Fly on the Wall

Purpose

Fly on the Wall is an ethnographic observation technique where the researcher becomes an unobtrusive observer in a real-world setting. The goal is to capture authentic behaviours, pain points, and environmental factors that people themselves may not articulate in interviews. It bridges the gap between what people say they do and what they actually do.

At a glance

• Group size: 1–3 observers
• Duration: 60 minutes
• Difficulty: Low to moderate
• Facilitation style: Silent observation, no interaction
• Setting: The user’s natural environment (office, shop, clinic, etc.)

How to run

1. Define the focus (10 min before): Decide what you are observing — a process, a space, interactions, or a specific task. Prepare an observation framework (who, what, where, when, how).
2. Get access and consent (beforehand): Arrange permission with gatekeepers. Inform people that observation will happen, but avoid over-explaining to minimise the Hawthorne effect.
3. Observe silently (45 min): Position yourself where you can see without being in the way. Take notes on actions, sequences, workarounds, emotions expressed, tools used, and environmental factors. Use timestamps.
4. Capture artefacts (during): Photograph or sketch the environment, signage, tools, and spatial layouts — with permission.
5. Debrief (15 min after): Share raw observations with your team. Separate facts from interpretations. Cluster related observations.

Materials needed

• Observation template or notebook
• Camera or phone (with permission)
• Sketch pad for environment maps
• Timer

Pitfalls & common mistakes

• Interacting with subjects: Asking questions turns this into a contextual inquiry — resist the urge.
• Confirmation bias: Recording only what supports your hypothesis. Capture everything, filter later.
• Too short a session: Behaviours emerge over time. Stay long enough for patterns to appear.
• Forgetting context: Note the environment, time of day, and any unusual conditions.

Inclusion & accessibility

• Ensure observation does not feel like surveillance — be transparent about purpose.
• Avoid observing vulnerable populations without ethics review.
• Share findings back with participants where appropriate.
• Consider cultural norms around being watched in different settings.

Variations

• Time-lapse observation: Take notes at fixed intervals (every 5 minutes) for a structured dataset.
• Shadowing: Follow one person through their day, still without interacting (a more focused variant).
• Video ethnography: Record the session for later collaborative analysis.

When NOT to use this

• When the behaviour you need to observe happens rarely or unpredictably.
• When the environment is private and observation would be intrusive (e.g., a patient’s home without proper consent).
• When you need to understand motivations — pair with interviews instead.

Source & further reading

• IDEO, *The Field Guide to Human-Centered Design* (2015)
• d.school, Stanford — "Fly on the Wall" method card
• Spradley, James P. — *Participant Observation* (Waveland Press, 1980)