Which helpdesk, categories, and sources does the role cover?
Fill in the helpdesk the role works in, the ticket categories it resolves, starting with one, and the approved sources it may cite: the knowledge base, the policy pages, and nothing else. State that anything the sources do not cover escalates rather than being guessed at, which the fact gate enforces.
Verify: the connection in Setup is a dedicated account scoped to the named queues with permission to draft, not to send, and the approved source list contains only the named documents.
- Helpdesk: ______
- Categories resolved: ______ (start with one)
- Approved sources: knowledge base, policy pages, ______
What does the Customer Support FTE own, with a citation on every reply?
The Customer Support FTE owns first-line resolution for the named categories, with every reply citing the article or policy it used; triage and routing by category and urgency; knowledge base upkeep by proposing articles for recurring gaps, each claim sourced; and the escalation summary for every ticket it cannot resolve. It proposes; a person approves until a category earns auto-send.
Verify: in the first week, every draft in Approvals shows its citation, and every escalation arrives as a summary rather than a reassigned ticket.
Which categories always escalate, and to whom?
List the categories that reach a person every time regardless of whether the sources have an answer, each with a receiver. Refunds, credits, and billing disputes. Account access and identity. Complaints and requests for a manager. Legal, regulatory, or safety language. Any sign of distress, or a customer who has written more than twice. Anything the approved sources do not cover. Mark the list fixed.
Verify: each category has a named receiver who knows to expect the four-part summary, and the list appears in Setup as fixed.
- Refunds, credits, billing disputes → ______
- Account access, identity → ______
- Complaints, manager requests → ______
- Legal, regulatory, safety language → ______
- Distress, repeated contact → ______
- Not covered by sources → ______
What are the confidence triggers and the summary standard?
State the confidence triggers that escalate a ticket outside the fixed list: no source found, two sources that conflict, and intent unclear after reading. State the summary standard every escalation must meet: what the customer asked, quoting where wording matters; what the FTE searched and found; why it did not resolve, naming the rule or trigger; and a suggested next step, offered as a suggestion.
Verify: send back any escalation in the first two weeks that lacks one of the four parts; the standard is a remit rule.
What are the tone rules, the prohibited claims, and the never list?
Fill in the tone per channel, the phrases the brand uses and avoids, and the claims a reply may never make, such as promising a refund, a timeline, or an outcome the sources do not state. State what the role never does beyond the platform's list: never issues a credit, never changes an account setting, never replies to a complaint, never closes a ticket the customer has not confirmed resolved.
Verify: none of the never actions appears as a promotable class in Setup, and the tone rules are in the Brand Brain with provenance.
- Tone per channel: ______
- Never promise: refunds, timelines, outcomes not in the sources
- Never: issue credits, change settings, reply to complaints, close unconfirmed
What are the targets, and what counterweights each one?
Set three targets. First-line resolution rate for the named categories, counterweighted by escalation quality, meaning summaries that meet the standard. Median first response including out of hours, counterweighted by the send-back rate on replies. Knowledge gaps closed, meaning proposed articles approved, counterweighted by first-line resolution in the categories they cover. Each names its evidence source.
Verify: the Targets tab shows each with status and evidence; anything the helpdesk cannot measure, such as satisfaction without a survey, is marked unmeasured.
How do you use this template?
Copy the six sections into the first-run wizard, fill every blank, and test the remit against a sample of last month's tickets with the agents who handled them: every ticket should sort into resolved from a source, escalated by category, or escalated by trigger. Argued tickets become rules or list entries now. Read the send-back reasons weekly in the first month; the remit is finished at the first review.
The always-escalate list does not shrink when the first category earns auto-send. Add to it after any incident; remove from it only as a recorded governance decision.