Abstract
This research explores how algorithmic audiencing, the tendency of social media platforms to suppress and amplify messages for economic gain, affects natural disaster response. We study how official warnings and community-led efforts adapt under these dynamics by analysing social media discussions surrounding Ex-Tropical Cyclone Alfred. Legacy communications (e.g., an agency posting a warning on their Facebook page) provide dense information but yield low reach, engagement, and targeting. In contrast, short-form visual posts (e.g., Instagram reels from local creators) carry limited informational value and weak targeting yet achieve wide reach and high engagement. Bridging these extremes, community leaders (e.g., Facebook group admins for migrant communities) act as curators, embedding official information into hyper-targeted group networks largely immune to algorithmic audiencing. These findings advance our understanding of algorithmic audiencing in disaster contexts and offer guidance for practitioners seeking to retain control and predictability of warning messages while leveraging community-driven communication.
Recommended Citation
Pirasteh, Farnaz; Bahl, Rashika; and Marx, Julian, "Algorithmic Audiencing in Natural Disaster Response:
Evidence from Ex-Tropical Cyclone Alfred" (2025). ACIS 2025 Proceedings. 65.
https://aisel.aisnet.org/acis2025/65