---
name: customer-feedback-analysis
description: Analyze customer reviews, support tickets, survey responses, and feedback to extract actionable insights. Use when someone needs to understand customer sentiment, identify patterns, prioritize issues, or turn qualitative feedback into product decisions.
---

# Customer Feedback Analysis

Help the user transform raw customer feedback into structured, actionable insights. Work with reviews, NPS responses, support tickets, survey data, social media comments, and any other customer voice data.

## How to Help

When the user shares customer feedback for analysis:

1. **Categorize the feedback** - Group by theme (product quality, shipping, pricing, UX, support, etc.)
2. **Identify sentiment patterns** - Tag as positive, negative, neutral, and note intensity
3. **Extract specific insights** - Pull out concrete, actionable takeaways, not just summaries
4. **Prioritize by impact** - Rank issues by frequency, severity, and business impact
5. **Recommend actions** - Suggest specific next steps based on the analysis

## Analysis Framework

### Step 1: Data Organization
- Group feedback into categories (product, service, price, experience, competitor mentions)
- Tag each piece with sentiment (positive/negative/neutral) and intensity (1-5)
- Note the source channel (review, support ticket, survey, social)
- Flag any urgent or escalation-worthy items

### Step 2: Pattern Recognition
- Identify the top 5-10 recurring themes
- Note which themes are growing vs. stable
- Look for correlations (e.g., negative shipping feedback correlating with specific regions)
- Highlight surprising or unexpected feedback

### Step 3: Insight Extraction
For each major theme, provide:
- **What customers are saying** - Representative quotes
- **How many** - Frequency or percentage of total feedback
- **Why it matters** - Business impact (retention, conversion, word-of-mouth)
- **What to do about it** - Specific, actionable recommendation

### Step 4: Competitive Intelligence
- Flag any mentions of competitors (by name or description)
- Note what customers say they like better about alternatives
- Identify switching triggers and loyalty drivers

## Core Principles

### Let customers speak for themselves
Always include direct quotes. Summarizing loses the emotional context and specific language that makes feedback actionable. "The zipper broke after two weeks" is more useful than "quality concerns."

### Separate signal from noise
One angry review is an anecdote. Twenty similar complaints are a pattern. Help users distinguish between individual outliers and systematic issues worth addressing.

### Quantify wherever possible
"Some customers mentioned shipping issues" is vague. "23% of negative reviews (47 of 204) mention shipping delays, up from 15% last quarter" is actionable. Always count, calculate, and compare.

### Look for the jobs-to-be-done
Behind every complaint is an unmet need. Behind every praise is a job well done. Help users understand not just what customers say, but what they're trying to accomplish.

### Connect feedback to revenue
Frame insights in business terms. "Customers who mention 'easy returns' in positive reviews have a 40% higher repeat purchase rate" is more compelling than "customers like easy returns."

### Track trends, not just snapshots
Single-point analysis is less valuable than trend analysis. When possible, compare current feedback to previous periods to identify what's improving, declining, or emerging.

## Questions to Ask Users

- "What type of feedback is this? (reviews, NPS, support tickets, surveys)"
- "What time period does this cover?"
- "Is there a specific question you're trying to answer with this feedback?"
- "Do you have comparison data from a previous period?"
- "What decisions will this analysis inform?"
- "Are there specific themes or issues you're already tracking?"

## Output Format

### Summary Report
1. **Executive Summary** - 3-5 bullet points of the most important findings
2. **Sentiment Breakdown** - Overall positive/negative/neutral split with percentages
3. **Top Themes** - Ranked list of recurring topics with frequency and representative quotes
4. **Competitive Mentions** - What customers say about alternatives
5. **Action Items** - Prioritized list of recommended next steps
6. **Methodology Note** - Sample size, time period, data sources, any limitations

### Quick Analysis (for smaller datasets)
1. **Key Takeaway** - The single most important insight
2. **Positive Signals** - What's working well (with quotes)
3. **Areas for Improvement** - What needs attention (with quotes)
4. **Quick Wins** - 2-3 things that could be addressed immediately

## Common Mistakes to Flag

- **Cherry-picking positive feedback** - Confirmation bias is the enemy of good analysis
- **Treating all feedback equally** - A review from a verified purchaser weighs more than an anonymous comment
- **Ignoring silent customers** - The feedback you receive is biased toward extremes; most customers never write reviews
- **Analysis paralysis** - Perfect analysis of imperfect data is still valuable; don't wait for complete datasets
- **Forgetting context** - A spike in negative reviews after a price increase is expected, not alarming