Oliok Field Notes / Kit 01

A practical kit for observing everyday systems

Method
Observe without surveilling
Keep facts separate

Observation is a simple way to understand how everyday systems actually work before deciding what to change. This kit helps you choose a bounded scene, record what happens, and separate what you notice from what you assume. It is designed for ordinary settings such as a queue, a shared workspace, a kitchen routine, or a handoff between people, with privacy and consent treated as part of the method.

Choose a bounded scene

Start with a scene small enough to watch without chasing every detail. Pick one place, one routine, and one short time window: the first ten minutes of opening a shop, the moment a delivery arrives, or the path from request to response in a shared inbox. Write down where observation begins and ends. Name who or what is included, and what is deliberately outside the frame. A bounded scene prevents the notes from becoming a vague opinion about everything. It also makes the exercise repeatable, because someone else can return to the same setting and compare the pattern.

Record sequence, time, and friction

Capture events in the order they happen. Use plain language: person enters, form is found, question is asked, item is moved, reply is delayed. Add rough times where useful, such as start, pause, wait, handoff, and finish. Note friction without blaming people. Friction might be a missing label, a repeated question, a locked cabinet, an unclear next step, or a tool that requires extra checking. Avoid recording private details that are not needed. The goal is not to monitor individuals. The goal is to understand how the system shapes actions.

Separate observation from interpretation

Keep two columns: what was observed and what it might mean. In the first, write only visible or audible facts: three people waited, the sign was behind the door, the same number was entered twice. In the second, mark interpretations as provisional. Ask for consent before observing people closely, and stop if the setting becomes private or sensitive. Remove names, exact addresses, and unnecessary identifiers. End with a repeatable checklist: boundary, sequence, approximate timing, friction, direct facts, possible meanings, privacy check, and one question worth testing.

Turn one observation into a safe reversible trial.

Try a small experiment