The Future of Market Research Is Judgement, Not Data.

I have spent enough time in market research to know that having more data does not necessarily leave you any closer to a good decision.

If anything, the opposite can be true. Most organisations are now surrounded by information: survey data, behavioural data, customer databases, social listening, reviews, sales data, dashboards, and a steady stream of secondary research. The challenge is rarely finding something that answers the question. It is deciding which evidence deserves weight, what has been misunderstood, what is missing, and ultimately, what the business should do with it.

As an industry, we are very good at talking about methodology, sample quality, analytical techniques, and the growing range of tools available to us. All of those things are important, but they are not the same as judgement. Research can be methodologically sound and still send a business in the wrong direction because the wrong question was asked, the context was missed, or an interesting finding was given more importance than it deserved.

I think this is where the future of market research gets much more interesting. The most valuable researchers will not necessarily be the people who produce the most analysis. They will be the ones who know when to challenge the brief, when not to trust the obvious answer, how to connect evidence that does not fit neatly together, and when there is enough evidence to recommend a course of action.

Research has never suffered from a shortage of data. What businesses increasingly need is someone capable of making sense of it.

AI Is Making Execution Cheaper. Judgement Is Becoming More Valuable.

AI has accelerated a shift that was already underway in research. Tasks that once took hours or required specialist support can now be done in minutes. Transcripts can be summarised almost instantly. Questionnaires can be drafted, desk research synthesised, and first-pass analysis produced with very little friction.

That is useful, but it also forces an uncomfortable question for the industry: if more of the execution can be automated, where does the researcher create value?

I do not think the answer is better prompting. The advantage will come from knowing when the output is wrong, when it is technically correct but contextually absurd, and when the question itself needs to be challenged before anyone starts analysing the answer. AI can produce something polished and plausible very quickly. But plausibility is not the same as judgement.

Research rarely deals in clean, objective truths. People contradict themselves, markets move, culture shifts, and what consumers say often diverges from what they actually do. A model can help us process those contradictions, but it cannot remove the need to interpret them. If anything, the easier the analysis becomes, the more important that interpretation gets.

The Researcher of Today and the Researcher of Tomorrow

If AI is taking friction out of execution, the obvious question is what researchers should be getting better at instead.

I don’t think the answer is to turn every researcher into a technologist. We need to understand the tools, certainly, but the bigger shift is in how we think about the job itself.

The traditional research role has often been built around delivering a piece of work well: designing the study, protecting the methodology, analysing the results, and presenting the findings clearly. Those skills still matter, of course, but in the age of AI, they are simply becoming less distinctive on their own.

The researcher of tomorrow will need to operate much closer to the decision.

If you run an insights function, commission research, or rely on it to make decisions, I think the way you judge its value has to change too.

The number of studies completed, dashboards delivered, or methodologies deployed tells you very little about whether research is improving the business. A better question is whether your researchers are changing decisions. Are they spotting risks before they become obvious? Are they identifying opportunities the business would otherwise have missed? Are they prepared to tell senior leaders that the question being asked is the wrong one?

That is a much higher bar than producing good research. It also explains why the strongest insight teams increasingly look less like internal service functions and more like strategic partners to the business.

The first shift is from answering the brief to interrogating it.

By the time research is commissioned, a business problem has usually already been translated into a research question. Sometimes that translation is exactly right, but sometimes it sends everyone off to answer a question that will never solve the real problem.

A good researcher can answer the question they were given … a great one can recognise when it is the wrong question.

That takes as much confidence as it does curiosity. It means understanding the decision behind the brief and being prepared to push back when the requested research will not get the business where it needs to go.

The second shift is from analysis to interpretation.

Analysis can tell us what happened in the data. Interpretation is about deciding what deserves weight.

A statistically significant result may have very little commercial importance. A small behavioural change may look insignificant in isolation but be the early signal of something much bigger. Two datasets can point in different directions without either being wrong, and consumers routinely say one thing while doing another.

This is where judgement becomes visible. The researcher has to decide which evidence is most credible, what still needs to be tested, and how much confidence the business should place in the conclusion.

AI can surface patterns faster, but it cannot make ambiguity disappear.

The third shift is from findings to recommendations.

Researchers are very comfortable saying, “This is what we found.” We are usually more cautious about saying, “This is what I think you should do.”

Some of that caution is healthy. Evidence has limits, and good researchers should be clear about them. But neutrality can become a hiding place when every finding is given equal weight, and the client is left to work out the implications alone.

Senior decision-makers do not need more information for its own sake. They need a point of view:  what matters, what does not, what is uncertain, and what the evidence suggests they should do next.

That does not mean pretending certainty exists where it does not.  It means being explicit about the limits of the evidence, then still being prepared to make a call.

The fourth shift is from methodological expertise to methodological judgement.

Knowing how to run a method is not the same as knowing when to trust the result.

That distinction becomes more important as researchers are asked to evaluate work they did not personally produce: automated coding, synthetic respondents, behavioural datasets, AI-generated summaries, or analysis created elsewhere in the business.

The real skill is recognising when something does not add up.

Experienced researchers develop a feel for this. A number looks wrong, or a segment is suspiciously clean. What respondents say does not fit what they actually do. A conclusion has travelled further than the evidence allows.

That instinct comes from seeing enough real people, real markets, and real projects to know how research can fail.

Finally, the researcher of tomorrow will need much stronger commercial literacy.

A recommendation is only useful if it makes sense in the real world of the business. A finding that looks important in a research report may become far less important once pricing, operational constraints, competition, regulation, or internal politics enter the picture.

This is where research starts to prove its commercial value. The return is rarely the report itself. It is the bad decision avoided, the opportunity spotted earlier, the product changed before launch, or the investment redirected before too much time and money have been committed.

The strongest researchers I know are interested in much more than research. They understand how the category works, how the business makes money, where it is vulnerable, what competitors are doing, and which shifts in culture, technology, or consumer behaviour could change the picture.

That is why I think the future of research has to be judged less by how efficiently we produce evidence and more by whether we help businesses make better decisions.

We Need to Train for Judgement

If we want more strategic researchers, we have to stop assuming they will simply become strategic with seniority.

Most researchers are developed around doing the work well: managing projects, protecting quality, analysing carefully, and delivering on time. Those are necessary skills, but they do not automatically teach someone how to challenge a client, make a commercial call, or take a position when the evidence is incomplete.

Judgement is learned in the work itself. Researchers need to be in the room when the business problem is still messy, when the client pushes back, when trade-offs must be made, and when someone must decide without perfect information.

That experience needs to come earlier in a research career, not after someone has already spent years proving they can execute.

We Need to Train Researchers to Colour Outside the Lines

Researchers are not always trained to colour outside the lines. Quant researchers, especially, are taught to be disciplined, consistent, and careful about what the data can support. Those habits matter, but they can also make us less comfortable with the ambiguity that strategy demands.

Marketers are generally more used to making calls with imperfect information. Researchers are often more comfortable explaining the evidence than pushing beyond it.

The interesting work sits somewhere between the two.

Strategy is rarely data alone, and it is rarely creativity alone. It is the point where evidence, context, experience, and imagination come together to shape a decision.

The shift is not away from discipline. It is toward being able to combine discipline with imagination and still make a call when the data does not point neatly in one direction.

From Insight to Foresight

The next step for market research is not simply better insight … it is foresight.

Insight helps explain what is happening and why. Foresight looks further ahead: what might change, what could disrupt the current trajectory, and what a business should start preparing for before the evidence becomes obvious.

For brands, that difference is commercially important. Insight often explains behaviour we can already see. Foresight extends the decision horizon, giving businesses more time to respond to shifts in consumers, culture, technology, or competition before those shifts are fully reflected in the numbers.

This is not about pretending researchers can predict the future. They cannot. The value lies in identifying plausible directions early, spotting weak signals, and asking what those signals could mean if they strengthen.

That is where researchers start to look a little more like futurists. Not forecasters with a crystal ball, but people who can connect developments across markets and culture, challenge the expected trajectory, and help businesses prepare for more than one possible future.

As AI makes it easier to summarise what has already happened and analyse what is happening now, that ability to look forward becomes a much clearer point of differentiation.

That thinking is also behind THEN. NOW. NEXT., our 35th anniversary campaign at Kadence.

THEN is the context: where markets have come from, what has shaped them, and which assumptions still influence decisions today.

NOW is the evidence in front of us: what consumers are doing, how categories are shifting, and where businesses are gaining or losing ground.

NEXT is where the real differentiation begins. It is about using judgement, foresight, and commercial context to help clients see what may be coming before it is fully visible in the data.

That has always been the point of good research: not simply to describe change, but to help businesses respond to it early enough to matter.

If your organisation is thinking about how research can move from insight to foresight,  we’d love to talk. Kadence helps businesses understand what is changing, what it could mean next, and where to act before the opportunity — or risk — becomes obvious.