---
user-invocable: true
name: user-feedback-synthesis
description: Turn messy user feedback into clear patterns and priorities
tokens: ~400
cloud-ok: true
---

# User Feedback Synthesis
#claudeai

## When to Use
You have feedback from multiple users (calls, emails, surveys, support tickets) and need to find the signal.

## What I Need
Paste the feedback - transcripts, notes, survey responses, whatever you have.

## Process

### 1. Categorize Every Piece of Feedback

| Category | What it means |
|----------|---------------|
| **Pain** | Problem they're experiencing |
| **Request** | Feature/change they want |
| **Praise** | What's working well |
| **Confusion** | They don't understand something |
| **Churn signal** | Reason they might leave |

### 2. Find Patterns

```
## Feedback Synthesis: [Date/Source]

**Sample size:** [X users]

**Top Pains (by frequency):**
1. [Pain] - mentioned by X users
   - Representative quote: "[quote]"
2. [Pain] - mentioned by X users
3. [Pain] - mentioned by X users

**Top Requests:**
1. [Request] - X mentions
2. [Request] - X mentions

**What's Working:**
- [Thing users love]
- [Thing users love]

**Confusion Points:**
- [Where users get stuck]

**Churn Risks:**
- [Why people might leave]

**Surprises:**
- [Unexpected insight]
```

### 3. Translate to Action

**Build next:** [Request that solves top pain]
**Fix now:** [Confusion point or churn risk]
**Keep doing:** [What's praised]
**Investigate:** [Surprising insight that needs more data]

## Key Questions to Answer
- What problem comes up most often?
- What would make the biggest difference to retention?
- What are users trying to do that we make hard?
- What do happy users have in common?

## Watch Out For
- Loudest user ≠ most representative
- Feature requests often mask underlying problems
- "Nice to have" ≠ "would pay for"
- 1 user saying something strongly ≠ pattern