Pawel Huryn
104 lines
3.6 KiB
Markdown
104 lines
3.6 KiB
Markdown
---
|
|
description: Analyze user feedback at scale — sentiment analysis, theme extraction, and segment-level insights
|
|
argument-hint: "<feedback data as CSV, text, or file>"
|
|
---
|
|
|
|
# /analyze-feedback -- User Feedback Analysis
|
|
|
|
Process large volumes of user feedback (reviews, surveys, support tickets, NPS responses) into structured insights with sentiment analysis and segment-level patterns.
|
|
|
|
## Invocation
|
|
|
|
```
|
|
/analyze-feedback [upload a CSV of NPS responses]
|
|
/analyze-feedback [paste app store reviews or survey responses]
|
|
/analyze-feedback [upload support ticket export]
|
|
```
|
|
|
|
## Workflow
|
|
|
|
### Step 1: Accept Feedback Data
|
|
|
|
Accept in any format:
|
|
- CSV/Excel with feedback text (and optional metadata: date, segment, rating)
|
|
- Pasted text (reviews, survey responses, Slack messages)
|
|
- Uploaded documents or exports from feedback tools
|
|
|
|
Ask:
|
|
- What kind of feedback is this? (NPS, reviews, support tickets, survey, etc.)
|
|
- Any segments to analyze separately? (user tier, plan, geography)
|
|
- What are you looking for? (general themes, specific issues, trends over time)
|
|
|
|
### Step 2: Analyze
|
|
|
|
Apply the **sentiment-analysis** skill:
|
|
|
|
- **Sentiment scoring**: Classify each piece of feedback (positive, neutral, negative)
|
|
- **Theme extraction**: Identify recurring topics and cluster related feedback
|
|
- **Frequency analysis**: Count how often each theme appears
|
|
- **Segment analysis**: Break down sentiment and themes by user segment (if data available)
|
|
- **Trend detection**: If dates are available, identify sentiment shifts over time
|
|
|
|
### Step 3: Generate Analysis Report
|
|
|
|
```
|
|
## Feedback Analysis Report
|
|
|
|
**Date**: [today]
|
|
**Feedback analyzed**: [count] responses
|
|
**Source**: [NPS survey / app reviews / support tickets / etc.]
|
|
**Period**: [date range if available]
|
|
|
|
### Overall Sentiment
|
|
- Positive: [X%] | Neutral: [Y%] | Negative: [Z%]
|
|
- Average sentiment score: [X/10]
|
|
- Trend: [improving / stable / declining]
|
|
|
|
### Top Themes
|
|
| # | Theme | Mentions | Sentiment | Segments Most Affected |
|
|
|---|-------|----------|-----------|----------------------|
|
|
|
|
### Theme Deep-Dive
|
|
|
|
#### Theme 1: [Name] — [X] mentions, [sentiment]
|
|
- **What users are saying**: [summary with representative quotes]
|
|
- **Root cause**: [what's driving this feedback]
|
|
- **Impact**: [how this affects retention, satisfaction, or revenue]
|
|
- **Recommendation**: [what to do about it]
|
|
|
|
[Repeat for top 5-8 themes]
|
|
|
|
### Segment Analysis
|
|
| Segment | Volume | Avg Sentiment | Top Theme | Key Difference |
|
|
|---------|--------|-------------|-----------|---------------|
|
|
|
|
### Notable Quotes
|
|
> "[quote]" — [segment, sentiment]
|
|
|
|
### Trends Over Time
|
|
[If date data available: chart-ready data showing sentiment shifts]
|
|
|
|
### Actionable Insights
|
|
1. [Insight + recommended action]
|
|
2. ...
|
|
|
|
### Gaps
|
|
[What this feedback doesn't tell you — suggested follow-up research]
|
|
```
|
|
|
|
Save as markdown. If input was structured data (CSV), also save enriched data with sentiment scores as CSV.
|
|
|
|
### Step 4: Offer Next Steps
|
|
|
|
- "Want me to **create user personas** from these feedback patterns?"
|
|
- "Should I **triage the top themes as feature requests**?"
|
|
- "Want me to **design an interview script** to go deeper on a specific theme?"
|
|
|
|
## Notes
|
|
|
|
- Sentiment analysis is approximate — flag edge cases (sarcasm, mixed sentiment, non-English text)
|
|
- Theme extraction should look for needs behind requests, not just surface-level topics
|
|
- If sample sizes are small per segment, note limited confidence
|
|
- For NPS data specifically, analyze Detractors (0-6), Passives (7-8), and Promoters (9-10) separately
|
|
- Output enriched CSV when input is structured, so the user can use it in their own tools
|