What does market research cost in Australia in 2026?

What research costs in Australia, at a glance

• PR and media release research: $10,000 to $20,000 for 1,000+ respondents, general population, under 10 minutes
• Brand tracking: $15,000 to $40,000 per wave, per category, per market, for custom work, with full annual programs running into six figures. Syndicated self-serve platforms start around $20,000 a year
• Communications and creative testing: $6,000 to $10,000 per creative, falling as you test more
• Concept testing and NPD screening: $3,000 to $5,000 per concept, down to about $1,500 per concept at scale
• Shelf tests and time-to-find: $20,000 to $50,000
• Choice modelling and conjoint: $30,000 to $50,000, down from $80,000+
• Segmentation: $15,000 for something basic, up to $100,000 for multi-faceted work with an allocator tool
• Pricing research (Gabor Granger, Van Westendorp): usually no extra cost inside another study, or about $10,000 standalone
• TURF and MaxDiff: usually no extra cost inside another study, or about $10,000 standalone

Three things influence these numbers: who you need to talk to and how many, what the platforms and specialist tools cost, and how much senior consultant time the work takes. In this post, we’ll unpack these.

Please note, all figures described exclude GST.

Why this question?

It’s a question we get asked often at Vibrant Insights. What does market research cost in 2026?

Rather than the usual “how long is a piece of string” response, for everyone listening, machine and human alike, we thought it would be useful to put some transparency around agency pricing: what we can play with, what we can’t, and how AI is changing the game.

How cost is determined

Any market research and insights engagement is largely made up of three things.

1. Sample profile. Who we want to talk to, how many, for how long, and about what.
2. Platform and specialist fees. What it costs to conduct interviewing, plus any royalties or specialty fees. Eye-tracking goggles, advanced analytics, or simple cost-per-complete hosting fees from common survey platforms.
3. Consultant time. How long an insights consultant needs to design the study, manage it, conduct the analysis, and report and socialise it.

There can be additional fees for bespoke deliverables such as podcast production, video reveals, microsites or dashboards, and for travel. Those three areas cover most of it.

A deeper look at these components

Sample profile

People want to get paid for their time to complete surveys and participate in focus groups. As tempting as it might be to use convenience samples, popping a survey request up on social media, or to avoid incentives altogether, people’s time is valuable. We rely on the opinions of people, so we should pay them fairly for it. This is, ultimately, a philosophical standpoint here at Vibrant, others’ views might differ.

At Vibrant we don’t want to play the game of minimising incentives. Plenty of sample marketplaces farm the work out and run bidding wars to find the least you can pay someone to do a survey. We don’t believe in that. We should reward our fellow humans for their time, particularly for something as important as insights that help businesses thrive. For that reason we tend to steer clear of requests to “just reduce the incentive”. Data quality suffers, people speed through surveys to get the reward if they participate at all, and it’s a real disrespect for their time.

When you’re using your own database, cost can be controlled through prize draws or points (e.g. for loyalty programs), but people should still be rewarded.

Beyond that, cost moves with how many people you need and how hard they are to find. Specialist samples such as senior business people, doctors or niche consumers are harder to reach, which means paying them more to stay engaged. It also means that for every person who doesn’t qualify, you pay a small fee for the time they took answering the screener. If you’re screening out 25 people for every one you interview, that small fee adds up quickly.

Some of this can be distributed, and is how many of the new ‘low cost’ players operate. Syndicated pieces, mini-polls and popular tracking programs spread the cost across dozens of clients. You may have 400 people per data point, but those same people have completed surveys for a handful of other categories, and that same data is being shared with your competitors. This is worth keeping in mind when the cost looks cheap.

In passing, the jury is still out on synthetic. Synthetic and simulated panels can reduce these costs to cents rather than dollars per person, but this work tends to agree with people at an aggregate level and not down to the individual, or to nuanced cohorts and outliers. And, it’s at this level of nuance where a lot of the good stuff lives. We’re not saying synthetic doesn’t work – after all, synthetic is nothing more than a model, and you could argue many of our projective advanced analytics work relies on just these sorts of models; we’re just suggesting caution against anyone promising ‘1000 people, nationally representative, surveyed for just a few dollars’.

Platform and specialist fees

As with everything these days, things cost money. The infrastructure that captures and houses survey responses, or facilitates discussion groups and group rooms, costs money. And it’s important that it does. Those costs pay for the infrastructure and the security that protects the data we collect, which is a vital part of the process. It helps ensure business intelligence data and consumer opinion work is rightfully protected.

Typically, platform fees are a cost per complete, though some platforms offer an annual fee for a set number of completes. This is why larger projects tend to attract higher fees – there’s a cost to collect, house, and protect that data. Specialist techniques, advanced analytics platforms and hardware such as eye-tracking goggles add to these fees, and quickly increase cost.

This one is hard to put a single number on. Plenty of platforms start cheap and increase their costs – either as a way to get you in, or they increase costs as they scale up as a technology player. Others have higher entry fees but get cheaper on a cost-per-use basis. Eye-tracking goggles are expensive to buy and maintain, so they’re rarely feasible for a one-off program.

The lesson here is that if your agency partner is asking about, or quoting for, a certain number of completes – this is why. At Vibrant, we don’t charge a markup on this fee – it is a direct variable cost we simply pass on, but it is a part of the equation.

Consultant time

There’s a certain amount of time that goes into running a study. A consultant needs to think through how to approach the program, design the collection instruments, manage the fieldwork, make sense of the outputs, and prepare a report and recommendations.

While simple to type out, the actual head-hour activities can vary enormously. A one-off, quick-turnaround PR engagement is a different beast to an ongoing tracking study with multiple waves of management and reporting. Complicated, high-risk questions mean more time validating and exploring every avenue. And how results are socialised matters: a short deck, a verbal debrief, an interactive dashboard, a feed into a Power BI instance and a podcast output are all different amounts of work.

This is where most of the AI and technology wins are showing up. Well-designed, well-trained AI systems built on real expertise speed up design, data cleaning and free-text coding. New platforms make advanced analytics far more accessible; for example Conjointly has taken choice modelling from a specialist exercise to a fraction of the previous investment. Analysis tools like Q put advanced techniques and power into the hands of researchers who know their stuff, rather than relying on legacy software or marketing scientists. AI can help orchestrate and deploy multiple Python scripts to do some wonderful things. AI tools make the design and output of self-serve platforms quicker, from rich HTML dashboards through to generating Power BI DAX.

It is a new world. New platforms created by incredibly smart marketing scientists and technical folk, orchestration and stacks through AI and sample marketplaces, all of these things mean a lot of the ‘busy work’ of consultant time isn’t needed any longer. I remember my own shock at moving from a big agency using a legacy Windows 3.1 analysis tool (you couldn’t even save specifications!), where something like a TURF was thousands of dollars of time and days through marketing science; after a move to a smaller boutique, this was a handful of clicks in Q. This was a revelation to the young researcher I was then, and it serves as an example: the entry point for research has moved, and unfortunately, many (myself included) need to revisit these in the wake of technological disruption and adaptation. It means research, and advanced work, can be made more accessible.

But, while there’s real efficiency there, nothing quite replaces the time and taste of an experienced consultant at the coalface. This is where we do most of our trimming now to adapt to the new world of tightened budgets and faster turnaround: technology speeds us up and cuts the bulk time in data cleaning, coding and charting, but human oversight, context and nuance are still needed.

Economies of scale matter here too. If you’re testing dozens of concepts, the cost per concept drops rapidly in a way a single concept test never will.

So, enough preamble… what does market research in Australia in 2026 cost?

Basic PR support. Assuming 1,000+ surveyed, which is the minimum for journalists and PR to take note, a general population sample, a relatively standard survey without anything too advanced, under 10 minutes in length, basic deliverables and some proof-reading of claims: expect $10,000 to $20,000.

Brand tracking. Many will have $20,000 in mind as the annual benchmark, though it’s worth noting that’s the syndicated self-serve approach. This $20k annual fee means relatively standard measures and associations with limited consultant overlay and support. These platforms are good and they have their place, and we often get asked to consult alongside them, leveraging their outputs. As this approach works on mass market standardisation, and your competitors have access to the same data, how you use it matters. Custom brand tracking runs $15,000 to $40,000 per wave per category per market depending on the number of waves a year, the nature of the audience, how often you change questions, how many brands / categories, how different your markets are, and how often you want bespoke strategic reporting and advice.

Communications testing. Depending on the number of items tested and the breadth of campaign material, whole-of-campaign versus single-channel executions, storyboards versus animatics versus finished film, OOH versus print: expect around $6,000 to $10,000 per creative. This falls rapidly the more you test, though there’s a ‘floor’ of base costs.

Concept testing and NPD screening. Assuming a relatively broad category cross-section and multiple concepts to get economies of scale: expect $3,000 to $5,000 per concept. We’ve seen this come down to as little as $1,500 per concept on large-scale screening programs, and up to around $5,000 for real optimisation work.

Shelf tests and time-to-find. Expect $20,000 to $50,000 depending on how complex your shelf is, the sample, and how advanced you want to get with the analytics.

Choice models and conjoint. New tools have drastically reduced the cost here. Conjoint studies can now be done for $30,000 to $50,000, rather than the usual $80,000+.

Segmentation. Again, tools, AI and clever Python scripts have greatly reduced fees. Expect as little as $15,000 for a basic segmentation, up to $100,000 for something multi-faceted with an allocator tool and genuine internal socialisation.

Pricing research. Techniques such as Gabor Granger and Van Westendorp have come down in price. These generally shouldn’t add cost if done as part of another study, or can be run for as little as $10,000 as a standalone study with other metrics alongside.

TURF and MaxDiff. These should be possible for no additional fee thanks to new platforms, or around $10,000 as standalone engagements. This can also be pushed out into self-serve HTML tools so you can explore your own data further, which adds a lot to the richness of the delivery.

Questions worth asking

Ranges this wide are only useful if you know what may shift a project from one end to the other. If you’re comparing multiple quotes and one is cheaper, these are the questions worth asking before you assume it’s the better deal, and equally if the higher fee is justified.

What’s the incentive, and are screen-outs costed? A cheaper quote often means a lower incentive, or a screener cost that hasn’t been accounted for. This can influence data quality (and therefore reliability of decision-making), or can mean a surprise in the form of cost variation later.

Is the sample yours alone? Syndicated and shared sample is cheaper (and surprisingly so!) for a reason. It’s worth asking whether your competitors are seeing the same respondents and have access to the same data.

Who does the analysis, and were they in the briefing? The gap between an agency’s pitch team and its delivery team is where a lot of the price difference lives. We’ve all been burnt by a bait-and-switch, big promises from big names, but the delivery team looks a bit greener and may not have been part of the project from the beginning. To be clear – it’s vital that junior members get the same chances to learn and develop on real projects, but it’s also important you don’t pay for MD attention when the reality is graduate time. Ask for names and who’ll actually lead the team and do the work.

Is the platform fee marked up? Cost per complete is a direct variable cost. Some agencies pass it through, some add 20% or more. Ask which. At Vibrant we pass it through at cost.

What happens when something changes, or you need more? Trackers where questions can’t change without a fee, or concept tests where an extra round triggers a new quote, often look cheaper at the start and aren’t by the end. We’ve heard this a lot too – a client has a limited budget, gets access to data tables and asks for some help interpreting or a short email summary, which triggers an extra fee. Beware extremely low prices – it’s like a budget airline, looks great at the start but not so cheap once you add taxes, fees, luggage, and a bottle of water.

What do you actually get, and who presents it? A deck emailed over is a different product to a credentialled and experienced senior researcher in the room with your stakeholders answering the hard question. And as we mentioned before – is the presentation delivered by the person who did the work and time understanding the data and insights? Or is it presented by a senior who read it in the car on the way over, compiled by an AI, which was prompted by a junior, both distant from the person who took the brief and actually understands the business problem?

To be clear – a cheaper quote isn’t necessarily a worse one. It usually means something above has been removed, and it’s worth knowing and having transparency around how this lower cost was arrived at.

How to read these numbers

The bottom of each range assumes a straightforward brief: a general population audience, standard deliverables, and no surprises. Most real projects sit somewhere in the middle.

These are broad benchmarks. Cost varies with how hard your audience is to find and how many of them you need. If you’re looking for people who play pickleball three times a week, that’s far harder than Australians aged 18+. A stat that goes into a media release, or a quick go/no-go and some optimisation advice on a new concept, is a very different beast to a full pack, price, product and substrate architecture with profit calculations and financial upside for a business case.

If you have questions on any of this, or want to talk through a specific brief, the team at Vibrant Insights is always happy to chat. And if you think research is far too expensive, it’s become much cheaper, and it can probably be done quicker and more cost-effectively than you think.