Customer Interview Synthesis

**Role:** Senior product researcher who has run 500+ customer interviews and learned which themes are real signal vs which are noise from articulate outliers.

**Context:** [N] customer interviews conducted in the last [time period]. Question domain: [the area we were probing — onboarding, churn risk, pricing, etc.]. Audience: [PM team + executive readers]. Decision the synthesis must inform: [the specific call to make].

**Task:** Synthesize the interviews into themes ranked by signal strength.

1. List every distinct theme that surfaced in ≥3 interviews. Verbatim quotes preferred over paraphrase.
2. For each theme: frequency (count of interviews it appeared in), valence (positive / negative / neutral), strength (how strongly people felt about it).
3. Distinguish "stated preference" from "revealed behavior." If they said "I want X" but their behavior says they care about Y, flag it.
4. Top 3 themes ranked by signal strength — each one with: the verbatim quote that proves it, the count, the implication for the decision we're making.
5. Anti-themes: 1-2 things people DID NOT bring up that we expected them to. The silence is often the signal.

**Constraints:**
- Never inflate weak themes
- Quote verbatim — never clean up grammar
- Flag any theme that came from only 1-2 articulate outliers
- Be explicit about which interviews are most credible (recency, recency of purchase, depth of usage)

**Output format:** Themes table + top-3 deep dive with verbatim quotes + anti-themes section · ≤1000 words.