Design Challenges and User Needs in Disaster Support Mobile Apps: A Large-Scale Analysis of Features and Reviews
Muhamad Syukron, Anuradha Madugalla, Mojtaba Shahin, John Grundy
TL;DR
This paper studies disaster support mobile apps to understand design challenges and user needs in real-world use. It analyzes 44 apps across Google Play and App Store and 28,161 reviews using hands-on feature extraction, BERTopic-based topic modeling, and VADER sentiment analysis to identify 13 core features across the disaster lifecycle and 18 user topics. It finds that while a majority of reviews are positive, significant negative feedback clusters around signup issues, connectivity, and app crashes, highlighting safety-critical design concerns. The work offers practical recommendations for developers and emergency agencies to improve reliability, accessibility, and user safety across lifecycle stages and geographies.
Abstract
Disaster support mobile apps play a central role during emergencies in conveying timely alerts to communities and providing recovery guidance. Their importance continues to grow as climate change increases the frequency and intensity of disasters, and as mobile devices become a primary information source for most people. However, despite their critical role, there has been limited systematic investigation into how these apps are designed and how users experience them in real-world contexts. This paper presents a large-scale empirical study of disaster support mobile apps aimed at addressing this gap. We investigated 44 apps available on the Google Play Store and the App Store and analyzed 28,161 user reviews to understand common functionalities, user expectations, and challenges encountered during use. Our feature analysis identifies 13 core capabilities spanning the stages of preparation, response, and recovery of the disaster lifecycle. Topic Modelling revealed 18 topics, with the highest discussions focusing on app alert functionality, the use of maps, and monetary transactions. Sentiment analysis revealed that while 53.09% of users provided positive feedback, 24.30% expressed negative opinions, and 11.25% remained neutral. It also showed that signup/signin issues, network problems, and app crashes were the most frustrating to users. These impacted user safety by preventing access when it mattered most. Our findings emphasize the need to approach disaster apps as safety-critical and human-centred systems, where reliability and ease of use are essential. We provide practical recommendations to support developers and emergency agencies in improving the dependability and accessibility of future disaster support apps.