Skip to main content

​
When to escalate

Escalate when:
  • The customer is upset or the topic is sensitive (e.g. billing, complaint).
  • The conversation needs a human or a specialist.
  • SLA is at risk or breached and a supervisor should be notified.

​
How to escalate manually

  1. Open the conversation in Inbox or Support → Conversations.
  2. Change status to Escalated (or your org’s equivalent).
  3. Assign the conversation to a supervisor or the right team.
  4. Optionally create a task (e.g. “Human follow-up”) and assign it so someone is explicitly responsible.
  5. Add a tag (e.g. “Escalated”, “Supervisor”) so you can filter and report.

​
Automatic escalation with workflows

You can use Workflows so that when a conversation is marked Escalated (or when SLA is at risk):
  • A supervisor is notified (e.g. email or Slack).
  • A task is created (e.g. “Escalated – follow up”) and assigned to a team or person.
  • The conversation is tagged for reporting.
Configure these workflows in AI Hub → Workflows (or your Workflows list). Triggers might be “Conversation status changed” or “SLA at risk”; actions might be “Notify supervisor”, “Create task”, “Tag conversation”.

​
SLA at risk

If your organization uses SLA (e.g. first response time, resolution time):
  • Workflows or alerts can run when a conversation is at risk or breached.
  • Actions can include: notify supervisor, create task, assign to a dedicated team, or tag. Check Workflows and Settings for SLA and notification rules.

​
After escalation

  • Supervisors or assigned agents handle the conversation, reply, and update status (e.g. back to Pending, then Resolved when done).
  • Use Analytics to track escalation rate, response times, and SLA by channel or team.