Why I am spending twelve weeks writing this
I run an IT firm in Christchurch. We put AI into our own operation before we sold it to a single client. That is the only credential that matters for what follows.
Most AI content is noise. It is written by people selling tools they have never run in a real business, and it is very easy to spot once you know what you are looking for. It talks about capability and never about consequence. It describes what a tool can do and never what it cost to make it work on a Tuesday afternoon with real staff and real clients waiting.
This series is different because every week comes from a firm that put AI into its own operation first. Two years of internal work. Dual ISO 27001 and ISO 42001 certification. More than 100 New Zealand businesses served. 98 percent client retention across our Canterbury clients from 2022 to 2025.
In two years there will be two versions of your firm. The next twelve weeks are about making sure yours is the one that runs without you.
The promise
You will not need more AI hype by the end of this series. You will have a system.
The tool is the smallest part. The governance is what makes it last. A tool bought and never governed is a liability, not an asset, and we have seen the rebuild cost when it breaks in month six and no one has the documentation to fix it. That bill is always larger than the licence someone was trying to justify.
The gap between the firms that get this right and the ones that fall behind is opening now. AI capability compounds. Reputation compounds with it. A small head start eighteen months ago is a wide gap today, which is also why waiting is not the neutral choice it feels like.
What governed adoption actually means
Governed adoption is the idea this whole series sits on, so it is worth being plain about it up front.
It means AI running in your business under a documented system that still works when it breaks. Somebody owns each workflow. Somebody wrote down why it was configured that way. There is a defined answer for what happens when the model changes, when the person who built it leaves, or when a client asks how their data was handled.
The opposite is a cowboy rollout. It looks identical in month one, because the demo works either way. The difference shows up in month seven, and by then the cost of unpicking it is real.
None of that is exotic or expensive. It is mostly discipline about order of operations, which is the least glamorous advantage in business and one of the most durable.
Who this is for
New Zealand and Australian service business owners, directors and partners. 20 to 200 staff. Agencies, consultancies, law, accounting, IT, design, architecture, engineering.
Not for MSPs. Not for enterprise. Not for consumers.
If you are the bottleneck in your own firm and you know it, this series is for you. Most owners I speak to can name their bottleneck in about four seconds once somebody actually asks them. Knowing has never been the hard part.
What is coming up over the next twelve weeks
- Week 1: are you making more money from AI, or just spending more on it.
- Week 2: the AI Divide. Same firm, two futures.
- Week 3: reality is the moat. Why proof beats AI hype.
- Week 4: race to elevate, not replace.
- Week 5: the margin moat. Systems that protect margin.
- Week 6: virtuous versus vicious friction.
- Week 7: price to value. Move upmarket on credibility.
- Week 8: retention is the growth engine.
- Week 9: owner-independence. Every problem is a systems problem.
- Week 10: control the inputs.
- Week 11: commit. The Ulysses Contract.
- Week 12: the 10-year business.
How the twelve weeks build
The order is deliberate, so it is worth knowing where you are heading.
Weeks 1 to 3 are problem awareness. Name the divide and make it clear which side of it you are currently on.
Weeks 4 to 7 are mechanism. How it actually gets done. Elevate rather than replace, protect margin rather than just cut cost, drive adoption through incentives rather than instructions, and price to value.
Weeks 8 to 10 are proof and application. Retention, owner-independence, and controlling the inputs you can actually move.
Weeks 11 and 12 are about commitment. Deciding, in a way you cannot quietly reverse, which version of your firm will exist in two years.
The standard for the series
Every claim, every week, is something we have run inside our own firm or inside a real New Zealand client firm. No theory I have not run myself. No invented case studies. No borrowed authority, other than named references where I have credited them.
In a market where confident writing is free, that constraint is the only thing that makes any of this worth your time.
What to do next
Subscribe to the newsletter so each issue lands in your inbox. Week 1 asks the question almost nobody asks before they buy AI: are you making more money from it, or just spending more on it.
If you would rather skip ahead and work out whether governed AI would move your numbers, take the AI Readiness Audit or book a 15 minute call with me. If it is not a fit, I will say so.
