On September 18, LinkedIn Chief Product Officer Hari Srinivasan announced a split-feed trial that gives the platform’s two-decade-old network-graph logic a visible competitor: a “Network” tab for connections and follows, and a “For You” tab that surfaces posts from, in his words, “real practitioners inside and outside your network.” The latter runs on the LLM retrieval and transformer-based ranking stack LinkedIn rebuilt in March 2026. Influent calls the shift what it’s: a move “from a network graph to an interest graph.”
For a five-to-thirty-person B2B firm whose principals write about what they actually do, this is the first structural change in years that favors them over scale. Inc.’s Alicia Teltz notes the For You tab evaluates topic relevance and professional-interest signals “even when there is no prior connection between author and reader.” The ceiling network size has always imposed on organic reach, in other words, is being tested as optional.
The second half of the picture is a filter running in parallel. LinkedIn’s “AI slop” report button, launched July 30, was used more than a million times in its first weeks. Flagged posts take a roughly 40% hit in views. Influent’s Garret Caudle, relaying LinkedIn’s creator product lead, describes the target as content “polished in its presentation but lacks substance” — the exact output profile of generic AI drafting at scale.
Christopher Penn of Trust Insights has walked through the mechanism on the firm’s podcast. The retrieval layer reads the language of profiles and posts first; the ranking layer then weighs engagement. Keyword-dense profiles and topically consistent posting compound. Scattershot posting doesn’t.
The ambient numbers tilt the same direction. Metricool data cited by Net Influencer shows Personal Profiles pulling 63% higher engagement than Company Pages in 2026, with 237% more comments at comparable impressions. Company Page likes fell 13% year over year, comments 17%, shares 11%. LinkedIn’s own reported video viewership grew 36% to roughly 154 billion views, a reminder that the surface area for new formats is still expanding while the brand-page surface is quietly collapsing.
Two cautions. Net Influencer, citing Social Media Today, points out that earlier LinkedIn feed experiments, including 2026’s topic-based suggested feeds, never graduated into full launches. And the For You logic sits uneasily alongside a separate shift already underway in B2B buying, which our earlier report on marketers preparing for AI-assistant buyers at HubSpot Unbound covered: the human reader on the other end of a post may soon be a model summarizing it.
Which is the structural read. LinkedIn is rebuilding distribution around topic identity at precisely the moment the audience itself is becoming machine-mediated. Firms that spend Q4 teaching the retrieval layer who they’re will be legible to both.
Sources
- https://www.inc.com/alicia-teltz/linkedin-is-rolling-out-a-split-feed-businesses-can-win-the-for-you-tab-by-making-expertise-obvious/91410274
- https://www.netinfluencer.com/linkedin-trials-a-network-tab-in-bid-to-revive-engagement-between-connections/
- https://www.socialpilot.co/blog/new-linkedin-features-and-updates
- https://influent.co/blog/linkedin-platform-updates
- https://www.trustinsights.ai/blog/2026/09/in-ear-insights-linkedin-algorithm-updates-october-2026/