Behavior signals
Behavior Signals
Behavior signals are interpreted patterns — hesitation, repetition, dwell, missed affordances — extracted from raw XR telemetry to describe what a user actually experienced.
Where raw telemetry produces millions of events, behavior signals compress them into decisions teams can act on. Gossip Analytics uses behavior signals instead of granular eye-tracking metrics so privacy and clarity stay aligned.
A behaviour signal is always defined at the experience level, not the event level. 'Hesitation at object X' is a signal; 'gaze direction sample #4213' is not. The interpretation layer is what makes signals usable across disciplines: designers, QA, product managers and accessibility researchers can each act on the same signal from their own angle.
Why it matters
Behaviour signals are what let XR analytics be shared beyond engineering. A raw telemetry stream is unreadable to non-technical roles; a signal like 'users hesitated at the door in 62% of sessions' is immediately actionable. Signals are also what allow XR analytics platforms to compare experiences across studios, headsets and release versions.
Example
A location-based VR operator sees a repeated 'stuck at scene 3' signal across venues. Instead of debugging the code, the team reviews the signal's underlying pattern (users repeatedly walked into a virtual wall) and repositions the doorway. The signal disappears in the next release.