X to shut down Communities feature in May 2026

· Source: Dataconomy · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Emerging Technologies & Innovation, Data Science & Analytics · Depth: Fundamental Awareness, quick

Summary

X will discontinue its Communities feature by May 2026, as confirmed by X Head of Product Nikita Bier. The feature, designed for interest-specific groups, achieved less than 0.4% user adoption and contributed to 80% of spam reports, financial scams, and malware on the platform. This low engagement and high resource consumption, taking up to half the team's time, negatively impacted X's overall performance. While some groups were active, they often served as acquisition channels for other platforms like Kick, deviating from the original intent. To facilitate transition, X introduced joinable links for group chats on its XChat platform on April 22, 2026, allowing up to 350 members, with plans to expand to 1,000. Moderators can pin these links in Communities before the May 30, 2026, retirement.

Key takeaway

For product managers evaluating feature performance, X's decision to sunset Communities highlights the critical importance of user adoption metrics and resource efficiency. If a feature consumes significant development time and contributes disproportionately to platform issues while seeing minimal engagement, it's a strong candidate for deprecation. Focus on features that demonstrate clear value and align with core platform goals, even if it means retiring those with "great vision" but poor execution.

Key insights

Low user adoption and high spam rates led X to discontinue its Communities feature.

Principles

Method

X is transitioning users from Communities to XChat group chats and custom timelines, using joinable links and Grok-curated feeds to maintain community organization.

In practice

Topics

Best for: Product Manager, AI Product Manager, Tech Journalist, Director of AI/ML

Related on AIssential

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Dataconomy.