Two people interacting during a street interview in a city environment
Research Method

Fly on the Wall

A passive observation technique for capturing real behavior without interfering. No questions. No prompts. Just watching, listening, and recording what people actually do when they think no one is designing for them.

Hand-drawn wireframe design on paper for a UX project
The Method

What it is and why it works.

Fly on the Wall is a pure observation method. The researcher positions themselves in the environment where the behavior happens — a workplace, a home, a public space — and records what occurs without directing, prompting, or interrupting the people being observed.

The strength of this approach is its honesty. When you ask someone how they do something, they give you the version they believe is correct. When you watch them do it, you see the version that actually happens — the workarounds, the shortcuts, the frustrations they have stopped noticing because they have adapted to them.

This method is especially powerful early in a project, when you are trying to understand a problem space before committing to a solution. It is less useful when you already have a specific hypothesis to test, because passive observation does not let you probe specific behaviors on demand.

Best for:

Discovery and problem framing. Identifying unspoken pain points, environmental constraints, and behavioral patterns that interviews would miss.

Not for:

Validating a specific feature, testing a prototype, or gathering quantitative data. Use contextual interviews or usability testing for those instead.

Step by step

How to run a Fly on the Wall session.

  1. Define what you want to understand.

    Write down one or two research questions before you enter the field. You are not testing a hypothesis — you are framing what you pay attention to. Without this, you will drown in data and surface nothing useful.

  2. Get access and consent.

    Arrange permission to observe the environment. Tell people you are there to learn about how they work or live. You do not need to explain your research questions, but you do need informed consent. Be transparent about recording.

  3. Find your spot and disappear.

    Position yourself where you can see the activity without being in the flow of it. The goal is to become part of the background. The first 20 minutes are usually stilted — people behave differently when they know they are being watched. Wait it out.

  4. Record what you observe.

    Take field notes: timestamps, actions, objects, sequences. Note the environment — lighting, layout, tools, interruptions. If you have consent, take photos or video. Capture what people do, not what they say about what they do.

  5. Look for patterns, not moments.

    A single interesting observation is an anecdote. A pattern repeated across three or more people is an insight. After the session, review your notes and look for recurring behaviors, workarounds, environmental friction, and moments where people improvised a solution.

  6. Translate observations into design implications.

    For every pattern you identify, write a single sentence describing what it means for your project. Not "users are frustrated" — but "users create their own filing system because the official one does not match how they actually work." That is the kind of statement that reshapes a design.

Field notes

Things that go wrong, and how to avoid them.

The observer effect.

People change behavior when they know they are being watched. Mitigate it by staying longer, staying quiet, and giving people time to forget you are there. The first half hour is noise.

Confirmation bias.

You came in looking for evidence of a problem. You will find it, whether it is real or not. Counter this by writing down observations that surprise you or contradict your assumptions — those are the most valuable notes you will take.

Note overload.

If you try to capture everything, you will capture nothing. Focus on actions, sequences, and environmental details. Ignore small talk unless it reveals something about the work.

Going alone.

Two observers see different things. Whenever possible, run sessions in pairs. Compare notes afterward — the gaps between what each of you noticed are often the most interesting findings.