As Genially scaled, real-time collaboration shifted from a feature problem to a platform-level reliability and trust challenge.
Multiple users editing the same content at the same time led to:
• Conflicts between simultaneous edits
• Lack of visibility about who was doing what
• Uncertainty during uploads and shared actions
• Reduced confidence in collaborative sessions
The existing collaboration experience didn’t scale to support real-time, simultaneous editing—especially in high-density scenarios with multiple active users.
As Genially grew, collaborative sessions became harder to trust: users lacked visibility into who was doing what, whether actions were still in progress, and if the system was in a stable state.
The shift from isolated actions to shared feedback reduced uncertainty and enabled confident real-time collaboration.
We defined a set of platform-level patterns to ensure clarity, trust, and predictability in high-concurrency editing scenarios.
Key patterns:
Live presence — cursors, labels, and selection states clarify who is acting.
Shared object states — predictable behavior when multiple users interact.
Upload transparency — clear feedback during async actions.
These patterns reduced conflicts and enabled predictable, trustworthy collaboration in high-density sessions.
We designed real-time co-editing around clear ownership, making it explicit who controls changes and when actions are in progress.
By surfacing editor attribution, ownership states, and in-flight actions, the system prevents conflicts and enables confident parallel work in high-concurrency scenarios.
During user testing, we observed a recurring breakdown in trust when collaborators became inactive without clear system feedback.
Participants were unsure whether an editor had intentionally stopped, lost connection, or was still holding control—leading to hesitation, duplicated actions, or unnecessary interruptions.
What we added:
Inactivity detection signals to infer disengagement over time
Clear visual state changes when an editor becomes inactive
Graceful ownership release to unblock collaboration without manual intervention
This iteration closed a critical trust gap that only surfaced under real usage, reinforcing the system’s reliability in long-running, high-concurrency sessions.
Key indicators based on early adoption and qualitative user feedback.
Live presence indicators and selection states significantly improved users’ understanding of who was doing what, in real time.
Clear ownership signals and shared object states reduced accidental overlaps and conflicting edits.
Explicit progress and system feedback reduced uncertainty during async uploads and background actions
Real-time collaboration only works when users can instantly understand who is acting, what is happening, and what state the system is in.
Simulating real concurrency early revealed edge cases invisible in static flows—and shaped more resilient interaction models.
Accessible colour, iconography, and spacing reduced friction and made collaboration more intuitive across skill levels.
Presence indicators, progress signals, and previews helped users stay confident and in control during live collaboration.