Bot Ownership and Escalation Map (Handoff)

Bot Ownership and Escalation Map (Handoff)

1. 01

   Name a single accountable owner for every bot before the client sees a dashboard

   Assign one named person on the agency side and one on the client side. Shared ownership across a pod is how a broken invoice bot runs unnoticed for a week.

2. 02

   Write the failure taxonomy the client will actually use when something looks wrong

   Separate three buckets: bot did nothing, bot did the wrong thing, bot did the right thing on stale data. Each bucket routes to a different responder.

3. 03

   Set the escalation clock in hours, not business days

   A rule-based workflow that stalls mid-month corrupts downstream reporting. Define a 4-hour first response for revenue-touching bots and 24 hours for internal ones.

4. 04

   Document the manual fallback for each automated step

   If a UI-level bot breaks because a portal changed its layout, the client needs a written procedure to complete the task by hand until the selector is repaired.

5. 05

   Log where human review gates sit inside the workflow

   Autonomous systems now trigger actions without a person in the loop, so mark every step that moves money, sends client-facing messages, or files a compliance record.

6. 06

   Hand over credentials through a vault, never a shared document

   Rotate any login the bot used during build. Record which accounts the automation holds and who can revoke them.

7. 07

   Schedule the 30-day ownership review as part of the retainer

   Confirm the named owner still holds the role, the escalation clock held during live incidents, and the fallback procedure was tested at least once.