ResearchPod Summary
Content creators on social media platforms often face disproportionate levels of online harm, such as hate speech. While existing research focuses on how creators use moderation tools for safety, this study investigates how they navigate the trade-offs between safety, impression management (maintaining a desired public image), and algorithmic visibility (gaining platform rewards). The researchers conducted a survey of 584 content creators to understand how they evaluate three specific moderation strategies: deleting, replying to, and ignoring harmful comments.
Creators evaluate these three strategies differently across the three dimensions of safety, impression management, and algorithmic incentives. Deleting was consistently rated as the most effective for achieving emotional safety. Regarding impression management, both deleting and replying were perceived as more beneficial than ignoring, as they allow creators to actively shape their public-facing environment. Interestingly, while creators perceived replying as the most beneficial strategy for algorithmic visibility, this perception did not significantly influence their actual decision to use it. Instead, the willingness to adopt a specific strategy was primarily driven by the expected benefits to emotional safety and impression management.
This study highlights that creators act as 'middle-level governance' actors who manage their own digital spaces. Their moderation choices are not merely defensive; they are strategic actions that balance personal well-being with the need to maintain an audience-friendly environment. The findings suggest that platform designers should move beyond simple safety tools and consider how moderation features impact a creator's public image and their ability to interact with their community, rather than assuming that algorithmic incentives are the primary driver of creator behavior.
AI-generated third-party summary by ResearchPod. Not official content or an endorsement by the paper authors or affiliated organizations.