An intercept survey meets people during an activity: leaving a shop, visiting an exhibition, browsing a website, or using a service. It can ask what brought them there, whether they achieved their goal, or how they experienced a particular moment.
The format overlaps with in-product surveys, but is broader than digital product use. Its defining feature is recruitment through the current encounter, with the resulting opportunities and limits that creates.
Decide who can encounter the invitation
A website prompt shown after two minutes excludes people who leave sooner. An exit survey outside a shop misses people using another door or visiting at another time. Describe those conditions as part of the sampling process.
An illustrative museum survey might sample visitors across different days, times, and exits to improve coverage. Asking everyone available during one quiet afternoon would answer a narrower question, however many complete the questionnaire.
Nielsen Norman Group’s discussion of when to use surveys includes intercepts as a way to gather information during a visit. The invitation still reaches a selected group, and those who agree may differ from those who decline.
Ask a question the moment can support
An arrival prompt can investigate intent but cannot yet ask whether the visit succeeded. A later prompt can ask about the experience, while recognising that some people have already left.
Keep the questionnaire short and avoid disrupting a consequential task. In person, use a clear invitation and an easy refusal; online, make dismissal straightforward. Informed consent and appropriate handling of any personal information remain relevant in either setting.
Use neutral wording and preserve the distinction between the person’s stated goal and the researcher’s interpretation of their behaviour. A visitor who leaves without buying may have achieved exactly what they intended.
Report the reach and the gaps
Record where, when, and how invitations were offered, alongside response and eligibility. If estimating beyond respondents, explain the sampling and analysis assumptions rather than presenting a convenience intercept as representative.
Combine the answers with relevant observational or behavioural evidence when useful. An intercept can explain a moment that analytics leaves ambiguous, while analytics may reveal how that moment fits a wider sequence. The interpretation should retain both sources’ limits instead of treating their combination as automatic confirmation.