Taking the temperature of the Australian market research industry, Q3, 2025

Like a dentist with horribly neglected teeth or an architect living in a characterless pre-fab home, is it possible the research industry, particularly in Australia, is woefully under-researched? 

As a community, do we actually have a good handle on the state of the industry overall? Particularly its strength and resilience during a time of rapid technological change?

This article is my maiden attempt to research the research industry — a now obvious solution to a growing list of questions I have compiled about the industry’s future and direction.

Beavering away in an independent agency, it’s easy to lose sight of the industry at large. And it’s hard to know if our workload and the typical briefs we’re answering are reflective of the broader market. But over the past 18 months or so I’ve noticed too many new developments. No longer able to ignore them, I needed to investigate for myself. 

Predictably, as a qualitative researcher, my approach was to conduct a small series — five in all — of one-hour depth interviews with industry participants.

For this first series, I focused on research sellers based in Australia (that means research agencies and their suppliers), although I did interview one client-side researcher (a ‘buyer’) based in London, an old colleague of mine, to get a broader view. 

For obvious reasons, I will not name them, any of their clients or commercially sensitive info mentioned during our discussions. 

These findings are very much preliminary and require tracking and synthesising over time. 

Thank you kindly to the five individuals who trusted me with their thoughts and perspectives. 

Strength of the market as of March, 2025: Mixed conditions and fortunes, but generally softer demand observed

All but one of the research professionals I spoke to reported weaker demand, and this held true even after the perennial January lull – when most Australians are enjoying summer holidays – should have ended.

Looking back over the last 12 months, two respondents — a boutique full-service agency and a panel provider — said sales were on par with the previous year but felt they were working harder for it. They explained that while they were delivering more projects, the average value of their projects had declined. Clients "sitting" on briefs, reducing brief scope, or "ghosting" was another common experience across the sample.

Tellingly, one respondent — a salesperson for a panel provider servicing agencies big and small across the country — had noticed an uptick in the number of agencies all vying for the same brief, pointing to greater competition.

Another respondent, citing a report that tracks the growth of the market research industry, said the industry is growing globally, Australia included, but also said, "How that spend is distributed is not very clear." Does that spend include subscriptions to CX packages like those offered by Qualtrics, for example? Or the burgeoning array of 'DIY' and AI-enabled research solutions?

This all said, respondents close to the ‘big agencies’ (you know who they are) said demand at the top end of town, if I can call it that, remained robust. A UK-based client-side research buyer, the only non-Australian-based person I interviewed, echoed this two-speed market dynamic on their side of the hemisphere: “The big agencies are doing alright. It’s the smaller ones that seem to be struggling a bit.”

So, what can I reliably deduce from speaking to such a small sample of industry participants about the strength of the market? What I didn’t hear is probably more illuminating. No one said words to the tune of ‘we’re manic’, ‘run off our feet’, or ‘we’re hiring’, which leads me to think that most agencies have seen busier times. Hardly surprising given the broader economic mood and the arrival of Trumponomics.

Market consolidation, the emergence of tech-centric ‘self-serve’ research providers, and "Uberisation"

“We seem to be going through a period of market consolidation,” one respondent told me as they related recent industry acquisitions such as Ipsos’s purchase of the Where to agency in January this year. As the owner of a market research venue, and therefore familiar with a great variety of research operators, I was curious to know how they pictured the market landscape in 2025: “There seems to be huge companies and then... freelancers. Not much in between. The middle tier seems to be getting squeezed.”

But a venue owner can’t tell us what’s happening in the online-only environment where entire projects can be launched and delivered within a matter of days, even hours. Enter the age of the tech-centric research supplier that is ‘disrupting’ and ‘democratizing’. A slew of new and reinvented research suppliers, particularly panel companies, are offering easy-to-use SaaS products where self-service is the name of the game.

Such solutions chime well with a trend for research buyers to reduce their dependence on external agencies and deliver more work internally using a small staff of researchers. Using Askable or Glow Feed is “frictionless,” said one respondent, once a company has an account; just top up your credit balance and your new project could be live within minutes.

But what about when a project is complex, messy, and of great strategic importance to the sponsoring company? Would a series of DIY surveys complete with automated charting or focus groups hastily recruited from an online service do the job? Arguably “yes,” said one respondent who briefly contracted to a tech-centric supplier to deliver a major stakeholder engagement project. As part of said company’s “managed” service, the company provided the technology (like access to their panel and survey creation software), while the contractor provided the end client with the stewardship and wisdom they needed to see the project through to completion.

Such a business model got us pontificating about the prospect of an ‘Uberized’ research industry where a handful of global tech players control the client relationships while a pool of contract researchers field messier, non-standard briefs when required.

The problem with this model, of course, is that the tech company sets the contractor’s hourly rates and sucks up all the value in a project.

Not exactly a rosy topic for those of us who love working closely with our clients, but something worth thinking about.

Use of synthetic data is still nascent, but projects delivered end-to-end by AI an opportunity

Respondents' use of synthetic data for quantitative research was limited. Despite some panel providers “pushing synthetic really hard,” no one had found an opportunity to use it in their project work. A major barrier is getting permission from cautious clients who remain unsure of its validity.

Most say its rightful place, for now, is as a ‘bolt-on’ to boost hard-to-find sample profiles (e.g., respondents living in the Northern Territory) or address non-primary research objectives when a survey has become so laden with questions it’s not practical to add any more. One respondent rejected the use of the word synthetic: “It’s just predictive modelling based on expansive data sets,” they countered. When you look at it that way, it doesn't feel so spurious.

As for the use of AI in the research industry, given the abundance of information already available, I will relay a simple but powerful anecdote from the UK-based client-side researcher I interviewed:

This person recently attended a qualitative research conference in mainland Europe where every presenter was keen to impress their AI wares upon the audience. While many of the presentations were “a bit blah,” one did manage to cut through: a paper co-delivered by a client and their research agency where AI tools did the heavy lifting at every step of the project journey. In other words, an end-to-end AI-driven qualitative research project. The AI produced the discussion guides, conducted autonomous moderation with hundreds of respondents via chat, and delivered the insights report. The humans involved simply prompted and redirected their AI helper when necessary. Inevitably, the audience, my contact included, bombarded the two presenters with questions about data quality and the possible impact of ‘talking’ to consumers without a human-to-human connection. Later that day at the post-conference networking drinks, the sponsoring client came up to my colleague and said something to the effect of, ‘I understand your concerns, but for the cost and time saving an AI-powered project offers, there are things I am willing to trade off. We’ve launched products off the back of this research, and they have been successful.’