New Zealand is a small country that sells food to the whole world. That one fact tells you most of what you need to know about where the money sits.
I looked at six ideas we first researched for India and asked a simple question - which of them still make sense in a market of about five million people? Here's where I landed.
The market in brief
New Zealand had about 5.32 million people in 2025 (source: World Bank, 2025) and GDP per capita of about US$49,590 (source: World Bank, 2025). Around 94% of people were online in 2024 (source: World Bank, 2024), so you can sell software without a field team.
The bigger number is this one. Food made up about 69% of merchandise exports in 2025 (source: World Bank, 2025). So the buyers with budgets are dairy co-ops, orchard groups, wineries and the exporters who ship for them. Overseas buyers keep asking for proof of where food came from, and if you can make that proof cheap then you have a reason to exist.
Setting up is simple. Companies register online with the Companies Office and tax runs through Inland Revenue (source: New Zealand Companies Office, 2026).
One caution before the list. Every idea below was built for India, so the rupee costs and Indian licences don't carry over. Read each one as a starting brief, not a plan.
1. Spray records for food exporters
Peppertrail ties each export lot to the farms, traders and processing steps behind it, so an exporter knows where residue risk sits before the lab test. In India it's built for chilli, cumin and turmeric, and the exporter pays a yearly fee per supply cluster. The Indian estimate is ₹45–80 L to build, with 6–9 months to launch.
What to adapt: swap spices for whatever your local exporters ship, such as honey, kiwifruit or dairy ingredients. The logic is the same - one rejected container can cost more than a year of the service. The hard part is also the same. Middlemen may not want to say where things came from.
2. Delivery robots for hotels
Trayline rents robots to business hotels with 80–250 rooms. At night, when one or two staff cover the building, the robot takes towels and room-service trays up in the lift. Hotels pay a monthly fee. The Indian build cost is ₹60 L–1.1 Cr.
What to adapt: in India the risk is that a night runner costs less than a robot. In New Zealand labour is dearer, so the sums may work better here [Requires validation]. Lift integration is still the first thing to check with any hotel.
3. Better crop yield sampling
Cutplot uses satellite images to plan where crop-cutting experiments should happen under India's crop insurance scheme. State agriculture departments pay. It needs ₹70 L–1.4 Cr and 9–15 months.
What to adapt: this one travels the least. It's tied to an Indian government scheme and its rules. If you want the idea in New Zealand then look at insurers and farm lenders who need yield data, and expect to rebuild the product around them.
4. Falls prevention classes for older adults
Surestride runs small-group strength and balance classes led live by a physiotherapist over video. The phone camera counts reps and times balance holds. Senior-living communities pay per member. It costs ₹50–95 L and 4–6 months to launch in India.
What to adapt: the Indian version leans on physio registration under the NCAHP Act. Here you'd check local physio registration instead [Requires validation]. Older Kiwis are spread across small towns, so a video-led class fits the geography well. Attendance is the risk everywhere. Below twice a week, the benefit fades.
5. Proof of origin for leather goods
Hidetrail maps the chain behind a leather exporter's products, back to the hide traders and slaughterhouses. EU buyers want this ahead of the deforestation rule (EU Regulation 2023/1115). It needs ₹55–95 L in India.
What to adapt: the same EU rule touches any exporter selling covered goods into Europe, so check which New Zealand exports fall under it [Requires validation]. If your farmers already keep good records then the service is cheaper to run here than in India.
6. Telepresence robots for parents living alone
Wickerly rents a simple telepresence robot to families whose parents live alone. A son or daughter abroad can drive it around the house during agreed hours. The adult child pays monthly. In India it costs ₹40–75 L and 5–8 months.
What to adapt: if the adult child lives in Australia or further away then the same "child far away" buyer exists here too. The risk doesn't change. Many parents will see a camera on wheels as surveillance, whatever the privacy switch does.
Which one I'd test first
If I were starting in New Zealand tomorrow, I'd pick the export proof ideas, Peppertrail and Hidetrail. They sell to the people with money, they work remotely, and Australia is a short flight away when you want to grow. The robot ideas are fun, but hardware in a small market is a hard way to learn.
For more, see the full list of startup ideas in New Zealand, the India idea hub and the quick-launch ideas.
No. They were researched for India. The costs are Indian estimates and the licences are Indian rules, so treat each one as a brief you need to check against local rules.
What's the biggest buyer group in New Zealand?
Food exporters and the co-ops behind them, because food is about 69% of merchandise exports (source: World Bank, 2025).
Should I start in New Zealand or Australia?
Many founders test at home and treat Australia as the first real step up. I think that's the right order for most of these ideas.
Talk to Yash
If you want to work out which of these fits New Zealand buyers before you spend on a build, I'm happy to think it through with you. Talk to Yash →
For illustration only. Ideas are hypothetical and names fictional; sourced facts may be out of date. Nothing here is legal, financial or professional advice, and Yash Khatri is not personally liable for any use of it. Read the full disclaimer.
For illustration only. Names are fictional and ideas hypothetical; sourced facts are real but may be out of date. Nothing here is legal, financial or professional advice, and Yash Khatri is not personally liable for any use of it. Read the full disclaimer.