Frustration signal
A frustration signal is any behavioral pattern that indicates a user is struggling, including rage clicks, dead clicks, error clicks, thrashing cursors, rapid back-and-forth navigation, and form field churn. Individually each is noisy; combined and scored, they rank sessions and pages by how much difficulty users are experiencing.
How it works
Detection rules run over event streams and emit signal instances with a session, timestamp, target, and type. Scoring aggregates them, usually weighting by signal reliability (error clicks are near-certain problems, cursor thrash is suggestive) and normalizing by session length. The output serves two consumers: humans, as a "most frustrated sessions" queue worth watching, and models, as labeled moments that anchor AI summaries of what went wrong. Because all the underlying signals derive from standard recorded events, frustration scoring works retroactively across a recording archive.
Where you encounter it
Session prioritization, UX health dashboards, alerting, and as the backbone of AI-assisted frustration detection.