In research and service design circles, the term now gets applied to almost anything that emerges from a workshop wall. A participant says something frustrating. A cluster of sticky notes forms around a recurring issue. A few themes are labelled and suddenly the project team is presenting a slide titled Key Insights.
But pain-points are not insights. Themes are not insights either. And confusing the two creates a deeper problem than terminology. It fundamentally changes the quality of thinking organisations produce from research. What most teams call insights are usually just organised observations.
A person saying they had to repeat their story five times throughout a healthcare process is an observation. Ten people saying it becomes a pattern. Grouping those patterns into a category called “fragmented experiences” creates a theme. None of that is insight yet. It is still description.
The insight is the explanation for why the pattern exists in the first place.
For example:
Observation
A patient repeats their medical history multiple times across providers.
Pain-point
People are frustrated by having to continually retell their story.
Theme
Fragmented continuity of care.
Insight
The system treats continuity as an information-sharing problem between organisations, while patients experience it as an accountability problem. Because no single actor owns continuity across the experience, patients become the coordinators of their own care.
That final statement changes the nature of the problem entirely. It moves the conversation away from generic complaints about disconnected systems and toward questions of governance, ownership, incentives and operating models. It reveals why the issue persists despite years of investment into interoperability and digital platforms. That is what an insight should do. It should increase understanding, not simply organise evidence.
The problem is that many modern research practices stop at categorisation and mistake that categorisation for synthesis. Affinity mapping, clustering and thematic analysis are useful techniques, but they are methods for structuring information, not generating meaning. A wall full of themes may look impressive, but structure alone does not produce understanding.
Real insight requires interpretation. It requires asking why a behaviour keeps occurring despite everyone already being aware of the symptom. It means looking beyond what people say at face value and examining the systems, incentives, constraints and relationships sitting underneath the experience. Often the most valuable insights emerge not from the loudest complaints, but from the tensions and contradictions hidden between them.
People are usually very good at describing friction. They are far less equipped to explain the institutional or structural conditions creating it. That gap between symptom and cause is where insight work actually begins.
This is also why genuinely useful insights tend to feel slightly uncomfortable when they surface. They challenge assumptions. They expose failures in logic that organisations have normalised over time. Weak insights merely restate what everyone already knows in softer language:
- Users find the process confusing
- Customers want simpler experiences
- People struggle to navigate services
None of these statements explain anything. They describe outcomes without identifying mechanisms. An actual insight reveals why those conditions continue to exist.
For instance, people often say large service systems are difficult to navigate because information is hard to find. In reality, information is rarely the central issue. The deeper problem is usually fragmentation of responsibility. Individuals move between agencies, channels, programs and providers that each understand only a small portion of the overall journey. The burden of stitching those fragments together falls onto the individual, particularly during periods of stress or vulnerability.
That is an insight because it reframes the source of the problem. It shifts intervention away from “more information” and toward coordination, accountability and experience continuity.
A useful test is to ask whether a statement actually changes strategic direction. If removing it from the research would make little difference to decision-making, it probably was not an insight to begin with.
This distinction becomes increasingly important as organisations mature their research practices. Many institutions have become exceptionally good at collecting evidence. Research repositories are overflowing with interview transcripts, journey maps, workshop outputs and thematic analysis. Yet despite the volume of material, the same systemic problems continue appearing year after year because the work often stops at documentation.
The organisation accumulates findings without producing understanding, yet understanding is the part that matters.
An insight is not the pain-point itself. It is the explanation of the forces producing that pain-point, the reason the behaviour persists, and the implication that changes how people think about the problem. It sits one level deeper than the observation. It connects human experience to system logic.
That is the difference between research that documents reality and research that explains it.
