A team is asked to “improve collaboration” and immediately starts discussing shared folders and comments. Exploratory research makes room for an earlier question: what does collaboration involve here, and where does it become difficult? The answer may concern access, responsibility or trust rather than a missing communication feature.

Exploratory research investigates a problem space that is not yet well understood. It develops understanding, candidate explanations and more precise questions, using methods that can adapt as the investigation reveals what matters.

Give the exploration a boundary

In a fictional housing-service study, the team might investigate how applicants gather information from other household members. That focus leaves room for unexpected practices while keeping the work connected to a decision. “Learn everything about applicants” would be too broad to guide recruitment or analysis.

Start with relevant existing evidence, including previous studies, service records and known constraints. GOV.UK’s discovery research guidance describes learning about users and the wider problem before committing to a service design.

Interviews and observation can explore unfamiliar experiences, while quantitative data may reveal patterns that deserve investigation. Exploration is a purpose, not a synonym for one data type or a guarantee of a small sample.

Follow useful surprises without losing the question

Recruit for experiences relevant to the uncertainty, including differences that might challenge the team’s initial assumptions. If an early conversation reveals that some applicants rely on relatives to complete the form, the next round may need to include those relatives rather than repeatedly interviewing only the named applicant.

Record changes to the guide, sample and focus. Adaptation is part of the design, but readers should still be able to understand how the investigation evolved and what each part of the evidence covers.

Use open questions without assuming that every answer explains the underlying cause. Contextual inquiry may help investigate a practice people find difficult to describe, while secondary research can test whether an apparent discovery is already documented elsewhere.

Make the output useful without overstating certainty

An exploratory study can produce a grounded account of a problem, possible segments, relevant language or hypotheses for further investigation. Distinguish those outputs from estimates of population prevalence or confirmed causal effects.

In the housing example, the finding may be that coordination across household members is a meaningful barrier in the studied cases. A later study could investigate how widely it occurs, or test a design that allows information to be gathered separately. The next step follows from what remains uncertain, rather than an obligation to turn every qualitative finding into a survey.

Exploration should end with a clearer account of what the team knows and which decisions the evidence can support. Sometimes it changes the proposed solution; sometimes it shows that the original problem was poorly framed. Both are useful outcomes when they prevent the team from answering a narrow question about the wrong thing.