Making video calls self-organized and inclusive

The Covid-19 pandemic highlighted the significance of digital exchange as a countermeasure to loneliness and a catalyst for activism. Spontaneous initiatives tried to involve individuals who were new to the digital landscape. However, at that time, current tools fell short in facilitating lively conversations and accommodating participants who lacked experience in digital spaces or just exhibited introverted tendencies.

Video calls happening outside hierarchical work contexts revealed a fundamental problem: existing tools were designed for corporate meeting culture, not for the self-organizing dynamics of informal gatherings. We missed the dynamics of a self-organizing crowd hanging out at one big table together: The big table in a bar, where so many discussions, jokes, comforting talks, utopias, and ideas happen.

I initiated the project and took on a product owner and facilitation role: keeping us on track with our collectively created vision, managing the backlog, moderating planning sessions and agile rituals, and facilitating feature and scope decisions. I conceptualized, planned, and conducted design research throughout the project and held responsibility for the budget.

Since we had no boss or business owner, we all owned the project. This fundamentally changed how we prioritized and what facilitating decisions meant. Rather than directing, I was coordinating among equals, and we are proud that despite this collaborative structure, we stayed on time and on budget.

User segment & pain

Primary users were people organizing or participating in digital social gatherings—especially those new to digital spaces or exhibiting introverted tendencies. The core pain: Existing video call tools forced hierarchical conversation patterns where assertive individuals dominated while quieter voices faded into the background or exited altogether. Users couldn’t figure out who would talk next, couldn’t understand anything when everybody talked at once, and had no way to have side conversations without disrupting the entire group or switching to separate chat tools.

Evidence at start

During the first SARS-CoV-2 shutdown, team members started a virtual bar and experienced firsthand how informal group conversations felt restricting and hierarchical. We observed that current tools failed to replicate the self-organizing dynamics of hanging out at one big table together.

Research challenge

We needed to validate whether spatial audio could create intuitive interaction patterns that enable informal, unmoderated video calls and understand how users would naturally orient themselves in virtual space.

Key research questions included: Do users understand the concept of a virtual space with dynamic sound? Do users understand that they can control what they are hearing by dragging themselves to the source? Can users imagine that this feature could improve video calls? Is it fun to use? (Full research documentation)

Methods

We chose rapid, qualitative methods to iterate quickly on interaction design before investing development resources: Lo-fi prototyping with Axure, Wizard-of-Oz testing to simulate technical functionality, qualitative user interviews, iterative prototyping, and dogfooding (forcing ourselves to use and improve our own tool).

We deliberately avoided quantitative methods in the first phase because we needed to clarify the “how” and iterate rapidly on design. Furthermore, we conducted our first user test after just 4 working days. In the second project phase, we added automated scalability testing while maintaining direct user feedback loops.

The concept of virtual space with dynamic sound is understood and approved
Evidence: All test users grasped the spatial metaphor within minutes of exploration
This validated our core hypothesis—we could proceed with spatial audio as the foundation.

Proximity-based volume is sufficient; stereo isn’t necessary
Evidence: Shadow testing with manual volume changes created the intended spatial experience
This led to a major architecture decision that simplified technical implementation while preserving our intended user benefit

Solution

Chatmosphere is an open-source video call platform that recreates the organic dynamics of in-person gatherings. Its spatial audio design adjusts volume based on proximity—as participants move their video feeds closer together, their voices become louder. This enables natural self-moderation and seamless transitions between parallel conversations within a single session.

To illustrate this principle, the circular video representation of users includes a surrounding circle indicating the maximum audible range. The interaction paradigm allows freely positioning one’s video representation using drag and drop.

• Test chatmosphere live
• Fork our project on GitHub
• Prototype Fund Round 1: Chatmosphere (in German)
• Prototype Fund Round 2: Skalierung (in German)

Open Design for Chatmosphere

During the Second funding round, we experimented with how to make the design assets open as well. By using PenPot as an open source alternative to Figma, as well as choosing open source design assets, we could prove that open source can also work in Design.
You can find our assets in our open design documentation.