Site visits produce notes that feel vivid in the moment and confusing a day later. A page may contain timestamps, fragments of dialogue, descriptions of movement, questions, and the observer’s own guesses. If those forms of evidence are blended together, a confident interpretation can easily be mistaken for something that actually happened.
An observation brief turns that page into a small, traceable record. It explains where and when the visit occurred, what was directly observed or heard, what the observer thinks it may mean, what contradicts that interpretation, and what should be checked next.
Start with the visit question and boundary
State the question that guided the visit. “Understand the library” is too broad. “See how first-time visitors find and use the self-checkout station” gives the notes a boundary.
Add the place, time window, and parts of the experience you did not observe. A short visit cannot represent every hour, person, or condition. Writing the boundary at the top prevents the brief from sounding more general than the evidence allows.
Label four kinds of notes
Read each line and label it before summarizing:
- Observed: an action or visible condition.
- Heard: words recorded as a quotation or close paraphrase.
- Inferred: your interpretation of what the observation may mean.
- Unknown: a missing fact that could change the interpretation.
For example, “A visitor looked confused” is an inference. “A visitor stopped, looked at two signs, returned to the entrance, and asked where to scan a card” is observable. The second version gives another reader enough detail to question or support the interpretation.
Normalize time, place, and sequence
Site notes often use phrases such as “later,” “near the desk,” or “someone else.” Replace them with the clearest reference the notes support: “At 10:18, beside the self-checkout station, a second visitor…”
Do not invent precision. If you did not record an exact time, use an approximate sequence such as “after the staff demonstration.” The goal is not a perfect transcript. It is to make the evidence easy to locate and compare.
Group evidence by moment, not by opinion
Arrange notes around moments in the experience: approaching the area, finding the station, starting the task, recovering from an error, and leaving. This structure helps readers see where friction may occur without forcing every note into a single conclusion.
A heading such as “Visitors do not understand the station” already states a verdict. “Finding the starting point” is more neutral and leaves room for both difficult and successful examples.
A worked example: a fictional library visit
Imagine a fictional 25-minute visit to a public-library lobby. The question is: how do visitors begin using the self-checkout station? The raw notes include:
- 10:07 — one visitor walks directly to the station and begins;
- 10:11 — another visitor taps the screen, looks behind the monitor, then asks a staff member where to scan the card;
- staff member points to a reader mounted below the screen;
- small card icon is visible beside the reader;
- 10:18 — a third visitor watches the previous interaction and then starts without asking;
- maybe the reader is difficult to notice.
The last line is an inference, not an observation. A careful brief might say:
Evidence statement: One of three observed visitors asked where to scan a card after looking around the station. Another began immediately, and a third began after watching the staff-assisted interaction.
Interpretation: The card reader may not be equally noticeable to every first-time visitor.
Counterevidence: One visitor used the station without visible hesitation.
Unknown: We do not know whether any visitor had used this station before.
The brief preserves the possible issue without turning one interaction into a universal claim.
Keep counts descriptive
Counts can clarify what you saw, but a small site visit is not automatically a representative sample. Write “one of three observed visitors asked for help during this visit,” not “a third of visitors need help.” The first statement describes the evidence. The second makes a wider claim the visit cannot support.
If several visits show the same moment, link each observation before naming a pattern. Keep counterexamples beside the pattern rather than removing them as noise.
Move from observation to the next question
An observation brief should end with a question or small check, not a fully formed solution. For the fictional library visit, the next question could be: “Can first-time visitors identify the card reader before touching the screen?” A small follow-up might compare what people notice before and after approaching the station.
If the team later combines site evidence with interviews, use the same distinction between source evidence and interpretation. This guide to writing a problem statement from interview notes shows how to connect context, obstacle, and consequence without prescribing a feature.
Copy this observation-brief template
Visit question:
Place and time window:
Observed area or step:
Not observed:
Evidence by moment:
- Observed:
- Heard:
- Inferred:
- Unknown:
Evidence statement:
Possible interpretation:
Counterevidence:
Limits of this visit:
Next question:
Smallest follow-up check:Use the same discipline when reviewing a demonstration: separate what was claimed from what you actually saw. This guide to product demo notes provides a parallel evidence check.
Run a final source check
Before sharing, ask whether every evidence statement can be traced to a note, timestamp, quotation, or permitted recording. Check that interpretations are labelled, counterexamples remain visible, and the limits of the visit are explicit.
Then take six notes from one visit you are permitted to use. Label each as observed, heard, inferred, or unknown. Write one evidence statement, one counterexample, and one next question. A useful observation brief makes the evidence easier to inspect without making it sound stronger than it is.
