Market research has long been a privilege of the big end of town. Got $50,000 and six weeks to spare? Great, you can know what your customers think. Everyone else? Good luck.

That model is being dismantled, and a New Zealand startup is doing some of the dismantling. In the latest episode of The Business of Tech, I sat down with James Donald, CEO of Auckland-based Ideally, fresh from closing a $16 million Series A that values the company at $100 million. 

Ideally is one of three AI-centric New Zealand startups to hit that psychological valuation milestone in the past month – a sign that our fledgling AI start-up ecosystem is gaining momentum.

James is a former Shell engineer turned serial founder whose previous company, Yonder, was acquired by a US travel tech firm. Now he's turned his sights on a $40 billion slice of the global market research industry – one where 90% of spend still flows to people-heavy agencies like Kantar and Nielsen. 

His pitch: AI can do what took those agencies weeks to do, in hours, at a fraction of the cost, and with results in the hands of the people inside a company who actually know what questions to ask.

The pros and cons of synthetic data

In our chat, we get into the heart of what Ideally is doing differently. One of the most interesting debates in AI right now is the rise of synthetic data – building artificial personas to simulate how real people would respond. James makes a pointed argument: when the stakes are high, and you need genuine nuance, synthetic just isn't good enough. 

We also dig into what James calls "living data" – the idea that consumer insight shouldn't die in a PDF buried in SharePoint, but should be a continuously growing, queryable understanding of your customer base. 

And we talk about the SaaSpocalypse – that February moment when hundreds of billions were wiped off the value of software companies worldwide. Ideally sits squarely in that story: an AI-native challenger gunning for the market share of legacy research platforms and expensive agencies alike, with a usage-based pricing model designed to turn in-house marketers into researchers, rather than leave it to outside consultants.

This is a great example of how AI is being used to shake up long-established industries.

The Business of Tech is available on Apple Podcasts, Spotify and wherever you get your podcasts.

Market research has long been a privilege of the big end of town. Got $50,000 and six weeks to spare? Great, you can know what your customers think. Everyone else? Good luck.
That model is being dismantled, and a New Zealand startup is doing some of the dismantling. In the latest episode of The Business of Tech, I sat down with James Donald, CEO of Auckland-based Ideally, fresh from <a href="https://www.nzherald.co.nz/business/we-couldnt-have-done-it-without-ai-ideally-raises-6m/G3UGK3SVYRED3AQY2VHRBQZLJQ/">closing a $16 million Series A</a> that values the company at $100 million.&nbsp;
Ideally is one of three AI-centric New Zealand startups to hit that psychological valuation milestone in the past month &ndash; a sign that our fledgling AI start-up ecosystem is gaining momentum.
James is a former Shell engineer turned serial founder whose previous company, Yonder, was acquired by a US travel tech firm. Now he's turned his sights on a $40 billion slice of the global market research industry &ndash; one where 90% of spend still flows to people-heavy agencies like Kantar and Nielsen.&nbsp;
His pitch: AI can do what took those agencies weeks to do, in hours, at a fraction of the cost, and with results in the hands of the people inside a company who actually know what questions to ask.
The pros and cons of synthetic data
In our chat, we get into the heart of what Ideally is doing differently. One of the most interesting debates in AI right now is the rise of synthetic data &ndash; building artificial personas to simulate how real people would respond. James makes a pointed argument: when the stakes are high, and you need genuine nuance, synthetic just isn't good enough.&nbsp;
We also dig into what James calls "living data" &ndash; the idea that consumer insight shouldn't die in a PDF buried in SharePoint, but should be a continuously growing, queryable understanding of your customer base.&nbsp;
And we talk about the SaaSpocalypse &ndash; that February moment when hundreds of billions were wiped off the value of software companies worldwide. Ideally sits squarely in that story: an AI-native challenger gunning for the market share of legacy research platforms and expensive agencies alike, with a usage-based pricing model designed to turn in-house marketers into researchers, rather than leave it to outside consultants.
This is a great example of how AI is being used to shake up long-established industries.
The Business of Tech is available on Apple Podcasts, Spotify and wherever you get your podcasts.

The Business of Tech, hosted by leading tech journalist Peter Griffin. Every week they take a deep dive into emerging technology and news from the sector to help guide the important decisions all Business leaders make.

Issues such as cybersecurity, retaining trust after a cyberattack, business IT needs, purchasing SaaS tools and more.

New Episodes out every Thursday. Follow or subscribe to get it delivered straight to your favourite podcatcher.

@petergnz

@businessdesk_nz

Proudly sponsored by 2degrees Business!

 

The Business of Tech, hosted by leading tech journalist Peter Griffin. Every week they take a deep dive into emerging technology and news from the sector to help guide the important decisions all Business leaders make.
Issues such as cybersecurity, retaining trust after a cyberattack, business IT needs, purchasing SaaS tools and more.
New Episodes out every Thursday. Follow or subscribe to get it delivered straight to your favourite podcatcher.
@petergnz
@businessdesk_nz
Proudly sponsored by 2degrees Business!
&nbsp;

The Business of Tech

The Business of Tech: AI is eating market research

https://api.omny.fm/orgs/7784f840-c291-422a-924b-ad9000bbad71/clips/3ad4ad3a-c1fa-462b-9c4a-b441016d7926/transcript