If market research leaders are worried about declining budgets, AI disruption and the perceived commoditisation of insight, we should begin with an uncomfortable question: How much of this is our own fault?

The IPA Bellwether shows market research is the only major area of marketing investment seeing sustained decline despite overall marketing budgets increasing. At the same time, conversation in the industry is dominated by concerns about AI, synthetic data, automation and the future of the profession.

The narrative is becoming familiar: AI is making research cheaper; technology is commoditising what we do; clients do not value insight like they used to. There may be some truth in this, but there is another possibility; perhaps we have spent years underselling what research is actually worth.

For decades, the industry has become increasingly focused on the mechanics of research. We talk about online sample sizes, methodologies, dashboards, platforms, fieldwork quality, automation and speed. We describe ourselves as generators of evidence and collectors of data. However, clients were never really buying any of those things.

Clients do not buy surveys, dashboards, or sample sizes. They buy confidence, reduced risk, and better decisions. Yet somewhere along the way, the industry began to confuse the ingredients with the outcome.

In doing so, we created the perfect conditions for our own commoditisation. If the value of research is access to respondents, then panels become interchangeable. If the value of research is chart production, then dashboards become interchangeable. If the value of research is reporting data, then AI will inevitably become interchangeable with researchers.

Perhaps that is why so much discussion around AI feels simultaneously alarming and underwhelming. Alarming because the technology is advancing rapidly; underwhelming because much of what is supposedly under threat was never the most valuable part of research in the first place.

Part of the problem may be that the industry has sometimes expanded beyond its natural reach. Not every decision requires research. When applied to low-risk, low-consequence decisions, it can appear slow, costly and time-consuming. Research creates its greatest value where uncertainty is high, stakes are meaningful and the cost of getting the decision wrong is significant.

The best researchers do not simply tell clients what consumers said, they explain what it means, why it matters and what should happen next. There are two types of decisions in the world: puzzles and problems. Puzzles have a fixed, correct answer; problems have multiple plausible answers. Most commercial decisions are problems, meaning most organisations are trying to find the best possible answer.

Should we launch? Reformulate? Reposition? Raise the price? Enter a new market? Redesign the pack? These are the decisions brands face constantly. They rarely suffer from a lack of data. The challenge is understanding what the data means and having confidence in the decision that follows.

In the industry’s pursuit of efficiency, we have increasingly embraced black box metrics, automated dashboards, synthetic respondents and templated approaches that promise answers with less human involvement.

Many of these tools are genuinely useful. AI has enormous potential to remove friction, automate routine tasks and help researchers process information more effectively. But there is a difference between automating administration and automating judgement. When clients receive dashboards instead of debriefs, benchmarks instead of perspectives and outputs instead of recommendations, research becomes easier to ignore. The voice of the consumer becomes just another input competing with internal opinion.

As an industry, we have often described ourselves as providers of information when our real value lies in helping clients make sense of uncertainty with that information.

Perhaps that is why agencies that continue to thrive often look very different from those predicted by conventional wisdom. They bring a point of view before the research begins, they connect findings to wider cultural, category and behavioural shifts, and they challenge assumptions rather than simply validate them.

Interestingly, some of the most talked-about new entrants to the industry are built around precisely this idea. Their proposition is rarely faster fieldwork, larger sample sizes or more sophisticated dashboards. Instead, they promise senior involvement, clearer thinking and closer proximity to decision-making. Whether operating as consultancies, independents or specialist agencies, they are responding to the same underlying demand: clients want help making decisions, not simply more data.

The future of research is unlikely to be determined by who can collect the most data or build the fastest dashboard. Those capabilities are increasingly accessible to everyone. The organisations that thrive will be those that can turn information into understanding and action.

If we want clients to value research more highly, we need to stop selling the mechanics of research and start selling its impact. Research was never at its best when it delivered data. It was at its best when it delivered impact.

Patrick Young is the UK managing director and global board member at PRS in Vivo