Recommendation-driven platforms can reshape not only what users see, but also who becomes visible and how community boundaries are negotiated. We examine Xiaohongshu discussions surrounding a perceived World Cup-related audience influx to understand how users attributed community change and how those explanations related to boundary-management strategies. We analyzed 262 comments from eight high-engagement threads using researcher-led thematic analysis, structured coding, and permutation-based association analysis. Users described personalization as both social separation and a boundary that could fail, interpreted platform growth through audience acquisition and commercialization, and contested whether gender should define community membership. Among 70 comments with explicit attributions and strategies, attribution and strategy were strongly associated (Cramer's V = .583): algorithmic attributions concentrated in technical boundary management, while social-actor attributions more often aligned with social enforcement or normative socialization. These findings connect causal sensemaking with boundary management and show how audience restructuring can create governance consequences on recommendation-driven platforms.
Figures & tables
Figure 1. Overview of corpus construction and analysis. The study began with eight high-engagement Xiaohongshu discussion threads surrounding the perceived World Cup-related audience influx. After collecting 836 comment records and deduplicating them to 808 comments, relevance screening produced a final analytic corpus of 262 comments. The final corpus was analyzed through researcher-led thematic analysis and separate structured operational coding. The structured coding variables were then used in the quantitative association analyses. A study overview diagram. Public Xiaohongshu discussions surrounding the World Cup-related audience change were collected from eight high-engagement threads, producing 836 comment records and 808 comments after deduplication. Screening identified 297 candidate comments, while the first author manually reviewed all 808 comments. The final analytic corpus contained 262 comments, including 146 core relevant and 116 contextual comments. The final corpus was analyzed through researcher-led thematic analysis and operational coding of stance, attribution, and strategy. Operational coding supported reliability assessment and association analyses of attribution by strategy and attribution by stance.
Source
Focal discussion
Collected
Analytic
D01
Increase in male users
90
31
D02
World Cup investment
191
76
D03
Attracting male users
107
35
D04
World Cup and community change
30
14
D05
Arrival of male users
17
5
D06
World Cup audience expansion
42
18
Table 1. Overview of the eight source discussions and their contributions to the analytic corpus.
Figure 2. Distribution of broad boundary-management strategies across attribution families among comments containing both an explicit attribution and an explicit strategy ( N=70 ). Percentages are calculated within attribution families. Algorithmic attributions were concentrated in technical boundary management, whereas social-actor attributions were more often associated with social enforcement and normative socialization. The permutation-based association test was significant ( p<.001 , Cramér’s V=.583 ). The structural-platform category contains only six observations and should be interpreted cautiously. A three-by-four heatmap showing boundary-management strategy distributions within three attribution families. Among 41 algorithmic attributions, 38, or 92.7 percent, were paired with technical boundary management. Among 23 social-actor attributions, 12, or 52.2 percent, were paired with social enforcement and six, or 26.1 percent, with normative socialization. The structural-platform row contains six comments distributed across all four strategy families.
Complex Systems Institute of Paris Ile-de-France CNRS, Paris, France. · m´edialab, Sciences Po, Paris, France. · Learning Planet Institute, CY Cergy Paris University, Paris, France.