Is the baseline recorded and stored?
Confirm the two-week baseline exists: hours by task, outcomes in the Targets' units, quality measures with their methods, rates from finance, and a line of caveats, stored with the date and the names of the people who logged it. Without it, nothing done later can be compared.
Verify: open the stored sheet and check it has all five parts and a date before today.
- 1. Baseline sheet stored, dated, five parts present
Are the connections dedicated, scoped, and confirmed?
Confirm every connection in Setup is a dedicated account or token, not a person's login; carries the narrowest permissions the remit needs at the tool; is scoped to one property; and has been confirmed at the tool by security with a name and date. Confirm any connector link for an external assistant was pasted where it belongs and not stored elsewhere.
Verify: read the connection list; ask security for the confirmation; try a forbidden action at the tool with the account and see it refused.
- 2. Dedicated accounts only 3. Narrowest permissions at the tool 4. Scoped per property 5. Security confirmation dated
Has the remit been tested, and does every escalation have a receiver?
Confirm the remit's three lists were tested against last month's real cases with the people who handled them, that the ambiguities the test surfaced were resolved, and that every category on the always-escalate list has a named receiver who has confirmed they expect a four-part summary. Confirm the never list is written and the targets have counterweights and evidence sources.
Verify: the remit in Setup shows the lists; ask each receiver whether they know what is coming.
- 6. Remit tested against real cases 7. Every escalation category has a confirmed receiver 8. Targets have counterweights and evidence
Has the Brand Brain been reviewed and corrected?
Confirm the manager read every item in the Memory tab, corrected what was wrong, removed what should not be there, added the tone and claim rules, and confirmed or replaced every inferred item. Confirm nothing personal about individuals is in it. A wrong item here becomes a sourced-but-wrong claim the fact gate cannot catch.
Verify: every item shows a provenance chip and a confirmed state.
- 9. Brand Brain reviewed, every item confirmed
Are the budget, model tier, and autonomy set?
Confirm the per-FTE budget is set inside the wallet, any external spend cap is set at the platform that bills it, the model tier is chosen with its honest label for the role's task mix, and every class of action is at Level 0 on the autonomy ladder with the always-escalate list marked fixed. Confirm approvers are routed per category so routine items reach the team member closest to the work.
Verify: Setup shows the budget, the tier, every class at Level 0, and an approver per category.
- 10. Budget and cap set 11. Model tier chosen; every class at Level 0; approvers routed
Has the manager read the first brief?
The last item is the first act of management: the morning brief the Digital FTE writes on its first day, read in full. It will say what it plans, what it is waiting on, and what it spent, which on day one is nothing. Reading it establishes the habit that the rest of the delegation depends on, and confirms the brief reaches the right person in the workforce view.
Verify: the brief is in the Today tab, addressed to the named manager, and the waiting count in the header matches it.
- 12. First brief read by the named manager
How do you use this checklist?
Work through it on the go-live morning with the person who will manage the role and, for items 2 to 5, the person who owns access. Tick each item only after performing its verification, and record the name and date beside it in the procurement or governance file. An item that cannot be verified is not ticked, and the first proposal waits until it can be. The whole list takes an hour if the preceding weeks did their work.
Keep the completed list; the quarterly governance audit reads it against what the workspace shows three months later.