Why do some teams thrive while others struggle, even when they have equally skilled members? Our research suggests that the difference often lies in how teams manage collective attention—the process of noticing, directing, and sustaining focus on what matters most at a given time.
In dynamic contexts such as open innovation, crisis response, or hybrid collaboration, teams cannot rely on rigid roles or pre-defined plans. Instead, success depends on members’ ability to guide one another’s attention in real time. We introduce the concept of an “architecture of attention”: the patterned ways in which teams use signals—such as tagging, highlighting resources, or marking milestones—to direct focus and coordinate action.
To examine this, we studied the OpenCovid19 initiative, where over 2,300 participants from 183 countries self-organized into 34 projects on Slack to design diagnostics, protective equipment, and other pandemic solutions. Analyzing more than 30,000 messages across 164 channels, we developed a measure of Attention Direction Ability—the efficiency with which behavioral signals triggered collective responses.
Our findings reveal that attention direction is not a single, static capability but a phase-dependent mechanism. In the early stages of a project, when teams are still forming and goals are fluid, efficiency in directing attention is most critical—well-placed signals can determine whether a project takes off or dissipates. In later stages, when tasks become clearer, it is the sustained volume of attention direction that keeps momentum and ensures follow-through.
These insights extend theories of team cognition and emergent coordination by showing that collective attention is not only an outcome but also a mechanism that can be deliberately shaped. For practitioners, the lesson is clear: effective teams design environments and practices that make the right signals visible at the right time. Whether in open science, startups, or corporate teams, building robust architectures of attention can help diverse groups transform potential into coordinated action.archy but by their ability to capture attention, share knowledge, and translate ideas into action.
We study these emergent communities using a mix of network science, computational modeling, and discourse analysis. Our aim is to understand the architecture of attention: how certain actions—like tagging others, initiating discussions, or sharing key resources—become signals that direct collective focus and shape participation. Early findings suggest that fluid roles, such as facilitators, admins, and project leaders, emerge through repeated action patterns rather than pre-defined positions, and that discursive diversity (the range of topics and voices) must be dynamically modulated to balance exploration and exploitation.
Beyond theory, this work informs the design of collaborative platforms: how to build onboarding systems, recommendation engines, and coordination tools that can support large-scale collective efforts. As we’ve seen during the pandemic, open science is not just an ethos—it’s an infrastructure, and understanding how it functions can help us prepare for future global challenges.
📄 Read more in our article in R&D Management: “Organizing an extreme crowdsourcing campaign to tackle grand challenges”
🗞️ Or in The Conversation: “Covid-19: the rise of a global collective intelligence?”
Interaction Data Lab 