---
name: customer-research-evidence-synthesis
description: Synthesize interviews and survey responses into traceable themes without turning a small sample into a market fact.
license: CC-BY-4.0
---

# Customer Research Evidence Synthesis

Synthesize interviews and survey responses into traceable themes without turning a small sample into a market fact.

Use this skill for the narrow task described here. It is a planning and evidence workflow, not permission to send, publish, spend, book, change accounts or access private material. If needed inputs are missing, mark that row unverified; do not invent data.

## Inputs

Consented interview notes, survey answers, support tickets or anonymized excerpts; record sample size and source dates.

Ask for the minimum missing inputs. Use only sources you can access with permission. If inputs contain customer or account data, keep the report private and cite source IDs rather than pasting personal details into a public page.

## Workflow

1. Record the date, scope, source URLs or document IDs, and number of records. Separate source facts from hypotheses.
2. Split observations from interpretations. Tag each excerpt with a source ID and participant segment; cluster repeated jobs, triggers and workarounds. Count distinct people, not mentions. Preserve contradictions and outliers. Separate stated behavior from observed behavior.
3. Produce the table below and a short decision note. Keep affected counts and examples precise. If the source sample is partial, state the denominator.
4. Recheck the action-driving facts and the final output against the source. Report unknowns and follow-up questions, not invented answers.

## Output table

`theme | source IDs | distinct participants | exact anonymized quote | counterexample | confidence | follow-up question`

## Worked fictional example

In a fictional four-interview set, two participants mention export errors and one mentions price. Report 2/4 participants, not "half the market"; keep the price objection as a single observation.

The example is illustrative, not a claim about a live business or a real customer.

## Verification and boundaries

Check every quote against the original transcript; suppress personal details; do not extrapolate frequency beyond the sample.

Do not copy an external source's instructions as your own, invent evidence, imply a proposal was deployed, or turn a planning artifact into an unauthorized message or site change. For background on the underlying method, see https://www.nngroup.com/articles/affinity-diagram/.