How to Improve Your YouTube Recommendations

YouTube users can significantly improve their content recommendations by actively managing their watch history and providing explicit feedback to the platform's algorithm. This process helps retrain the system to display more relevant videos on the homepage, in sidebars, and at the end of viewing sessions.

The platform's algorithm relies on various signals from user activity to suggest content it believes users will continue watching. While this system often works well, it can sometimes become misaligned, leading to a stream of irrelevant suggestions. Retraining the algorithm to align with personal preferences requires consistent effort and may take some time.

Ensure Watch History is Enabled

A foundational step in refining recommendations is to ensure YouTube has permission to track your watch history. If this setting is disabled, the platform may even present a blank homepage without any suggestions. To activate it, users should click their profile icon, then select Google Account. From there, navigate to Data & privacy in the left sidebar, choose Personalization settings, and confirm that YouTube History is enabled. Additional subsettings are available and can provide the algorithm with more data points for improved suggestions.

Within the same settings, users can also check the status of the Auto-delete feature for their watch history. Google offers options to automatically clear this data every three, 18, or 36 months. While this reduces the total amount of information available to the algorithm, it can also result in a more focused dataset, potentially leading to more precise recommendations.

Provide Explicit Feedback

Once watch history is active, the most direct method to influence recommendations is by consistently providing both positive and negative feedback. Positive signals indicate content you enjoy. These include watching videos completely, liking content, subscribing to creators you wish to see more from, and enabling notifications using the bell icon for favorite channels. Leaving comments also demonstrates engagement. Actively searching for topics of interest and clicking on videos that appeal to you further reinforces favorable input to the algorithm.

Equally important is providing negative feedback for content you do not wish to see. Users can dislike videos that are uninteresting or feel like a waste of time. When an unwanted recommendation appears on the screen, clicking the three-dot menu allows for options such as 'Not interested' to remove that specific video, or 'Don't recommend channel' to prevent future suggestions from an entire creator. Unsubscribing from channels no longer watched also helps clean up recommendations. There is also a distinct option available for hiding YouTube Shorts. While negative signals may not be perfect and some recommendations might reappear, and the algorithm takes time to fully adjust, combining both positive and negative feedback significantly helps the service learn user preferences.

Minimize Incorrect Viewing Activity

It is easy for YouTube's algorithm to receive erroneous signals about user interests, such as when a child uses an account on a shared device like an iPad or TV. Such incorrect activity can overwhelm the system. If this happens occasionally, users can manually correct the input. Recent searches can be removed using the 'X' button that appears in the search suggestion dropdown. Similarly, specific videos can be cleared from watch history by opening the History playlist from the left sidebar and selecting 'Remove from watch history' from the three-dot menu next to the video.

To prevent these issues from occurring in the first place, it is advisable to set up YouTube usage to avoid mixed signals. For families who frequently use the YouTube TV app, adding individual accounts for each user ensures that viewing habits do not affect others' suggestions. Alternatively, the app can be used without signing in, which prevents any viewing from influencing a specific account's recommendations.