Definition:
Shadowing is a contextual research technique in which a researcher accompanies and observes a participant while they perform real activities. Its purpose is to understand actions, decisions, interruptions, tools and conditions that may be difficult to reconstruct through an interview.
Observation may be supplemented with brief questions and with notes, photographs, audio or video when permission has been granted. It is not covert surveillance or the remote recording of any digital behaviour. It requires a defined purpose, informed consent and clear limits on what will be observed and documented.
How a shadowing session is conducted
Research begins by defining an activity, environment and set of questions. The accompaniment should alter the routine as little as possible, although the observer’s presence is never completely neutral. A basic process includes these working stages:
- Preparation: Define participants, duration, tasks, consent and the data that may be collected.
- Observation: Record actions, sequences, time, tools, interruptions and environmental conditions.
- Contextual questions: Request clarification without turning the entire activity into an interview.
- Closing: Check initial interpretations and allow the participant to add context.
- Analysis: Organise evidence, identify patterns and separate observed facts from inferences.
The researcher may take a mainly passive role or interact at agreed moments. The choice depends on the task and the risk of disrupting it. In both cases, a time-based record helps connect each observation to its specific situation.
Information that shadowing can provide
Shadowing allows researchers to compare what a person says they do with what occurs during the real activity. A difference does not mean that the participant is mistaken: many decisions are automatic or depend on circumstances that are difficult to recall.
- Real sequences: The order followed to complete a task and steps that are skipped or repeated.
- Friction points: Waiting, errors, tool changes and requests for help.
- Workarounds: Improvised solutions used to compensate for limitations in a product or process.
- Context: The influence of space, other people, devices and time constraints.
This evidence may become an insight when interpreted alongside other sources and connected to a defined question. A single observation does not by itself demonstrate a general need or indicate its frequency.
Applications in user experience
The technique is used in UX research, service design, organisational studies and process analysis. Its value lies in examining an experience in its real environment, not in guaranteeing a solution or commercial outcome.
- Digital products: Observe how a tool fits alongside other tasks and channels.
- Services: Understand journeys combining face-to-face support, calls and online processes.
- Internal work: Identify dependencies, information handovers and undocumented manual tasks.
Findings may enrich a customer journey or a design thinking initiative. They should be translated into testable hypotheses before changing a product, service or process.
Differences from other methods
A focus group gathers perceptions through a group discussion, whereas shadowing observes individual activities in context. An interview explores explanations and memories; shadowing reveals actions and conditions as they occur.
It is not equivalent to digital analytics or social media monitoring. Those sources record aggregate events or platform interactions, but do not necessarily explain the environment or reasons behind an action. They can complement observation of online behaviour, although they answer different questions.
Limitations and participant protection
Shadowing requires time to recruit participants, observe, document and analyse. People may also change their behaviour when they feel observed. Quality depends on researcher experience, case diversity and traceability between evidence and interpretation.
- Privacy: Avoid information outside the objective and protect personal or confidential data.
- Consent: Explain how evidence will be used and allow the session to stop or records to be excluded.
- Scope: Do not generalise qualitative findings as if they represented the entire population.
Combining shadowing with interviews, metrics or subsequent tests helps establish whether a pattern recurs and how important it is. The result is contextual evidence, not an automatic statistical representation.