Week one: what do you set up?
Week one with a Digital FTE is about connection and contract. Grant scoped access to the systems the role needs, confirm the targets it proposes or replace them with your own, and write down what must always escalate.
Spend the time on the escalation list specifically. It is the cheapest control you will ever configure and the one that prevents the incidents that damage trust in the first month. Name the topics, the thresholds, and the situations that must always reach a person.
- Connect the systems the role works in, with scoped permissions
- Confirm or replace the targets it proposes
- Write the escalation list: topics, thresholds, and situations
- Name the owner and the approver, and tell them
Week two: how closely should you review?
Week two is close review, and you should expect to send work back. Read the reasoning as well as the output, because the reasoning is what tells you whether the remit landed correctly.
Give specific reasons when you reject. A send-back that says not quite right teaches nothing; one that says this claim needs a source, or this tone is too formal for our audience, updates the standard. The specificity you invest here is what makes weeks three and four shorter.
Week three: what should you tune?
Week three is tuning, driven by the pattern in your send-backs rather than by impressions. Group the rejections from the first fortnight and look for repeats: the same missing element, the same misjudged tone, the same wrong priority.
Repeated send-backs for the same reason mean the remit or the guardrails are wrong, not that the system is failing. Fix the instruction rather than continuing to correct the output, because correcting output is work that never ends.
Week four: what is the first promotion?
Week four is the first autonomy promotion, and it should be narrow enough to feel almost trivial. Pick the single class of action you have approved most often without amendment, and promote only that class inside explicit limits.
Keep every other gate in place. The purpose of the first promotion is to establish the mechanism and prove it is reversible, not to release capacity. The release comes later, once the pattern of promote, observe, and adjust is familiar.
What should you not do in the first month?
Do not add a second Digital FTE inside the first month, and do not widen the remit mid-stream. Both make it impossible to tell what caused what, which is the same mistake that ruins pilots.
Also resist the temptation to stop reviewing once the output gets good. The approval queue is where you notice a drift in standard early; a month of unread approvals is how a small misunderstanding becomes a pattern.
What does a healthy first month look like?
A healthy first month shows a rising approval rate, falling send-backs on the same reasons, and a manager whose time has shifted from producing to deciding. Those three together mean the arrangement is working.
It should also show at least a few genuine rejections. A month with no send-backs at all usually means the reviews were not real, which is a problem to fix before promoting anything.
What should the Brand Brain look like by day thirty?
By day thirty the Brand Brain should read like a document a careful new hire could work from: every fact about the business, its products, its voice, and its rules confirmed or corrected by the manager, with each item's provenance chip showing whether it came from onboarding, a document, a connected system, or an inference from approvals. The inferred items deserve the closest attention, because they are rules the Digital FTE drew from patterns in your send-backs, and a rule inferred from three decisions may be an over-generalisation.
The first month is when memory errors are cheapest to fix, because every wrong item becomes a wrong proposal until corrected and the fact gate cannot catch a claim sourced from wrong memory. Review the full memory in week one, correct items as proposals reveal them through the month, and confirm the whole thing at day thirty as the first quarterly confirmation. A Digital FTE with a clean Brand Brain at day thirty produces proposals that are right in substance and need review only for judgement and tone, which is when the approval rate starts to rise on its own.