87% of New Zealand businesses are using AI. 88% are going nowhere.

No, the headline is not a mistake.

Those two numbers sit side by side in the latest research on AI adoption in New Zealand. Read them again slowly, because they tell you everything about where most businesses actually are right now.

Almost every organisation has started. Almost none have moved past the starting line.

The pilots are real. The decks are impressive. The partner/vendor presentations are frequent, polished, and full of promise. And yet the gap between “we are exploring AI” and “AI is working in our business” remains stubbornly, expensively wide.

This is not a motivation problem. It is not a budget problem. And it is not even a technology problem - the technology is extraordinary and getting better by the week. It is an execution problem. Most businesses are trying to figure out AI while simultaneously running a business. The transformation agenda does not pause the operational one.

AI doesn’t fix broken foundations. It magnifies them.

That is the uncomfortable truth sitting underneath most AI pilots right now. If your data discipline is inconsistent, AI will amplify it. If your processes are fragmented, AI will expose them. If your foundations are strong, AI will accelerate them. The technology reflects what is already there….. for better or worse.

The enterprise trap

And here is where it gets expensive. Enterprise AI solutions are everywhere right now. Every major vendor has a suite. Every platform partner has a roadmap. And the pitch is always the same: comprehensive, integrated, transformational.

The problem is that hard coding a sophisticated AI solution into a business whose data foundations are not ready does not accelerate you. It locks in the mess. You spend eighteen months (let’s be real 24-36 months) integrating a system that is only as intelligent as the data feeding it……and your data, like most businesses, has gaps, inconsistencies, and legacy structures that were never built with AI in mind.

Legacy systems are real. They are everywhere. And yes, they are genuinely limiting. But the answer is not always to rip them out and replace them with enterprise-grade infrastructure. That path is expensive, disruptive, and often unnecessary. The businesses moving fastest right now are not the ones who bought the biggest solution. They are the ones who identified the highest-value use case, strengthened the specific foundations that use case required, and built something that works,  before expanding from there.

Small, agile, staged. That is not a compromise position. In this market, right now, it is the smart one.

The ANZ reality nobody in the vendor decks talks about

Most of the AI case studies being presented to ANZ retailers were built in markets that bear little resemblance to ours. The economics of operating in New Zealand or Australia with the market size, the geographic spread, the cost to serve, the competitive intensity, and the ambitions that almost always have to be funded on a shoe-string - create a fundamentally different commercial reality to the one underpinning a Walmart or a Kroger (or even Kmart and Briscoes).

ANZ retailers are not lagging because they lack sophistication. They are making rational decisions inside a set of commercial constraints that most global vendors have never had to account for. A $400 million retailer in New Zealand does not have the data infrastructure, the IT resourcing, or the balance sheet of a $40 billion one in the United States. The ROI calculation is different. The risk tolerance is different. The definition of “scalable” is different.

What does not change is the opportunity. AI-mature retailers are projected to generate 2.3 times the revenue growth of their lagging competitors. That gap is not reserved for the big end of town. It is available to any business that gets the sequencing right and gets the right people in the room to make it happen.

The window to get the foundations right before the market forces the decision is probably 12 to 18 months.

The gap nobody is filling

There is no shortage of AI consultants. There is an acute shortage of people who understand retail commercially and complexity - the margin structures, the omnichannel complexity, the customer behaviour, the supply chain realities, the legacy systems and who can also build and implement AI that works in that specific operating environment, at the right scale, without destabilising the business in the process.

Most AI engagements follow the same pattern. A strategy is delivered. A roadmap is presented. The consultants move on. The internal team is left holding a document and a direction, but not the capability or the implementation support to make it real. The pilot quietly joins the others on the shelf. Or worse, you have to pay more (and more and more).

This is the gap rxcolab.ai was built to close.

rxcolab.ai is a specialist retail AI venture for ANZ, built in partnership between RX Group and thecolab.ai. RX Group brings deep ANZ retail commercial experience across strategy, store networks, category, and operating models. thecolab.ai brings world-class AI engineering capability and a track record of building platforms that go into production, not into presentations.

The proposition is deliberately simple: navigate, implement, stay accountable. Find the use cases that move margin. Build on foundations that are ready and strengthen the ones that are not. Fit AI into existing workflows without rip-and-replace. And stay in the room until it delivers - not until the engagement ends.

Where to start

For businesses that are sitting, watching, and waiting: that is a rational response to genuine financial pressure. But waiting is also a decision, and it has a cost that compounds quietly.

For businesses that are ready to move but do not know where to start that sticks: the right first step is not a six-month strategy engagement. It is a focused, facilitated session that cuts through the noise, pressure-tests your assumptions, identifies your highest-value AI opportunity, and gives you a clear prioritised path to production.

Working solutions, not slide decks. And if that session leads to a project with us, the fee is credited back. Your starting point becomes your foundation, not an additional cost.

87% of New Zealand businesses are using AI. 88% are stuck.

We built rxcolab.ai for the ones who are ready to move.

We don’t sell software. We make sure it actually delivers.

Find out more at rxcolab.ai