Think. Notice. Build.

Three Phases

Every research engagement moves through three phases. Think: frame what you're after — know the question before you look. Notice: go find out what's actually happening — the fieldwork. Build: make the answer real and test if it holds.

These map to the scientific method. Think is the question and hypothesis. Notice is observation and experiment. Build is analysis and conclusion. We just don't call it that because nobody wants to feel like they're in a lab.

Think

Framing the problem before spending time or money looking at the wrong thing. This is where stakeholder interviews surface assumptions, problem framing workshops separate what we know from what we guess, research questions get sharpened, and screener design ensures we talk to the right people.

The output is a research plan that answers: what are we trying to learn, from whom, using what method, and what decisions will this inform.

Notice

Going out and paying attention. The actual fieldwork, deployed based on the question, not the budget.

Generative methods — contextual inquiry, in-depth interviews, diary studies, participatory design, concept testing — for when you need to understand the problem space. Evaluative methods — usability testing, heuristic evaluation, A/B testing, card sorting, accessibility audits — for when you need to assess a solution. Quantitative methods — surveys, analytics review, benchmarking — for when you need to measure how much.

Build

Turning findings into things people can act on. Research readouts tied to specific decisions. Journey maps grounded in observed behavior. Persona frameworks built from patterns in data. Opportunity maps that show where the gaps are, sized by impact and feasibility.

Different audiences get different packaging. Product teams need specific, actionable detail. Leadership needs strategic implications and risk framing. Same research, different format.

What Makes This Different

The researcher who designs the study is the one in the room. No handoffs. We use the participant's language in synthesis, not ours. We record everything and make raw footage available — no filtering. And mixed methods by default, because qual tells you why and quant tells you how much.