David Nothling-Demmer brought home the best line from SuiteWorld 2026. Asked what founders should do about AI, NetSuite Founder Evan Goldberg said to work out what makes customers choose you, protect it, use AI to make it better, “and then change everything else”.
It’s good advice. For sales-led private companies in New Zealand and Australia, it’s also half a plan. The missing half is a number.
Here’s why that matters. In a sales-led business, the special sauce usually isn’t in the back office. It’s the relationships, the way your best people read a customer, how fast a quote goes out, and the trust you’ve built over years in a market where everyone knows everyone. That’s revenue. Once you start using AI to “make it better”, you’ve stopped tidying the admin and started working on the engine that pays the bills.
That calls for a higher standard of proof.
A measurement gap, not an AI gap
Talent Group’s AI in the Workplace 2026 report surveyed 1,505 leaders across Australia and New Zealand. It found 53 percent of organisations now have AI in their workflows or built into their strategy, up from 13 percent a year earlier. Only eight percent are measuring AI’s impact with clear metrics or KPIs.
It’s worth being careful with that second number. It doesn’t mean only eight percent are getting value. Plenty of the rest probably are. It means most of them can’t show it. Adoption has taken off, and measurement hasn’t caught up.
In my experience, that’s where AI work quietly stalls. The technology usually isn’t the problem. The problem is that nobody agreed at the start what success would look like. Then, when the board asks “what did we get for that?”, the honest answer is a shrug and a few stories about time saved.
Time saved isn’t a result, though. It’s an input. The result is what your people do with the hour they got back.
Put a number on the special sauce first
Goldberg’s filter is “what makes customers choose you?” The commercial follow-up is “how would we know if it got better?”
For a sales-led company, the answer is usually a number you already track, or should. It might be how quickly a new enquiry gets a real response, how many quotes become orders, the margin on an average deal, how long a new salesperson takes to hit target, or how many good customers come back.
Pick one. Measure where it sits today, before anyone switches anything on. Then write down what change would make the investment worth keeping, and what result would make you stop.
That’s not a framework. It’s the same discipline any good owner already uses for a new hire or a new sales territory. We just tend to forget it when the new hire is software.
Start small, but make it measurable
David’s piece argues for starting small and putting people first, and I agree on both. Steve Fox of Every Man Jack summed up his approach as “train, transform and track”. In my experience, track is the word most businesses skip.
A small pilot with one clear revenue number does more good than a big rollout with none. The team gets a scoreboard they can actually see, which does more for adoption than any change programme. The board gets something it can govern. And you, as the owner, get the confidence to scale what works and drop what doesn’t before it becomes a line item nobody wants to question.
The question for the board
For private company boards, the shift is simple. AI shouldn’t show up as an IT item or a promise of cost savings. It should show up like any other revenue investment: here’s the customer outcome we’re protecting, here’s the number we expect to move, and here’s how we’ll know by next quarter.
Directors don’t need to understand the models. They need to ask one question every time AI comes up: what number will this move, and how will we know?
Owners who can answer that will find AI spending easier to justify and their teams easier to bring along. Their special sauce will be safer for it, too.
So protect what makes customers choose you, change everything else, and put a number on it. In a market this size, the winners won’t be the businesses that used the most AI. They’ll be the ones that can show what it earned.
