User-Driven Feed Pruning Shapes Creator Reach on YouTube
AI-generated image · US National Wire
New guidance on refining YouTube's recommendation engine highlights how viewer signals directly control content discovery.
YouTube's recommendation engine serves as the primary gatekeeper for creator visibility, relying on a complex array of user signals to determine which videos appear on homepages, sidebars, and end-screens. According to reporting from Engadget, users can actively reshape these recommendations by providing specific positive and negative feedback to the algorithm.
Positive signals that boost creator reach include watching videos in their entirety, hitting the "Like" button, subscribing to channels, enabling notification bells, and leaving comments. Conversely, users can prune their feeds by hitting "Dislike," selecting "Not interested" via a three-dot menu, or choosing "Don't recommend channel" to remove specific creators from their suggestions entirely.
Beyond active feedback, the algorithm is heavily influenced by account settings and data management. Engadget notes that disabling watch history can result in a blank homepage with no suggestions. Users also have the ability to auto-delete history every three, 18, or 36 months, or manually remove specific entries to refine the data set the algorithm uses. To prevent "erroneous signs" from skewing recommendations—such as when children use a parent's account—users can employ Incognito mode, private browsing, or an account lock feature on TV apps.
Because the recommendation engine is the fastest way for viewers to find new content, these user-led adjustments to the algorithm's data set directly impact which creators are surfaced and which are suppressed.