Simple contagion drives population-scale platform migration

A team of researchers from ETH Zurich and Meta has published a landmark study providing the first population-scale causal evidence of how social media platform migrations actually occur. By linking 276,431 scholars on Twitter/X to their new profiles across the entire universe of 16.7 million Bluesky accounts from January 2023 to December 2024, and using Brazil's court-ordered suspension of Twitter/X as an exogenous shock, the researchers demonstrated that migration is driven by 'simple contagion'—meaning adoption spreads through direct peer influence rather than requiring complex coordination or threshold mechanisms. This challenges prevailing theories in network science that emphasize complex contagion for significant behavioral shifts.

The study's key technical achievement was a scalable, high-precision cross-platform matching pipeline that enabled this causal analysis. Three clear patterns emerged: adoption concentrated among users deeply embedded in Twitter's social graph, public political expression predicted migration (consistent with homophilous inflows to Bluesky's left-of-center information space), and early reconnection with prior contacts predicted longer tenure and engagement. The findings recast platform migration not as a full exit but as a 'multi-homing' strategy users employ to insure against governance uncertainty on any single platform. This research provides a crucial empirical framework for understanding future platform transitions in an increasingly fragmented digital landscape.