have much prior history) and the homepage basically collapses into a minimalist screen - search bar, menu, not much else. That's not a bug. It's the platform saying, "No data, no feed." ([9to5google.com](https://9to5google.com/2023/08/08/youtube-watch-history-changes/))

Why creators should care

If more viewers "un-algorithm" their YouTube, the upside is brutal and beautiful: attention gets quieter, but cleaner. People who choose you in a Subscriptions-first world are more likely to watch long-form, finish episodes, and actually remember you.

The downside: top-of-funnel discovery gets weird. If someone hides recommendations, removes Shorts, or blanks their homepage by keeping Watch History off, your content has fewer accidental entry points. That hurts especially if your strategy is "one banger thumbnail + the sidebar will do the rest."

Distribution-wise, this is another reminder that "subscribers" isn't a vanity metric - it's a delivery mechanic. Except YouTube's been quietly sanding down that mechanic with UI tests and injected "relevant" sections (especially on TV, where people binge). ([androidauthority.com](https://www.androidauthority.com/youtube-subscription-section-test-3642927/))

Monetization? Same story. When viewing shifts from endless browse to intentional sessions, watch time can drop... but conversion can climb. Fewer random viewers. More actual fans. Better RPM doesn't always come from more views. Sometimes it comes from the right 10,000 people showing up like clockwork.

One more layer: PewDiePie didn't arrive at this by accident - he's been publicly deep in the "build my own tools" arc, including running local AI setups and experimenting with training/fine-tuning models. That mindset - creators shipping utilities, not just content - is spreading. ([tomshardware.com](https://www.tomshardware.com/tech-industry/artificial-intelligence/pewdiepie-goes-all-in-on-self-hosting-ai-using-modded-gpus-with-plans-to-build-own-model-soon-youtuber-pits-multiple-sentient-chatbots-against-each-other-to-find-the-best-answers?utm_source=openai))

What to do next

• Start optimizing for the Subscriptions feed like it's the homepage. Because for a growing slice of people... it is. Clean titles. Clear series naming. Thumbnails that read small. Upload cadence that doesn't feel like random noise.

• Build one off-platform "direct line." Email list, SMS, Discord - pick one. Not seven. If the platform UI shifts (again), you still have a way to launch without begging an algorithm for permission.

• Make "intent content" a pillar. Tutorials, breakdowns, evergreen explainers, behind-the-scenes - stuff people search for on purpose. If recommendations get hidden, search and subscriptions pick up the slack.

• Assume viewers will keep removing distractions - and plan funnels accordingly. Shorts can still work, but treat them like trailers that point to a destination (a long video, a newsletter, a product). Not like a slot machine you feed forever.

If you want the meta-lesson: the future isn't "algorithm vs subscriptions." It's "how fast can you adapt when your audience rewires the platform underneath you."