In an era where platforms like WhatsApp, Telegram, and YouTube dominate digital communication, understanding collective attention is important to analyzing societal impacts. This paper presents a comprehensive and interdisciplinary approach that combines computer and social sciences to explore patterns of collective attention in online chats. We apply conversation analysis and communication theories like Computer-Mediated Communication and Conversation Analysis, supplemented by psychological insights into attention, to study user engagement, response dynamics, and digital interaction nuances. Our three-pronged methodology addresses cognitive aspects of attention: alertness, processing capacity, and selectivity. We analyze conversation dynamics across platforms with distinct user behaviors, focusing on WhatsApp, Telegram, and YouTube chats. Political chats showed more prolonged periods of alertness and deeper engagement, especially on WhatsApp and Telegram, while non-political Telegram chats exhibited greater topic diversity. Moreover, YouTube’s rapid pace limited users’ processing capacity, leading to higher thematic variation. This study enhances our understanding of digital interaction and contributes to fields like computational social science by comparing how platform characteristics and content types influence attention and engagement. The insights from this research can be directly applied to improve user experience on various online platforms. We introduce a novel methodological framework to guide future research on online chats and collective attention.