Leaders Ledger · Rachel McBride
← Working frames · Rowan Advisory ↗
Leaders Ledger #1: Clear, and stuck
Capable leaders. Real pilots. Real effort. Still stuck on AI. After 53 conversations across New Zealand, four useful moves before the organisation is fully aligned.
During April and May 2026 I sat alongside 53 New Zealand leaders. They came from charities, small and medium-sized businesses, larger organisations and the public service. Almost all were saying a version of the same thing: I know we should be doing something with AI. I have no idea what that means for us.
They were not short of curiosity or effort. Many had done training. Most had pilots or experiments underway. Several were further ahead personally than their organisations were.
Two structural problems kept appearing. Senior teams did not have shared language for the workforce, strategic and operational choices involved. And their measures and incentives still rewarded the old work, done efficiently.
What can a leader do now?
Most of the 53 leaders did not have the authority of a founder with a fully aligned board. They had day jobs, sceptical peers and boards still working through privacy, risk and security. That raised a more useful question than “what is our AI strategy?”
Four moves appeared across the conversations.
Stop
Do not start with tools. Start with your position: what AI is for, where people must remain in charge, and what you will not do. Use that position to choose the work worth changing and to refuse the rest.
The aim is not more AI use. It is better work.
Build
Build something small and real. It might be an agent at home, a way to organise personal finances, or a modest work experiment inside your authority.
Building teaches what reading and training cannot. You learn how much iteration is required, where the output fails and where your own judgement matters. A small failure is useful because the cost is contained and the lesson is yours.
Own
Stay the person who decides. Let AI draft, research, compare and argue both sides. Keep the judgement and accountability with you.
This matters because a confident answer can still be wrong. The more fluent the system sounds, the more deliberate the human review has to be. Owning the work means being able to say where you stepped in, what you checked and why you made the final call.
Show
Find something that delivers real value and show it working. A demonstration gives people direct experience and creates a better conversation than another abstract argument about strategy.
You can usually show something small before you have a formal mandate. That does not replace the authority needed to change budgets, roles or rules.
What a mandate is for
A mandate matters. It lets a leader change budgets, roles, incentives and decision rights. Most of these 53 leaders did not have one.
The four moves are what remain available before that alignment exists. Stop, build, own and show create evidence. They do not confer authority.
The leaders I would back do not claim to have the answer. They have better questions, something real they have built, a clear account of what they refused to do, and evidence they can put on the table.
What did you stop?
What did you build?
What did you own?
What did you show?