<aside> 👁️ Witness is about seeing clearly before acting. Observe the work as it actually happens, not as people describe it.
</aside>
Objective: Document where work breaks down, why it breaks, and how much it costs. Provides structured evidence to move from observation to prioritized action.
When to use: After completing initial observation sessions, as patterns emerge. Pull this out when you have raw observations and need to organize them into a prioritized map.
Inputs: Observation session notes, process documentation, interview transcripts.
Position: Architect leads. Scout validates findings against field reality.
How to fill out: One row per friction point. Capture the name, where in the process it occurs, frequency, time impact per occurrence, who is affected, and current workarounds. Rate severity using the thresholds below.
<aside> 📊 Severity: High = >60 min/day impact OR >20% of transactions. Medium = 15-60 min/day OR 5-20%. Low = <15 min/day AND <5%. Adjust thresholds for your volume.
</aside>
Objective: Capture how work actually happens, not how documentation says it should. Raw observation data that feeds Friction Mapping and User Flow Mapping.
When to use: Throughout Witness. This is the foundation; everything else builds on it. Schedule 60-90 minute observation sessions.
Inputs: Access to people performing the work, scheduled observation time, permission to observe without disrupting.
Position: Any team member; often Architect or Scout.
How to fill out: One row per observation. Record what happened (not what should happen), mark anything surprising, and flag items that need follow-up. Confirm your summary with the person you observed.
<aside> 💡 Key questions during observation: "What just happened there?" "Why did you do it that way?" "What would happen if you skipped that step?" "How often does that exception happen?"
</aside>