Making it stick past week one

The setup is not the hard part.

Week two is, when the novelty fades and people drift back to the tools they already know.

People stall when the system feels like extra work added on top of their real work.

Three moves keep that from happening.

Move 1 — One real win per person on day one

Adoption takes or it does not in the first useful output.

Skip the practice task.

A toy problem teaches nothing that survives their actual job, and it tells them the system is a training exercise instead of the tool they work in.

Put each person’s first session on something they genuinely owe this week — a real email, a real report, a real client they are preparing for.

I produce something useful that day, and they learn by getting their own work done.

The goal of week one is one sentence from each person: "it did my work, and the work was good."

Your part is to clear the path: make sure each person knows which domain handles their work and walks in with a concrete thing to bring.

Say to me, for each person: "Patrick owes a renewal proposal this week — set his first session up on that, in the sales domain, and tell me what he should bring."

Move 2 — The daily brief is the wedge

One win gets them in.

The daily brief is what brings them back tomorrow.

The brief is a saved routine the system carries: every morning I read their email, calendar, and open work, then hand them a short read of what matters today.

Type /brief and it runs.

That is the habit that makes the system the first thing they open, not a thing they remember to use.

A person who starts the day in the brief stays in the system all day; a person who has to remember to come back forgets by Thursday.

Set the brief as the standing first move of the morning for everyone you onboard.

Say to me, for each new person: "Set Mike up so her first action each morning is the daily brief — and show her how to run it."

Move 3 — A champion on every domain

A domain with no one curating it goes stale, and stale is where adoption dies.

When no one feeds the knowledge, I stop getting better at that work, and the team quietly concludes the system is not improving — which, in that domain, is true.

Champions are your domain owners, and they are your multipliers.

A good owner pulls their whole function onto the system because it keeps getting better at their function — and it keeps getting better because they keep curating it.

You assign them in the roster, the recorded list that pairs each name with the domains they own, taught in Lead-3.

Naming an owner there is the same act as granting them the access to curate that domain.

Resource them with the time to own it — an owner named on a slide but given no hours is the same as no owner.

Say to me: "Confirm every live domain has an owner in the roster, and flag any domain that is named but unowned."

Credit the people who feed it

The system compounds only if people feed it, and people feed what gets noticed.

Every correction, every piece of feedback, every saved output lands in three always-on logs the system carries — a feedback log, a decisions log, and a roadmap — that an owner curates into real improvement.

So the contribution is already recorded; your job is to make the recorded act a credited one.

When someone corrects their agent, flags a workflow that keeps missing, or saves a good output as reusable knowledge, name them and credit them.

The message is that improving the system is part of the job, not a distraction from it.

Say to me: "Show me who has been feeding the logs this month — corrections, feedback, saved knowledge — so I can credit them by name."

You will know it has taken when

People reach for their agent first.

Every domain has an owner curating it.

New requests arrive as feedback, not complaints.

And the agents are visibly better this month than last.

The one thing to remember

Adoption is three moves you hold through week two: one real win per person, the daily brief as the standing first move of each morning, and a named champion on every domain who curates it and credits the people who feed it.

Each one targets the same failure — the system feeling like a thing on the side instead of the way the work gets done.

Next: reading the compounding