New Zealand likes to tell itself an innovation story. Some of that story is true. The problem is that the strategy underneath it still looks too slow, too institutional and too disconnected from how modern innovation now happens.

Is New Zealand Falling Behind in Innovation? A 2025 Reality Check

New Zealand likes to tell itself an innovation story.

High-impact science, clever agri-tech, rockets from Mahia, a growing games sector and ambitious founders building from the edge of the world.

Some of that story is true.

The problem is the strategy sitting underneath it.

On paper, New Zealand is refocusing its science, innovation and technology system, building new funding pillars and publishing fresh strategies for digital technologies and artificial intelligence. The Government has announced changes to the science, innovation and technology system, MBIE has published material on New Zealand’s Research, Science and Innovation Strategy, and the country now has a 2025 AI Strategy: Investing with confidence.

In practice, the country is still working from a mental model that belongs in about 2010, not 2025.

The result is an innovation strategy that is roughly 15 years behind the reality of how modern innovation actually happens.

What Wellington currently calls an innovation strategy

Over the last few years, several big pieces have been put on the table.

• A new funding strategy for the science system, built around pillars such as economy, advanced technology, environment, health and society.
• A refocus of the wider science, innovation and technology system to give clearer direction and lift productivity and growth.
• A Research, Science and Innovation Strategy that positions New Zealand as a global innovation hub in a productive, sustainable and inclusive future.
• Industry Transformation Plans, including a Digital Technologies Industry Transformation Plan intended to grow the digital sector and lift economy-wide productivity.
• A 2025 Artificial Intelligence Strategy focused on accelerating private-sector AI adoption and creating an environment where businesses can invest with confidence.

The language in these documents is familiar: pillars, hubs, long-term direction, strategic investment, productivity, inclusive growth.

None of that is wrong.

The issue is what is missing.

Almost all of it still treats innovation as something that starts in formal science, flows through funding programmes and institutions, and eventually emerges as commercialised products at the other end.

That linear “science to market” pipeline was already under pressure 15 years ago. In a world of cloud platforms, open-source software, AI models, product-led growth and global digital markets, it is nowhere near enough.

The reality check: indicators that do not match the rhetoric

If New Zealand’s innovation strategy was working, you would expect to see it in a few obvious places: research intensity, productivity, digital scale and talent dynamics.

R&D investment is still well below leading economies

Stats NZ reported that total R&D spending reached $3.7 billion in 2023, up 17 percent from 2022.

That sounds positive until you look at R&D intensity.

• New Zealand’s R&D spend is around 1.5 percent of GDP.
• The OECD average is materially higher.
• Leading small advanced economies invest significantly more again.

RNZ has previously summarised the challenge bluntly: New Zealand’s research and development investment remains well below other leading economies.

Despite tax incentives and new funding structures, the gap is still there.

Productivity remains a structural weakness

The OECD and the New Zealand Treasury have been saying the same thing for years.

New Zealand ranks highly on many wellbeing measures, but incomes sit below the OECD average because of persistently weak labour productivity. The New Zealand Treasury’s productivity commentary has repeatedly highlighted the country’s productivity challenge.

Innovation strategy that does not move productivity is performance, not progress.

This is why practical innovation has to connect to actual business improvement, not just national-level strategy language. For organisations, that means linking innovation to process improvement, workflow automation, better data and measurable operating outcomes.

The digital and AI sectors are still small relative to the opportunity

The Digital Technologies Industry Transformation Plan notes the importance of the digital sector and its potential contribution to future productivity.

That is significant, but for a country that wants to be an innovation hub, it remains modest.

The 2025 AI Strategy is mainly framed around private-sector adoption and confidence, not building frontier capabilities in areas such as foundation models, compute infrastructure, AI platforms or globally scaled deep-tech companies.

In other words, policy is still talking mainly about using AI tools rather than building the platforms, infrastructure and companies that others depend on.

This matters for Changeable’s work with New Zealand organisations. Many businesses do not need frontier AI research, but they do need a practical AI strategy that connects AI adoption to productivity, service quality, cost, risk and decision-making.

Talent and capital keep leaking

New Zealand continues to battle familiar problems.

• Brain drain to larger markets, particularly Australia.
• Domestic capital markets that are small and risk averse.
• A business sector dominated by SMEs with limited internal innovation capacity.
• Too few pathways for deep technical talent to scale globally from New Zealand.

An innovation strategy that does not address where talent chooses to live and work is playing a different game than the one the economy is actually in.

Why this looks like a 2010 strategy in a 2025 world

The 15-year lag is not mainly in the language. It is in the assumptions underneath the policies.

Innovation is still treated as science plus commercialisation

Most official documents still frame innovation as publicly funded research, handed off to commercialisation offices, supported by grants and tax incentives, and eventually turned into products and services.

That model made sense when physical science and lab-based R&D dominated.

In software and AI, the loop often runs differently.

• Product teams or founders ship something useful on top of global platforms.
• They iterate based on usage data.
• Technical learning happens inside the product flow, not in a separate pipeline.
• Distribution, customer feedback and platform access matter as much as initial invention.

New Zealand’s strategies mention digital and advanced technologies, but the machinery is still tuned heavily for multi-year research programmes and funded projects rather than continuous, product-led innovation.

Policy still optimises for inputs, not outcomes

The system puts significant energy into the number and type of funding pillars, governance structures, advisory groups, R&D tax settings and contracting mechanisms.

These are inputs.

Outcomes would look more like:

• Higher productivity in specific sectors.
• Export revenue from IP-rich products and services.
• A visible pipeline of mid-stage and later-stage technology companies.
• Adoption of AI and automation in everyday business processes.
• Better procurement pathways for local firms with validated technology.
• More companies scaling from New Zealand rather than leaving to scale elsewhere.

New Zealand publishes strong narrative reports about research and innovation investment, but there is much less transparent strategy-level focus on whether that spend is shifting the real economy.

That same pattern appears inside individual organisations. Leaders often measure AI activity, workshops, pilots and tool licences before they measure actual workflow improvement, decision quality or productivity gain.

This is why Changeable starts with AI use case discovery and practical outcome definition before recommending tools.

Risk is framed in yesterday’s terms

Debates around gene technology illustrate the pattern.

For years, policy was dominated by precaution, delay and binary framing. In 2025, the Government moved toward modernising gene technology regulation and enabling more current use of these tools.

The pattern is familiar:

• A long period of precaution and delay.
• Global science and industry move on.
• New Zealand then rushes to catch up.

AI is at risk of following the same path if regulators and institutions focus only on controls after the fact, rather than experimentation frameworks, sandboxes and proactive capability building.

Good AI governance does not mean doing nothing until all uncertainty is gone. It means creating safe ways to learn, test and scale.

The strategy is institution-centric, not platform-centric

The current reforms create new funding entities, investment plans, research organisations, strategies and coordination structures.

Those matter, but they are still institution-centric.

A 2025-ready innovation strategy would think more like a platform.

• Shared national data assets and synthetic data environments.
• Shared AI and compute infrastructure accessible to startups, researchers and SMEs.
• Procurement pathways that let government become an early customer for local innovation.
• Open standards and APIs that make it easier for others to build on top.
• Industry-specific AI adoption pathways that connect technology to productivity.

Without that, each initiative becomes another silo that has to integrate back into a fragmented system.

What a 2025-ready innovation strategy would do differently

If you were designing from scratch today, trying to serve a small, open economy at the edge of the world, the priorities would look different.

Start from productivity and export outcomes

Instead of starting from funding instruments, start from a small number of measurable outcomes.

• Lift labour productivity closer to the OECD median within a clear timeframe.
• Increase the number of firms earning significant export revenue from IP-rich products and services.
• Increase business R&D intensity toward leading small advanced economy benchmarks.
• Grow the number of companies scaling globally while remaining meaningfully based in New Zealand.

Then build innovation policy around those targets, not the other way around.

This is the same logic organisations should use when building an AI strategy. Start with the measurable business outcome, not the tool, funding category or technology trend.

Treat software and AI as horizontal infrastructure, not niche sectors

Modern innovation, in almost any industry, rides on cloud infrastructure, APIs, AI models, data pipelines, analytics and workflow automation.

New Zealand’s AI Strategy starts to address adoption, but a frontier-aligned approach would go further.

• Co-invest in shared compute and model access for researchers, startups and SMEs.
• Support open-source and local model development where it makes strategic sense.
• Embed AI capability into every industry productivity plan, not just digital technologies.
• Help SMEs use AI and automation to lift productivity in real workflows.

The question should not be “do we have an AI sector?”

The better question is: “Is every productive sector using AI, software and data to close the productivity gap?”

Move from project funding to platform building

Rather than distributing money across many projects, a 2025 strategy would concentrate effort on a smaller number of shared platforms.

Examples could include:

• National data spaces for agriculture, health, climate, logistics and infrastructure.
• Testbeds for autonomous systems, climate tech, biotech and advanced manufacturing.
• Interoperable digital identity and consent frameworks that reduce friction for new services.
• AI governance and assurance frameworks that help organisations adopt safely without reinventing the wheel.

These platforms de-risk innovation for many firms at once and make it easier for global partners to plug in.

Use government as a deliberate early customer

One of the strongest levers a small state has is its own procurement.

A modern innovation strategy would:

• Set explicit pathways for procuring from New Zealand-based innovative firms in key areas.
• Use outcome-based tenders that let companies propose novel solutions.
• Build multi-year adoption programmes around validated tools, not one-off pilots.
• Create clearer routes from trial to operational deployment.

This matters because many New Zealand innovators do not just need grants. They need credible early customers, reference sites and repeatable adoption pathways.

The same problem appears inside AI adoption. Many organisations run pilots, but few create a pathway from pilot to operational use. That is why implementation planning, governance and change design need to be built in from the beginning.

Build a talent and capital pipeline that matches the ambition

A serious innovation strategy needs a serious talent and capital strategy.

That means:

• Stronger pathways for deep-tech founders and technical talent to spin out of universities and Crown Research Institutes into startups.
• Co-investment funds that specialise in capital-intensive, long-horizon innovation.
• Active attraction and diaspora strategies for New Zealanders in global technology hubs.
• Better links between universities, technical education, business capability and export-oriented firms.
• Practical AI capability building for SMEs, not just large enterprises.

The UK Government’s trade and investment material describes New Zealand as having a high-impact science ecosystem and strong innovation culture. That reputation matters.

The missing piece is converting that reputation into a repeatable pipeline of globally competitive firms that choose to base themselves here.

What this means for New Zealand businesses

The national innovation problem is not only a government problem.

Businesses also need to stop waiting for a perfect national innovation system before acting.

For SMEs, councils, professional services firms and mid-sized organisations, the practical question is:

What can we do now to improve productivity, decision-making, customer experience and operational capability with the tools already available?

That is where AI, automation and better process design become very practical.

Most businesses do not need a moonshot to improve. They need to:

• Find the work that is repeated, hidden or manual.
• Improve processes before automating them.
• Use AI where it improves speed, quality or decision support.
• Build practical governance around data, privacy and human review.
• Develop internal capability so people can actually use the tools well.
• Measure outcomes, not activity.

This connects directly to Changeable’s work in process improvement, workflow automation, AI agents, data models and AI governance.

So is New Zealand doomed to fall behind?

No.

The underlying strengths are real.

• High-quality science.
• Trusted institutions and national brand.
• A digital technology sector with global niches.
• Strong agri-tech, climate-tech, aerospace and creative technology stories.
• A growing recognition that the current system needs refocusing.

The problem is not intent.

It is tempo and framing.

If New Zealand keeps treating innovation as science funding plus tax credits, it will keep getting what it has now: strong stories, good pilots and chronic productivity underperformance.

If it is willing to adopt a 2025 frame that is honest about AI, software, data, talent, procurement and platform building, then the country can move from being 15 years behind reality to quietly building the next 15 years on purpose.

That requires an innovation strategy that is less about pillars on a diagram and more about outcomes in the real economy: working smarter, exporting more valuable things and giving the next generation a reason to build here rather than somewhere else.

What Changeable helps with

Changeable helps New Zealand organisations turn innovation ambition into practical operating improvement.

• AI strategy that connects technology to business value, productivity and implementation.
• AI use case discovery to test whether an idea is viable before investing.
• Process improvement to remove bottlenecks before automation.
• Workflow automation that targets real operational friction.
• AI agents for knowledge retrieval, triage, analysis and task support.
• Data models that make reporting, insight and decision support more reliable.
• AI governance that supports safe experimentation and responsible adoption.
• AI maturity and readiness assessment to identify capability gaps before scaling.
• Fractional AI leadership for organisations that need senior guidance without a full-time AI lead.

Frequently asked questions

Is New Zealand really falling behind in innovation?

New Zealand has real innovation strengths, but the country continues to underperform on productivity, R&D intensity, scale-up pathways and the conversion of ideas into globally competitive firms. The issue is not lack of talent. It is the operating model around innovation.

What is wrong with New Zealand’s innovation strategy?

The problem is that much of the strategy still looks institution-centric and funding-led. A 2025-ready strategy needs to be more platform-led, outcome-focused and honest about AI, software, data, talent and procurement as core innovation infrastructure.

Why does productivity matter in an innovation discussion?

Innovation should eventually show up in productivity, export value, better services, stronger firms and higher-value work. If innovation policy creates stories and pilots but does not shift productivity, then the system is not converting ideas into economic impact.

How does AI change the innovation equation?

AI makes innovation faster, more software-driven and more dependent on data, compute, platforms and product iteration. Countries and organisations that treat AI as a niche sector risk missing its role as horizontal infrastructure across the whole economy.

What can individual businesses do?

Businesses can start by identifying practical AI and automation opportunities, improving processes, building internal capability, putting governance in place and measuring real productivity outcomes. They do not need to wait for perfect national policy.

How can Changeable help?

Changeable helps organisations build practical AI strategy, clarify use cases, improve workflows, design AI governance and implement automation in ways that create measurable value.

Steve Wilson is the founder of Changeable and Ministry of Insights, providing AI strategy, governance and automation consulting for organisations navigating the gap between AI ambition and operational reality.

For people and teams still building confidence with AI before implementation, visit Zero to AI.