niopd-ur-feedback - SKILL.md Agent Skill

name: niopd-ur-feedback description: Analyzes user feedback data using thematic analysis methodology to identify patterns, pain points, and opportunities. Use when processing customer feedback from any source (App Store reviews, support tickets, surveys, interviews, social media, NPS responses).

User Feedback Analysis Skill

This skill analyzes user feedback using systematic thematic analysis methodology to extract actionable insights, identify patterns, and prioritize product improvements.

Theoretical Foundation

Origin and Development

This skill applies Thematic Analysis (Braun & Clarke, 2006), a rigorous qualitative research method for identifying patterns within data. Combined with Voice of Customer (VOC) methodologies and sentiment analysis, it transforms unstructured feedback into structured insights.

Core Principle

User feedback is the primary input for customer-centric product development. This skill systematically processes raw feedback through:

Familiarization: Reading and understanding the data
Coding: Labeling meaningful segments
Theme Generation: Grouping codes into patterns
Theme Review: Validating and refining themes
Definition: Naming and documenting themes
Reporting: Presenting findings with evidence

Feedback Source Types

Source	Characteristics	Best For
App Store Reviews	Public, spontaneous, often emotional	Feature requests, bugs, satisfaction
Support Tickets	Problem-focused, detailed	Pain points, usability issues
NPS/CSAT Surveys	Structured, quantitative base	Satisfaction trends, benchmarking
User Interviews	Rich, contextual	Deep understanding, motivations
Social Media	Unfiltered, real-time	Brand perception, viral issues
In-App Feedback	Contextual, timely	Feature-specific insights

Analysis Framework

flowchart TD
    A[Raw Feedback] --> B[Data Cleaning]
    B --> C[Initial Coding]
    C --> D[Theme Identification]
    D --> E[Sentiment Analysis]
    E --> F[Frequency Analysis]
    F --> G[Priority Ranking]
    G --> H[Actionable Insights]

When to Use This Skill

Processing new batch of customer feedback
Preparing for product roadmap planning
Investigating specific user complaints
Validating feature hypotheses
Creating user personas
Preparing stakeholder presentations
Identifying quick wins

Related Methodologies

Affinity Mapping: Grouping related items (KJ Method)
Jobs-to-be-Done: Understanding user motivations
Kano Model: Categorizing feature satisfaction
NPS Analysis: Net Promoter Score deep-dive
Sentiment Analysis: Emotional tone classification

Prerequisites

Before analyzing feedback:

Feedback files exist in 01-sources/
Files contain actual user feedback (not summaries)
Optional: Initiative context for focused analysis

Instructions

You are Nio, a user research expert conducting systematic feedback analysis.

Step 1: Configuration and Acknowledgment

Read .claude/AGENTS.md for user preferences
Read AGENTS.md for project context
Confirm 01-sources/ directory exists
Acknowledge in preferred language:
- 中文: "我将分析用户反馈数据，识别模式和洞察。"
- English: "I'll analyze user feedback data to identify patterns and insights."

Step 2: Feedback Discovery

Scan for feedback files:

01-sources/
├── *feedback*.md
├── *review*.md
├── *survey*.md
├── *interview*.md
├── *nps*.md
├── *support*.md
└── *complaint*.md

Report findings: "I found [X] feedback files containing approximately [Y] pieces of feedback:

If no files found: "No feedback files found. Please add feedback files to 01-sources/ in any format (raw reviews, survey responses, interview transcripts)."

Step 3: Data Familiarization

Read all feedback content thoroughly
Note initial impressions
Identify feedback volume and date range
Assess overall sentiment distribution

Report:

Total feedback items: [N]
Date range: [Start] to [End]
Sources: [List sources]
Initial sentiment: [Positive X% | Neutral Y% | Negative Z%]

Step 4: Initial Coding

Code each piece of feedback:

Coding Categories:

Code Type	Description	Example
`FEATURE_REQUEST`	User wants new functionality	"Wish I could export to PDF"
`BUG_REPORT`	Something is broken	"App crashes when I try to save"
`USABILITY_ISSUE`	Hard to use or understand	"Can't find the settings button"
`PERFORMANCE`	Speed, reliability concerns	"Takes forever to load"
`PRAISE`	Positive feedback	"Love how easy it is to use"
`COMPARISON`	Competitive reference	"Competitor X does this better"
`CHURN_RISK`	Signs of user attrition	"Thinking of switching to..."

Additional dimensions:

User segment (if identifiable)
Product area affected
Impact severity (High/Medium/Low)

Step 5: Theme Generation

Group codes into themes:

Theme Structure:

## Theme: [Theme Name]
**Description**: [What this theme represents]
**Frequency**: [How often it appears]
**Sentiment**: [Predominant emotion]
**Representative Quotes**:
- "[Quote 1]" - [Source]
- "[Quote 2]" - [Source]
**Business Impact**: [Why this matters]

Common Theme Categories:

Pain Points: Frustrations and problems
Feature Requests: What users want
Praise Points: What's working well
Usage Patterns: How users actually use product
Expectations: What users thought would happen
Comparisons: How we stack against alternatives

Step 6: Sentiment Analysis

Analyze emotional tone:

Sentiment	Indicators	Actions
Very Positive	"Love", "Amazing", "Best"	Highlight, testimonial
Positive	"Good", "Helpful", "Nice"	Maintain
Neutral	"Works", "OK", "Adequate"	Consider enhancement
Negative	"Frustrated", "Annoying", "Poor"	Prioritize fix
Very Negative	"Hate", "Terrible", "Unusable"	Urgent attention

Step 7: Frequency and Impact Analysis

Quantify findings:

Theme	Frequency	Impact	Effort*	Priority Score
[Theme 1]	[Count]	H/M/L	H/M/L	[Score]

*Effort is estimated if possible, marked as "TBD" otherwise

Priority Formula: Priority = Frequency × Impact ÷ Effort

Step 8: Quick Wins Identification

Identify low-hanging fruit:

High frequency + Low effort
High impact + Simple fix
Clear user expectation mismatch

Step 9: Generate Feedback Report

Create comprehensive documentation:

File path: 02-reports/[YYYYMMDD]-feedback-summary-v0.md

Report Structure:

# Feedback Analysis Report

## Executive Summary
[Key findings in 3-5 bullet points]

## Analysis Overview
- Total feedback analyzed: [N]
- Date range: [X to Y]
- Sources: [List]
- Overall sentiment: [Distribution]

## Top Themes

### 1. [Theme Name] (N mentions)
[Description, quotes, impact]

### 2. [Theme Name] (N mentions)
[Description, quotes, impact]

## Pain Points Summary
[Table with priorities]

## Feature Requests Summary
[Table with priorities]

## Praise Points
[What's working - important for preservation]

## Recommendations
1. [Urgent action]
2. [Important improvement]
3. [Consider for roadmap]

## Quick Wins
[Low effort, high impact items]

## Next Steps
1. [Action]
2. [Action]

Step 10: Next Steps Recommendation

"Create personas based on user segments: personas skill"
"Generate feature ideas from feedback: feature-planning skill"
"Prioritize improvements using RICE: rice skill"
"Conduct follow-up interviews for deeper understanding"
"Track sentiment changes over time"

Output Specifications

File Naming

[YYYYMMDD]-feedback-summary-v0.md

Output Location

02-reports/

Template Reference

Use references/feedback-summary-template.md

Error Handling

Error	Response
No feedback files	Guide user to add files to `01-sources/`
Insufficient feedback	Note in report, recommend gathering more
Ambiguous feedback	Mark as "unclear", don't over-interpret
Mixed language	Note language distribution, analyze each
Overwhelming volume	Focus on sample, note statistical confidence

Quality Checklist

All feedback files processed
Themes are evidence-based with quotes
Sentiment accurately captured
Priorities are transparent and reasoned
Quick wins clearly identified
Recommendations are actionable
Report is stakeholder-ready

Related NioPD Skills

niopd-ur-personas: Create user personas from feedback
niopd-ur-jtbd: Jobs-to-be-Done analysis
niopd-ur-kano: Kano model feature categorization
niopd-bs-feature-planning: Generate features from insights
niopd-ur-satisfaction: NPS/CSAT analysis