Hashtags of hope: technology, stakeholders and disaster response

This research examines how individuals exercised digital agency by reaching out to emergent actors on Twitter  (now called as X), and how this reshapes understandings of stakeholder salience in humanitarian supply chain management (HSCM) during disaster response for humanitarian organisations. A corpus of 120,057 tweets collected via hashtag-based Application Programming Interface extraction is analysed through three complementary methods: (1) descriptive frequency analysis of stakeholder callout patterns, (2) social network analysis (SNA) and (3) topic modelling to establish the purposive, humanitarian intent of the identified interactions. Affected populations directed the majority of callouts to celebrities, exceeding appeals to government, media and NGOs. Accounts of the top celebrities structurally bridged disconnected stakeholder clusters. Finally, the interactions were found to be purposive and humanitarian, spanning direct appeals, donation mobilisation, information diffusion and volunteer coordination. Humanitarian organisations should establish pre-crisis partnership agreements with verified celebrity amplifiers and invest in specialist digital intelligence intermediation units, structurally separate from operational command, to filter, verify and route social-media-sourced signals to field commanders. The study contributes to debates on inclusive, participatory disaster governance and the use of technology to reduce information asymmetry that systematically disadvantages the most vulnerable. This study makes two original contributions. First, it extends the Power-Legitimacy-Urgency framework by introducing network amplification power, a normative, relational power modality distinct from institutional command authority, thereby formally incorporating technology-mediated, non-institutional actors into stakeholder theory. Second, it grounds this theoretical extension in multi-method empirical evidence, combining SNA, latent Dirichlet allocation topic modelling and inferential statistics on a large-scale, real-world corpus of disaster communication.