X's Algorithm Appears to Penalize News Links, Sparking Debate Over Platform's Value for Publishers

An analysis of the 200 most-recent tweets from 18 major publishers on X reveals a striking pattern: engagement appears to drop significantly when tweets include links to external content. This finding comes amid ongoing debate about the platform's value for news organizations trying to reach audiences.

The data shows that tweets from high-follower accounts like @nytimes (53 million followers) generate far less engagement than similar content from smaller accounts like @GlobeEyeNews (866,000 followers). The difference becomes particularly pronounced when links are involved.

This trend raises several questions about X's algorithm and its implications for journalism:

Why links might hurt engagement

• X may be prioritizing native content to keep users on the platform longer
• The algorithm could be designed to discourage traffic to external sites
• Link-heavy tweets might be seen as less engaging by the platform's metrics

What publishers are experiencing

• Declining referral traffic from X
• Lower engagement rates on link-containing posts
• Pressure to adapt content strategies for the platform

The broader context This finding comes as news organizations increasingly question the value of social media platforms for distribution. With Meta's recent moves to prioritize AI-powered content and X's ongoing evolution under new ownership, publishers are reassessing their social media strategies.

The analysis suggests that X may be becoming less hospitable to traditional news sharing, potentially forcing publishers to either adapt their content or seek alternative distribution channels. This could have significant implications for how news reaches audiences in the future.

For now, the data indicates that publishers may need to reconsider their approach to X, potentially focusing more on native content or exploring other platforms that better support link sharing and external traffic generation.

Link to original analysis