Few practices enjoy the near-universal support that primary research does. Suggest skipping user interviews, field visits or customer workshops and you’ll likely be treated as reckless, arrogant, or both. Entire industries have emerged around the belief that better research leads to better decisions.
Yet I’ve increasingly found myself asking an uncomfortable question: does primary research genuinely improve outcomes, or has it become an expensive ritual that helps organisations feel more confident about decisions they were going to make anyway?
This isn’t an argument against research. I’ve interviewed customers, facilitated workshops, observed frontline staff, mapped journeys and synthesised findings into reports that were presented to executives, boards and government departments. I’ve seen research uncover critical problems that nobody in the organisation understood. I’ve also seen months of research generate little more than a professionally documented version of what everyone already knew.
Part of the problem is that many people misunderstand what primary research is actually capable of doing. A common criticism is that interviewing fifteen people cannot possibly be statistically representative of millions of customers. That’s true, but it is also missing the point. Most qualitative research is not attempting to measure the prevalence of a behaviour across a population. It is trying to understand why something is happening in the first place.
If ten customers independently describe the same workaround, frustration or barrier, the significance is not that exactly 66.7% of participants experienced it. The significance is that the issue exists at all, and that the organisation may have been completely unaware of it.
Where things become more complicated is when organisations start treating exploratory insights as if they were objective facts. Anyone who has worked in large organisations has seen this happen. A handful of interviews become evidence of a widespread trend. A quote from a participant gains an authority that far exceeds what the underlying evidence can support.
This is where research begins to drift from investigation into storytelling. Findings are assembled into narratives that help stakeholders make sense of complexity. Sometimes those narratives are useful. Sometimes they are simply persuasive.
There is another reality that practitioners rarely discuss openly. Much of the value of research has little to do with the participants being studied. Instead, the value comes from the people observing the research. Watching a customer struggle through a process often changes minds more effectively than any report, dashboard or business case.
In that sense, organisations frequently use research as a socialisation tool rather than a scientific one. They use it to build consensus, reduce resistance, create shared understanding and give stakeholders confidence to proceed. There is nothing inherently wrong with that, but it is very different from the way research is often described. We tend to talk about discovering insights when much of the practical benefit comes from helping people believe them.
The difficulty is that confidence is not the same thing as correctness. Research can increase confidence in a poor decision just as easily as it can improve a good one. It can conduct extensive discovery and still misunderstand the underlying problem. The presence of research does not guarantee the quality of the judgement that follows.
This becomes particularly apparent when projects continue researching long after they have learned anything new. Additional interviews are commissioned. More workshops are scheduled. New rounds of validation are undertaken. The activity continues not because uncertainty is being reduced, but because decision-makers remain uncomfortable making a commitment, becoming a substitute for leadership.
Every experienced researcher eventually encounters diminishing returns. The first few conversations often generate entirely new information. The next group provides confirmation and nuance. Eventually the findings stabilise and very little changes. Continuing beyond that point may produce more data, but it rarely produces proportionally better decisions.
The irony is that some of the most successful products and services in history emerged with remarkably little formal research. Their creators observed something, formed a hypothesis and acted. Equally, some of the most thoroughly researched initiatives have failed spectacularly despite extensive evidence gathering. Outcomes are influenced by execution quality, market conditions, timing, organisational capability and countless other variables that research alone cannot control.
Perhaps the question organisations should ask is not whether they need more research, but what decision the research is intended to support. If the answer is unclear, the research probably is too. Good research should reduce uncertainty around a specific choice. It should help distinguish between competing options, expose assumptions or reveal consequences that were previously hidden. If it cannot do those things, its value becomes much harder to justify.
Research remains one of the most powerful tools available to designers, product teams and policymakers. But treating it as inherently valuable is no different from treating technology, process or strategy as inherently valuable. Its worth depends entirely on what it changes.
The organisations that gain the greatest benefit from research are not necessarily those that do the most of it. They are the ones that understand when enough has been learned to make a decision. That judgement is rarely found in a methodology, a recruitment plan or a discussion guide. It sits with the people responsible for acting on what they have heard, and their willingness to accept that certainty was never available in the first place.
