What is Journey Tracker
Journey Tracker is for measuring how conversations move through a named flow. You define the flow as a template — what event starts it, the milestones along the way, and what event finishes it — and Atender attaches a tracked journey to every conversation that matches the start trigger.
Think of it as a funnel definition you can layer over conversations: a Returns journey might start when a customer says “I want to return” and end when the return label is sent, with milestones at Reason captured and RMA issued. A New customer onboarding journey might start on first contact and end when CSAT comes back positive.
The point isn’t to automate anything — automations do that. The point is to measure where conversations slow down or stall, so you can spot the milestones that take too long and the ones conversations drop out at.
The pieces
Every journey template has four parts:
Start trigger. An event that opens a journey for a conversation. Common starts: Conversation Created, Conversation Tagged, Message Received. You can add conditions to narrow it — for example, only start the journey when the channel is email and the conversation tag is returns.
Milestones. Ordered checkpoints between start and end. Each milestone is itself a trigger (event + optional conditions) — when the event fires for a conversation in the journey, that milestone is marked reached. Milestones aren’t required; you can model a journey with just a start and end if you only care about total time. They aren’t strict either — a conversation can reach a later milestone without passing earlier ones, but the data will show that.
End conditions. One or more events that finish a journey. Common ends: Conversation marked as done, CSAT Rating. As soon as any end condition fires, the journey closes.
Timeout. Optional. If a journey is still open after N hours, it auto-closes as “timed out.” Useful for journeys that should resolve quickly — anything still open after 48 hours is, in practice, abandoned.
What you can use as a trigger
Journeys read the same event stream as automations. The trigger picker exposes about two dozen event types — conversation lifecycle events (created, assigned, status changed, marked done, snoozed, reopened, escalated), message events (sent, received), AI events (response generated, handoff), agent events (first response, presence changes), SLA events (breached, met), CSAT events, side-conversation events, and call events.
Each trigger can have a list of conditions on standard conversation fields (channel, status, tags, agent, team) or on event metadata. Conditions support operators like equals, not equals, contains, starts with, in.
Enable / disable
Each template has an Enabled toggle. When a template is disabled, no new journeys are started against it — but in-flight journeys keep tracking until they hit an end condition or timeout.
What journeys aren’t
Journeys are read-only observers. They don’t trigger anything, they don’t route conversations, and they don’t enforce that a conversation reaches each milestone in order. They’re a measurement layer. If you want a milestone to do something — assign a team, send a message, mark a tag — that’s a job for an automation rule.