Back to blog

How to actually stay off YouTube without blocking it entirely

·6 min read

You need YouTube. Conference talks, tutorials, product demos. Your job requires it. But you also lose hours to YouTube. Video essays, rabbit holes, the algorithm feeding you 'just one more'.

The obvious solution is to block YouTube. But then you can't watch work videos. So you unblock it 'just for this one video', and 45 minutes later you're watching a documentary about deep-sea submarines.

This is the YouTube problem: the useful content and the distracting content live on the same domain. Domain-level blocking forces an all-or-nothing choice. Here's how to solve it.

Why 'just watch the one video' doesn't work

The reason you can't 'just watch one video' is not a failure of willpower. It's the way YouTube is designed. The moment a video ends, the next one autoplays. The sidebar recommends videos tailored to keep you watching. The homepage is optimised to surface content you'll click on.

This is not an accident. YouTube's business model is attention. The longer you stay on the site, the more ads you see. Every part of the interface is designed to keep you there: the autoplay countdown, the endless scroll, the notifications about new uploads from channels you follow.

When you unblock YouTube to watch one work video, you're not just giving yourself access to that video. You're giving the algorithm permission to pull you into the next one, and the next one after that. The only way to resist is constant vigilance, which is exhausting and often fails.

The allowlist solution

URL-level blocking solves this by letting you block YouTube by default, but allow specific videos. Here's the workflow: you block youtube.com in your focus app. YouTube is now inaccessible. The homepage, search, subscriptions, shorts, all blocked.

Then you find a video you need to watch. You copy the URL (youtube.com/watch?v=abc123), add it to your allowlist, and refresh. That video loads. But when you try to click on a recommended video, or navigate to the homepage, or open YouTube Shorts, you see the block page.

The only part of YouTube that's accessible is the specific video you allowed. Everything else stays blocked. This removes the decision point. You don't have to resist clicking on recommended videos, because they're not clickable. The algorithm can't pull you in, because the interface is blocked.

What this looks like in practice

Let's say you're working and you need to watch a conference talk. You search for the video, find the URL, and copy it. Then you open your focus app (Kyte, in this case), add the URL to your allowlist, and refresh the page.

The video loads. You watch it. When it ends, the autoplay countdown starts, but the next video is blocked. The sidebar shows recommendations, but clicking on them takes you to the block page. The only way to watch another video is to add its URL to your allowlist, which creates just enough friction to stop the impulse.

When you're done, you close the tab. Later, if you absentmindedly try to open YouTube, you see the block page. The only way to watch more videos is to deliberately find their URLs and add them to the allowlist, one at a time.

Handling channels and playlists

What if you want to allow an entire channel, not just a single video? URL-level blocking can handle this too. If you trust yourself with a specific channel (a conference channel, an educational creator, a work-related podcast), you can add a wildcard rule that allows all videos from that channel.

For example, you could allow youtube.com/watch?v=* from a specific channel, or youtube.com/@channelname/*. This gives you access to everything from that channel while keeping the rest of YouTube blocked. The trade-off: if the channel posts anything distracting, you'll see it. But for focused channels, this is a reasonable compromise.

Playlists work the same way. If you have a work playlist, you can allow the playlist URL. This lets you watch all the videos in the playlist without giving you access to the rest of YouTube.

Why domain blocking fails for YouTube

Domain blocking treats YouTube as binary: it's either all blocked or all unblocked. This works for sites you never need (social media you don't use for work, news sites that waste your time). But YouTube is different. The useful content and the distracting content are mixed together, and the algorithm is designed to pull you from one to the other.

When you unblock YouTube at the domain level, you're not just unblocking the one video you need. You're unblocking the homepage, the search page, the subscriptions feed, and the entire recommendation engine. The site is designed to keep you there, and you've just given it permission to try.

URL-level blocking breaks this dynamic. You allow the specific page you need. The rest of the site stays blocked. The algorithm can't tempt you, because the interface isn't accessible.

The YouTube problem is not unique to YouTube. Reddit, Twitter, and any site with an algorithmic feed has the same dynamic: useful content mixed with infinite distraction. Domain blocking doesn't work for these sites, because blocking them entirely means losing access to valuable content.

URL-level blocking solves this. Block the domain by default. Allow the specific pages you need. Everything else stays blocked. It's not perfect, but it removes the constant negotiation and lets you access what you need without falling into the algorithm.

Block YouTube by default, allow videos one at a time. Try Kyte.

Get started