Research Methodsfrom observation to insight

If Heart is cognitive empathy calibrated to evidence, research is how you earn that evidence. It means first-hand contact with the people you are designing for, close enough and structured enough to produce insight the design can be held accountable to.

When a central team runs research, it is easy to lose touch with the people you are designing for. The work reaches you scheduled, gated, and handed down as synthesis. Stay close anyway. Sit in on sessions when you can, read the raw transcripts rather than the summary decks, and build the relationship with the research team that lets you ask for the encounters your decisions need. They own the process. You own the decision, and the understanding behind it has to be yours.

Value Shaping (2.2) asks the questions that matter here: who is this for, and what does it mean to them? Only first-hand contact can answer them. Four methods do the most work at the practitioner level: observation, interview, diary study, and jobs-to-be-done. They are ordinary, and under-practised, because they produce slower, messier artefacts than a synthesis slide does.

• Observation: Watch someone do the real thing, in their own context, with as little steering from you as possible. As they file an expense report, triage their inbox, or pick up a delivery, notice what they do, what they mutter to themselves, what they avoid, and what they have stopped noticing. Erika Hall’s framing: most of what people report is their theory of the problem, one step from the problem itself.¹ Observation is how you catch the difference.
• Interviews: The method most often done badly, and the easiest to do badly without noticing. The common mistake is treating an interview like a questionnaire: a fixed list of questions asked in order to pull out answers. A good design interview is closer to a conversation about specific things that actually happened. “Tell me about the last time you tried to do X” gives you evidence. “Do you usually do X?” gives you self-report, which is almost always wrong. People tell you their theory of what they do, not what they did. Indi Young’s method: listen for how a person actually thought their way through a real situation, underneath the opinions they offer about it.²
• Diary studies: Observation stretched across time. Someone records what they do, when, and what set it off, across a week or a month. It catches what a single session never will: how use changes once the novelty wears off, the habit loops that form, the moment someone drifts away for good. Heavy to run well, and under-used because of it. Worth it when a decision rides on how people behave over time.
• Jobs-to-be-Done: More a way of reading research than a method of its own, and it changes how you use the other three. Christensen’s frame asks: what job is this person hiring the product to do, and in what situation?³ Read an interview that way and “what did you like about this feature?” becomes “what were you trying to get done, the last time you reached for it?” The second gives you evidence, because it lands on a real moment and the outcome that person was after. Jobs-to-be-Done also bridges design and business (4.3). Here, it helps you organise what research gives you and carry it to the next step.