Facebook is internally testing a tabbed version of News Feed for mobile that includes the Most Relevant, Most Recent, and Already Seen feeds
Facebook may make it easier to escape its ranking algorithm and explore the News Feed in different formats. Facebook has internally prototyped …
Context & Ripple Effects
This prototype is the latest round in a decade-long cycle of Facebook experimenting with fragmenting the single ranked News Feed. Back in late 2015 it tested multiple topic-based feeds replacing the main Feed, alongside a dedicated video feed and a Trending section — none of which stuck at the time.
What makes the 2020 version different is that the tabs are organized around trust and control rather than topics: Most Relevant keeps the algorithm's picks, Most Recent offers a chronological escape hatch, and Already Seen addresses repeat content. It reads as a direct response to the ranking-legitimacy problem Facebook itself flagged when it launched global News Feed ranking feedback tests asking users which posts they actually value.
First-order effects
- Users on the test build gain an explicit, always-visible way to bypass the ranking algorithm — Most Recent functions as a chronological mode without digging through settings.
- Publishers and creators who optimize for ranked reach face a new variable: if Most Recent becomes a real surface, recency and posting cadence regain value that pure relevance-ranking had eroded.
Second-order effects
- Advertisers' inventory fragments across tabs with different engagement dynamics, forcing measurement and bidding to distinguish algorithmic impressions from chronological ones.
- Rivals that already offer chronological options face less differentiation on 'escape the algorithm' grounds, while Facebook blunts criticism of opaque curation without abandoning ranked content as the default.
Third-order effects
- The pattern points toward a permanently multi-feed app rather than one canonical Feed — a structure Facebook ultimately formalized two years later in the Home and Feeds split, separating recommended content from followed-content chronology.
- If user-controlled ordering becomes a standard surface, regulators and critics pressing for algorithmic transparency gain a concrete benchmark: platforms can be measured on whether a genuine non-ranked alternative exists, not just whether one is promised.
The trend: Facebook periodically re-tests feed fragmentation, and each iteration shifts the split from topical organization toward separating algorithmic recommendation from user-controlled chronology.