Sometimes the most informative person to interview is unusual. They manage several accounts, recovered from a failed migration or stopped using a feature everyone else in the customer community praises. Purposive sampling deliberately selects people or cases because their experience can illuminate the research question.

It is a non-probability approach, also called judgement sampling. Selection follows a reasoned purpose rather than a chance mechanism. That makes the rationale central: the researcher should be able to explain why these participants were useful and what their inclusion leaves unresolved.

Choose the variation the question needs

A study may recruit people meeting a specific criterion, seek specialist expertise, examine contrasting circumstances or focus on unusual cases. These choices answer different needs. Maximum variation can help explore how an experience differs across contexts; an extreme case can reveal a constraint that a typical case would not encounter.

BetterEvaluation’s overview of purposive sampling describes selection based on characteristics relevant to the inquiry. In product work, the same principle means recruiting for the activity or decision being studied, not merely assembling people with an attractive job title.

For a fictional scheduling product, a team investigating last-minute shift changes might recruit managers who recently handled absences, employees asked to cover and people who declined. Interviewing only the managers who approved the purchase would offer a much narrower account of the process.

Make the selection rationale reviewable

Write eligibility criteria and the intended mix of cases before recruitment. Explain which differences matter and why, then use appropriate screeners to establish relevant experience. Avoid requesting intrusive proof when a proportionate question or check would suffice.

Recruitment may evolve as findings emerge. An interview could reveal that temporary staff encounter a distinct problem worth investigating. Record that change and its reason, so readers can distinguish a deliberate extension from opportunistically collecting only supportive accounts.

Look for cases that might challenge the emerging explanation. If every participant was selected because they experienced the same difficulty, the study may explain that difficulty well while offering little basis for judging how widely it occurs or what happens when it is absent.

Keep description separate from population estimates

You can accurately report that four of six participants described a workaround. The proportion describes this sample; purposive selection alone does not establish that two-thirds of the wider population use it. Further inferential claims would require an appropriate design or explicit modelling assumptions and supporting evidence.

Qualitative findings can still be valuable beyond the immediate sample. Detailed accounts of circumstances, actions and consequences help readers assess whether a finding may apply elsewhere. Generalizability depends on that reasoning and additional evidence, rather than a sample-size rule.

Report who was sought, how people were recruited and whose perspective remains absent. If referrals are used through snowball sampling, consider how social or professional networks shape the cases reached. Purposeful choice makes research focused; transparency about that choice makes its contribution easier to judge.