Write It Down: The Case for Verbatim, Unfiltered Discovery Notes

There is a consistent temptation during discovery interviews to interpret as you go: to hear something a stakeholder says and immediately translate it into insight-language, to place it in its likely category, to correct the framing from solution-language to need-language, to smooth the messy contradictions into coherent observations.

Resist this. Write everything down verbatim, unfiltered, and uninterpreted. Sorting is the next phase's work.

Why Filtering During Discovery Costs You

When a designer interprets during an interview, they are doing two cognitively demanding tasks simultaneously: maintaining conversational engagement and analytical processing. Something must give. Usually it is the listening — the designer starts half-hearing what is said because they are busy deciding what category it belongs to.

The other cost is subtler and more significant: premature interpretation destroys information. The raw statement "we've mapped parts of the experience before, but we don't have a living view" contains more signal than any translated version of it. It tells you something about the speaker's awareness of past attempts, their frustration with static artifacts, and their mental model of what "living" means. A translated version — "stakeholder perceives a gap in journey continuity" — loses all of this. It is correct but thin.

Verbatim notes preserve the specificity, the emotional register, and the organizational language that makes insights actually useful rather than merely accurate in a general sense.

"Write everything in approachable language, keep contradictions, keep even wild assumptions. They will be resolved later, in Tidying Up."

What Verbatim Looks Like

In practice, verbatim discovery notes look rough. They contain fragments. They include statements that contradict each other. They mix customer observations with organizational opinions with half-formed ideas. They sound like someone talking, not like research documentation.

This is exactly correct. The tidying phase will impose order. The discovery phase should preserve chaos.

Specific things worth capturing verbatim:

Direct quotes from interviewees. The exact words someone uses — "we've mapped parts of the experience before" versus "we've done journey mapping before" — carry meaning that paraphrasing erases. Direct quotes are also more persuasive in alignment workshops than summaries: they create the sense that real people said real things.

Contradictions. When a stakeholder says the experience is working well in the same breath as describing a persistent problem, do not resolve the contradiction. Write both statements down. Contradictions often reveal the gap between official narrative and operational reality — which is exactly what you are trying to surface.

Emotional signals. When someone's tone shifts, when there is a pause before answering a specific question, when frustration or enthusiasm surfaces unexpectedly — note it. These signals indicate which problems are felt most acutely and which are accepted as background noise.

Letting Patterns Emerge Rather Than Imposing Them

The reason verbatim notes are more valuable than pre-interpreted ones is that patterns in raw data are more reliable than patterns in already-filtered data. When a designer interprets as they collect, they import their own analytical assumptions into the material. The patterns they find later will partly reflect their pre-existing hypotheses rather than the actual distribution of the evidence.

When the raw material is preserved and clustering is done later — in a Tidying session, away from the pressure of live conversation, with the full corpus of interviews visible — patterns emerge from the data rather than being imposed on it.

The moment you start finding the same phrase appearing in interviews from different teams, or the same frustration surfacing independently from a colleague, a customer, and a customer service agent, that convergence is powerful. It would not be visible if the notes had been pre-interpreted into a taxonomy that the designer had already decided to use.