User interviews produce a lot of notes that turn into nothing. The team did 12 interviews. Each generated 6 pages of notes. The notes are in Notion, in Google Docs, in someone's personal directory. A month later, somebody asks "what did we learn from the interviews?" and the answer is a vague summary nobody can source. The interviews informed the team's intuition; they did not produce the structured insight report that would let the team make decisions from the data. The cost is paid in research that does not change anyone's behavior.
Synthesis turns raw interview input into structured insights: Jobs-to-Be-Done statements, recurring pain points with frequency and severity, and supporting evidence quotes that make the claims defensible. The discipline is well-known among researchers and rarely applied because reading 12 interviews and clustering the themes takes a focused day. The /echo-interview skill compresses that day into minutes by reading the raw input and producing the structured report.
What useful synthesis produces
A useful synthesis report has four parts. JTBD statements: "when [situation], the user wants to [motivation], so they can [outcome]" framing applied across the interview set. Recurring pain points: each pain stated concisely, with frequency (how many interviews mentioned it) and severity (how disruptive it was reported as). Quotes: representative verbatims for each insight so the claim is sourced. Recommendations: what the team could do in response, prioritized by potential impact.
How /echo-interview works
The skill reads the raw input (interview transcripts, notes, support ticket samples, NPS verbatims) and clusters them by theme. Each theme becomes a candidate insight. The skill weights themes by frequency and severity to produce the recurring-pain ranking. Quotes are extracted for each theme so the synthesis is sourced. Recommendations are derived from the patterns: a pain mentioned in 8 of 12 interviews with high severity is a higher-priority recommendation than a pain mentioned in 2 with low severity.
The most common synthesis mistake is over-weighting the loud minority. /echo-interview reports frequency and severity separately so a single very-vocal interviewee does not dominate the ranking.
Tonone's /echo-interview skill synthesizes raw user research input into JTBD statements, ranked pain points, and an insight report with supporting quotes.
When to use /echo-interview
/echo-interview is the right call after a round of user interviews, after collecting NPS verbatims or churn surveys, or when support tickets need to be clustered by theme. Skip it for individual interview reads where the team is exploring rather than synthesizing. For JTBD analysis specifically without the synthesis step, /echo-jobs is the right call.
| Capability | Tonone | Generalist chatbot | Cursor / Copilot |
|---|---|---|---|
| Themes weighted by frequency + severity | Yes, both reported separately | Single ranking | Variable |
| JTBD statements extracted | Yes, structured framing | Free-form summary | Variable |
| Source quotes per insight | Yes, sourced verbatims | Generalized claims | Variable |
| Recommendations prioritized | Yes, by impact | Unranked list | Variable |
| Works on mixed input types | Yes, transcripts + tickets + NPS | One input type | Variable |
Related skills
/echo-interview synthesizes mixed input. /echo-feedback clusters support tickets and reviews specifically. /echo-jobs produces JTBD analysis with switching triggers. /echo-segment builds personas from mixed data.
Install
/echo-interview ships with the Echo agent in Tonone for Claude Code. Install Tonone, paste the raw input, and the skill produces the structured insight report.
1. Add to marketplace
2. Install Echo
Frequently asked questions
- What does /echo-interview do?
- It synthesizes raw user research input (interview transcripts, support tickets, NPS verbatims) into JTBD statements, recurring pain points with frequency and severity, and an insight report with sourced quotes.
- What input types does /echo-interview accept?
- Interview transcripts, customer feedback documents, support ticket samples, NPS verbatims, app store reviews, churn surveys. The skill handles mixed input.
- When should I use /echo-interview?
- After a round of user interviews, after collecting NPS or churn surveys, or when support tickets need to be clustered by theme.
- How is /echo-interview different from a generalist summary?
- A generalist returns prose. /echo-interview produces structured JTBD statements with frequency, severity, and source quotes.
- How do I install /echo-interview?
- Install Tonone for Claude Code via tonone.ai/get-started. /echo-interview ships with the Echo agent. Tonone is free and MIT-licensed.