Why the map is needed
Scattered interviews and notes on their own do not give a solution: they give noise. The task of the research stage is to turn the noise into a structure where every conclusion shows its ground. This is exactly what the CustDev Agent does: it researches the audience and builds a verifiable opportunity map. "Verifiable" is the key word: every element of the map is linked to a primary fragment that can be opened and re-read.
Stage 1: Preparation
The agent checks the goal of the research, the segment and the sufficiency of the materials. This is an honest input check: if there is little material for the declared goal, it is reported right away, not after a beautiful but empty report. The mistake of most manual reviews is starting the coding without answering the question "what must this research answer".
Stage 2: Coding
The materials are broken down into four types of fragments: observations, situations, jobs and obstacles. An observation — what the person actually did; a situation — in what circumstances; a job — what task they were solving; an obstacle — what got in the way. This separation protects from the main error of interpretation: the desire to immediately "understand what the client wants". First, what was said and done is recorded.
Stage 3: Synthesis
The repeating signals are grouped, but the exceptions are not thrown away — they are shown explicitly. If four respondents say one thing and the fifth — another, it is visible in the map, not dissolved in averaging. Here the contradictions between the respondents are also revealed — they become separate material for hypotheses.
Stage 4: The artifact
The result is a research report with the map of topics, quotes, contradictions, hypotheses and a verification plan. Every topic refers to the primary fragments, so any conclusion can be checked against the sources. The hypotheses are not passed off as facts: each has a verification plan — the next series of interviews, an observation or a test. About the limits of such a conclusion on a small sample — in the article "Five interviews - data or illusion?".
How to use the map
The working scenarios of use are fixed in the agent's passport: gather scattered interviews into a single structure, separate the direct quotes from the interpretations, find hypotheses for the next test and get a reproducible research memo. The map is good when the next person can repeat the researcher's line of thought from the primary sources — without speculation and "everyone knows it anyway".
Questions and answers
What does the artifact include?
A research report: the map of topics, quotes, contradictions, hypotheses and a verification plan — with links to the primary fragments.
Are the quotes separated from the interpretations?
Yes: the coding separates the observations and the situations, and the synthesis groups the signals while keeping the connection to the source fragments.
Does the map replace live interviews?
No: the agent's passport states directly that it does not replace the researcher's conversation with the users — it structures what has already been collected.
What is the opportunity map made of?
Of topics with quotes, contradictions, hypotheses and a verification plan with links to the primary fragments.
How does the agent go from quotes to topics?
The coding separates the observations and the jobs, the synthesis groups the repeating signals and shows the exceptions.
What happens with the exceptions?
They are shown explicitly, not smoothed away: the exceptions are the source of counter-hypotheses.
Can we go back to the primary source?
Yes: the map is linked to the primary fragments, the conclusion is traceable down to the quote.
What is the form of the final artifact?
A research report with the map of topics, quotes, contradictions, hypotheses and a verification plan.
Is the preparation of the materials required?
Yes: the goal, the segment and the sufficiency of the materials are checked at the preparation stage.
Who decides which hypotheses to test?
The team: the agent proposes the map and the verification plan, the prioritization stays with the people.