X tweaks its algorithm to make social connections matter again

X is adjusting its recommendation system to favour reciprocal relationships, signalling that pure reach and engagement may not produce the kind of network users want to participate in.

What happened

The platform changed its algorithm to give more visibility to mutual connections—accounts that follow each other. Product head Nikita Bier said reply sections had become dominated by unfamiliar users and that stronger weighting for reciprocal relationships could help interest-based communities form more naturally.

The change does not remove recommendations from outside a user’s network. Instead, it changes the balance between discovery and familiarity by treating an existing two-way connection as a stronger signal of relevance.

For creators and brands, the adjustment could affect which replies receive attention and how posts spread. Accounts that have built genuine communities may benefit, while engagement tactics designed primarily to reach strangers could become less effective.

Why it matters

Recommendation systems shape the social character of a platform. An algorithm optimised heavily for reactions can surface conflict, provocative replies and viral content because those behaviours generate measurable engagement. That may increase time spent in the short term while making users less willing to post or participate over time.

Reciprocal connections offer a different signal: they suggest an ongoing relationship rather than a one-off interaction. Giving that signal more weight could improve conversation quality, although it also risks reinforcing closed networks and reducing exposure to new perspectives.

The bigger picture

Social platforms are reconsidering the trade-off between discovery and trust. The next phase of algorithm design may focus less on maximising every post’s possible audience and more on creating smaller, durable communities.

X’s change is a relatively simple intervention, but it tests an important idea: a platform can improve retention not only by showing more compelling content, but by making users feel that the people around them are recognisable and relevant.