Why are users abandoning my agent?
Users abandon when the agent fails them in a way your logs call a success: an unanswered intent, a hallucinated answer, a loop, or a slow response. Flowlines detects abandonment as the session_abandon signal, ties each one back to the intent and failure that preceded it, and shows which cohorts abandon most.
Abandonment is the clearest behavioral vote a user can cast: they left mid-task. But the abandonment itself isn't the problem, it's the symptom. The question is what happened in the turns before it.
Flowlines fires session_abandon on premature exits and correlates each one with the intent the user came in with and the signals that fired during the session. The pattern is usually legible: a cluster of abandonments concentrated on one unmet intent, or following a specific failure mode.
Rolled up by cohort, you see whether abandonment is broad or concentrated in your at-risk users, and tied to versions, whether a specific deploy made it worse.
Last updated 2026-05-28