Primary research creates new evidence for an investigation. A team may interview customers about a recent decision, observe a task, field a questionnaire, or introduce an experimental change. The method is chosen to fit the question, with the collection process under the researcher’s direction.
Original collection can provide a close fit to a product or audience, but it is not inherently better than existing evidence. A poorly recruited survey remains weak evidence even when every response is new.
Identify what existing research cannot answer
Review relevant material first. The unresolved question may concern a new audience, a changed product, or a detail that earlier studies did not investigate. State that gap before deciding what to collect.
For an illustrative repair-booking service, existing reports might explain general concerns about reliability but say little about how customers choose an available appointment. A focused usability study could examine that interaction, while interviews might be needed to understand scheduling constraints outside the interface.
The government’s introduction to user research describes how direct research can inform service design. The practical step is to connect the activity to a decision the team can make differently afterwards.
Design the collection around the evidence required
Define the audience and recruit for relevant experience. Prepare materials that do not lead participants towards the preferred answer, and pilot the procedure where misunderstanding or technical problems could compromise it.
Choose observation when actual interaction matters, an interview when an account and context are needed, and a suitable quantitative design when the decision requires estimation or comparison. Mixed methods can connect different forms of evidence when there is a clear reason to combine them.
Manage informed consent, data access, and retention as part of the study. The ability to collect material does not by itself justify collecting every available detail.
Report the evidence with its limits
Explain who took part, what they encountered, how the analysis was conducted, and what uncertainty remains. Keep observations distinct from the team’s interpretation and recommendation.
The result can then join the organisation’s existing knowledge and become secondary evidence for a later question. Good primary research leaves enough context for that future use, so its value is not restricted to the meeting in which the findings were first presented.