YouTube custom feeds expand to allow multiple user-defined landing pages

The expansion of YouTube custom feeds signals a shift toward user-curated algorithmic control as a retention tool. By allowing viewers to define their own landing pages, YouTube is directly addressing consumer demand for depersonalized or highly specific content silos, a trend popularized by Bluesky's open feed architecture. This move is strategically significant as decentralized platforms begin hosting longer native video, threatening YouTube's dominance in mid-form content. As platforms move away from a single 'black box' algorithm, the industry should watch for whether these custom feeds increase session duration or if users find the manual curation too high-friction for daily use.

Bluesky has emerged as the primary reference point for user-controlled content discovery, and its video strategy is expanding. In early 2026, Bluesky announced native video uploads up to three minutes in length with automatic thumbnail generation and a dedicated video tab, positioning the platform as a potential destination for mid-form creators who previously relied on YouTube for that format. The platform's open feed architecture, which lets any developer build and publish algorithmic feeds that users can subscribe to, has attracted more than 40,000 custom feeds as of mid-2026, according to Bluesky's public developer documentation. This ecosystem of third-party curation is precisely the model YouTube's custom feed experiment appears designed to internalize within a closed platform.

Google has simultaneously been tightening the rules around how AI systems surface and recommend content, a dynamic that intersects with YouTube's feed personalization strategy. In May 2026, Google updated its spam policies to explicitly prohibit attempts to manipulate generative AI responses in Search, signaling stricter enforcement against deceptive practices that could influence AI-driven discovery. Around the same time, Google published new documentation on optimizing websites for generative AI features in Search, emphasizing non-commodity content and agent-friendly structures. These policy moves matter for YouTube custom feeds because they define the boundaries within which algorithmic recommendation systems can operate, particularly as AI agents increasingly mediate content discovery on behalf of users.

On the technical side, Akamai's recent launch of AI Brand Presence offers a parallel data point on how platforms are adapting content delivery for AI-mediated discovery. Akamai reported that its own pilot of AI Brand Presence produced an 85% increase in AI citations and a 364% surge in brand visibility for general searches, demonstrating that actively managing how AI systems interpret content can materially shift traffic patterns. For YouTube, the custom feed expansion represents a similar bet: that giving users explicit control over how content is categorized and surfaced will increase engagement metrics enough to justify the added interface complexity. The key difference is that crawling the open web, while YouTube is optimizing for internal recommendation signals within a walled garden.