How can I improve my YouTube recommendations in 5 easy steps

I budget monthly for streaming services, but YouTube still plays a key role in my entertainment routine. In contrast to the relatively limited catalog of movies and shows on these services, YouTube hosts hundreds of millions of videos on its servers, covering a wide range of topics from millions of creators. This wide range of content is beneficial to viewers, but it also means that YouTube’s recommendations can often be far off target.

In recent weeks, I’ve noticed an increase in disgruntled users complaining online about this same issue. Some people are unhappy to see videos that were previously suggested ad re-recommended, while others object to suggestions that do not align with their subscriptions or searches. In short, users aren’t seeing suggestions for videos they want to watch or are happy finding. This is an annoying problem that ultimately ruins the YouTube experience, whether you pay for it or not.

Thankfully, I haven’t experienced these recommendation issues recently, and that’s likely thanks to my years of careful and thoughtful YouTube use. But what does it actually mean? After some introspection, I’ve outlined some key guidelines that help keep my YouTube recommendations focused and targeted. Hopefully, these will help you too.

While YouTube’s algorithms play a role in recommending content and creating sensible scapegoats, a curated YouTube feed is ultimately dependent on user discipline.

As much as I wish YouTube knew exactly what I wanted to watch, when I wanted to watch it, that’s just not possible. It can only make inferences based on viewing history, current watch trends, subscriptions and searches. Therefore, it is important that we are careful about the content we view on our accounts.

What does this mean in practice? If I’m curious about topics or channels that I’m not yet subscribed to or that I’m not interested in, I’ll view that content in incognito mode or through a secondary, irrelevant account. Here, I can evaluate whether I really enjoy the style, content, and experience of that channel, and then decide whether to subscribe. Importantly, this subscription will directly influence my recommendations later on, so this decision will also have a cascading effect.

This pedantic practice largely keeps my main account a clean space where I get recommendations based on a narrow scope of content I really enjoy. And isn’t that what the recommendations mean?

Of course, I don’t always “evaluate” new channels. If a new channel’s content matches the topics I like to watch, I’ll often watch it on my main account. Importantly, the more concentrated your viewing history is, the more likely you are to encounter similar channels. It’s a satisfying feedback loop.

As I mentioned, YouTube also uses your search terms to inform its recommendations, and you can use this to your advantage.

This tip overlaps with my first and primary point, but if you search “latest political news,” you’ll likely see more and more recommended videos that match that particular topic.

For example, I’m on a mild balcony gardening bent, so searching for “gardening tips,” “building a balcony garden,” or “balcony gardening strategies” helps YouTube focus my attention.

Of course, this only works for your core topics of interest. If you’re really interested in mankind’s various lunar missions and beyond, I would suggest searching “history of spaceflight”, but not if you just want to briefly know the news related to the Artemis program. For this, I will once again add my secondary account.

And, on that note, using a secondary YouTube account is a big part of keeping recommendations focused on your main account. I could have mentioned this in either of the first two sections, but it is so important that it bears repeating here.

Separating your digital lives into silos is a great way to keep cross-contamination, including YouTube and other Google services, to a minimum.

Notably, this system does not need to take the size of a dedicated secondary account. You can use Incognito mode in the YouTube app, a secondary YouTube app that you use without an account, or a web browser designed for this purpose.

So how and when should you use this secondary space?

• When you want to watch a random YouTube video that doesn’t fit your established interests.
• When you’re looking to test a channel from a creator you’re not immediately sure about, or one that’s beyond the scope of your immediate topic.
• Music videos, if you don’t want your feed to be filled with music-adjacent recommendations.
• Any video sent to you by others.
• Any videos that you don’t want to impact your main account. It ultimately depends on your preference.

Subscribing to a new YouTube channel is powerful. This not only benefits the creator, but it also creates enough buzz for your YouTube account. After subscribing, you’ll likely see a flurry of recommendations coming from this account. This makes it incredibly important to be confident before pressing that subscribe button.

Every few months, I’ll go through my YouTube subscription list and remove anything that I no longer enjoy watching or haven’t watched in years. I also make sure I watch a video or two from a channel I haven’t seen recommended in a while, but have forgotten I’m subscribed to. This speeds up the recommendation system, and as soon as you finish watching this video, YouTube highly recommends content from this channel.

I’ve seen a lot of users online suggesting that using these two features on YouTube helps steer the recommendations down the right path, but I’m not so sure.

Although Google News and Discover are two completely different experiences and platforms, using these same controls only encourages that content to appear on my feed. What works is when I positively engage with content I like and ignore what I don’t. This lack of conversation appears to be a strong driver of recommendations.

For YouTube users, this means liking videos that you like more than using the “Not interested” or “Don’t recommend channel” options. This, then, also benefits the channel creator.

I will note that occasionally I will use one of these features, but largely to highlight an issue with the channel. If it has a large amount of AI-generated content, I’ll definitely let Google know about it report Option.

Fresh Start: Nuke Your Watch History

Finally, if your YouTube account has just gone too far, it’s best to clear your entire viewing history. You can even go a step further and systematically remove channels from your subscription list, but starting with a clean watchlist is a good place to start.

Judging by the flood of comments on various forums, users are flooded with YouTube recommendations that either do not suit their tastes, or are completely different from what they have seen before, or are simply serving up sleaze. Thankfully, with a few careful viewing changes, you can have a more direct impact on the content YouTube suggests.

These guidelines are based almost entirely on anecdotal experience, but I believe they have some merit – these steps have kept my account clean and focused for years. While the odd random recommendation will always make its way in, my YouTube home screen feels like I’ve created some kind of place where content is forced upon me.