A dispatcher describes assigning a delivery as a quick task in the scheduling system. Following the work for an afternoon may reveal phone calls, handwritten notes and checks with a colleague before the assignment is made. The software interaction is brief because much of the work happens around it.
Shadowing follows a participant through a period of activity, observing the sequence, setting and interruptions as they unfold. It is useful when an interview is likely to compress a complicated routine into a cleaner account than everyday practice permits.
Follow the activity without becoming its director
In the fictional dispatch study, the researcher might accompany someone through a shift or a defined handover. Choose the period to capture the process of interest, including quieter moments and exceptions where relevant. One convenient hour may reveal only a small part of the working pattern.
Agree when questions can be asked. Frequent interruptions can change the activity being studied, while waiting until the end risks losing the context of a decision. Brief questions at natural pauses, followed by a fuller debrief, often help balance those demands. GOV.UK’s guidance on contextual research discusses observing services in real-life settings.
Shadowing overlaps with contextual inquiry, which explicitly combines observation with inquiry into the work. The label matters less than describing how the researcher followed, questioned and recorded the activity.
Record the surrounding work
Keep timestamps and distinguish what happened from what you think it means. “Called a colleague before assigning the job” is an observation; “did not trust the system” is an interpretation that needs further evidence. Record relevant artefacts and conditions with permission, including interruptions and hand-offs.
The researcher’s presence may affect behaviour. Do not assume that this influence disappears after a fixed period or that early observations can simply be discarded. Note signs that the participant is demonstrating an ideal process and explore how that relates to ordinary practice.
Arrange boundaries around colleagues, customers and private information before fieldwork. A participant’s agreement does not automatically cover everyone encountered during a shift. Be prepared to pause or move away when the activity falls outside the agreed scope.
Turn a detailed day into a bounded finding
Compare observations across relevant roles, settings and times rather than treating one person’s routine as universal. Ask about unusual events during a debrief and examine whether an apparent workaround serves a purpose the proposed redesign might accidentally remove.
The findings can inform task analysis, realistic test scenarios and questions for further research. A useful report preserves enough context to explain why the work takes its observed form. “People use spreadsheets” is less informative than an account of the coordination need the spreadsheet meets and the conditions under which the official tool falls short.