jeffvincent avatar

support-data-analyzer

Analyze customer support data (Excel/CSV with PIT, Support Tickets, CSAT) to categorize issues, prio

提供方 jeffvincent|开源

Support Data Analyzer Skill

A Claude Skill for analyzing customer support feedback data and generating prioritized strategic reports with real customer examples.

What This Skill Does

Transforms raw support data (Excel/CSV) into actionable insights by:

  • Categorizing thousands of issues into clear themes
  • Prioritizing by volume, frustration, and revenue impact (P0-P3)
  • Extracting real customer quotes to illustrate pain points
  • Trending recent issues vs. historical baselines
  • Recommending specific actions with timelines and impact estimates

When to Use This Skill

Invoke this skill when you have:

  • Excel/CSV files with support data (PIT, roadblocks, tickets, CSAT)
  • Need to understand top customer pain points
  • Want prioritized recommendations for product/eng teams
  • Need to show trends (60-day vs. all-time)

Quick Start

1. Prepare Your Data

Your Excel file should have these sheets (all optional except Support Data or PIT):

PIT & Roadblocks (strategic customer feedback):

  • Submitted At Date
  • Category (e.g., "Object", "Properties", "Pipelines")
  • Sub Category (e.g., "Associations", "Create Property")
  • Frustration Level (1-5 scale)
  • MRR (CS) and MRR (Sales)
  • Use Case Title and Use Case Body

Support Data (support tickets):

  • Created UTC Date
  • Support Product Area
  • Support roadblock
  • Ticket Name and Content

CSAT (optional - satisfaction surveys):

  • Created At Date
  • Score (1-5)
  • Text (feedback)
  • Event Trigger

Ideas Forum (optional - feature requests):

  • Idea Title
  • Count of Upvotes
  • Portal Gross MRR

2. Invoke the Skill

User: "Analyze this support data and prioritize issues by customer impact.
       Focus on the last 60 days."

[Attach: Data Platform data - Code Orange.xlsx]

3. Review the Output

The skill generates a comprehensive markdown report:

  • Executive Summary: Top 3 critical issues, key metrics, trends
  • P0-P3 Breakdown: Each theme with 5 customer examples
  • Strategic Recommendations: Immediate actions with timelines
  • CSAT Analysis: Correlation with support themes
  • Revenue Protection: MRR at risk by theme

Example Use Cases

Use Case 1: Quarterly Product Planning

"Analyze all our support data and show me the top 10 customer pain points with revenue impact." → Generates all-time analysis with P0-P3 priorities, MRR totals, and Ideas Forum validation

Use Case 2: Sprint Planning (Last 60 Days)

"What are customers complaining about most in the last 60 days? Show trends vs. historical." → Generates filtered analysis with trend arrows, "NEW to top 5" flags, and frustration increases

Use Case 3: Deep Dive on Specific Area

"Can you analyze just the property management issues in detail?" → Generates focused report with sub-theme breakdown and 7-10 examples per sub-issue

Use Case 4: Customer Examples for Leadership

"Create a doc with real customer examples for each major pain point." → Generates separate customer feedback document with detailed quotes and interpretations

Files in This Skill

support-data-analyzer/
├── Skill.md                              # Main skill instructions
├── README.md                             # This file
└── resources/
    ├── PRIORITIZATION_FRAMEWORK.md       # P0-P3 assignment criteria
    ├── REPORT_TEMPLATE.md                # Markdown report structure
    └── EXAMPLES.md                       # Sample analyses with annotations

Configuration Options

Time Filtering

  • "Last 60 days" - Recent trends, worsening issues
  • "Last 90 days" / "Q4 2025" - Quarter-based planning
  • No filter (default) - All-time historical analysis

Focus Areas

  • "Focus on association issues" - Drills into sub-themes
  • "Show me property management issues" - Single category deep-dive

Output Formats

  • Default: Single comprehensive report
  • "Create two files: analysis + customer examples" - Separate docs

Interpreting Results

Priority Levels

P0 (Critical) - Ship stoppers, high revenue at risk

  • Volume >150 (60d) or >1,000 (all-time) OR
  • High frustration + substantial volume OR
  • Revenue >$75K (60d) or >$4M (all-time)

P1 (High) - Significant pain, medium-high volume

  • Volume 100-150 (60d) or 700-1,000 (all-time)
  • Moderate frustration with good volume

P2 (Medium) - Specialized high pain or moderate volume

  • Volume 30-100 (60d) or 200-700 (all-time)
  • High frustration but niche use case

P3 (Lower) - Administrative, low frequency

  • Volume <30 (60d) or <200 (all-time)

Trend Indicators

  • ↑ +14% - Metric increasing (usually bad for frustration)
  • ↓ -4pp - Percentage point decrease (good if frustration)
  • NEW to top 3 - Issue surged recently, investigate urgently
  • ⚠️ - Frustration ≥3.5/5, high pain when encountered

Customer Examples

Each example includes:

  • Quote: Real customer feedback (not synthesized)
  • Frustration: 1-5 scale (if from PIT data)
  • MRR: Revenue at risk (if available)
  • "What this means": Interpretation of business impact

Tips for Best Results

Data Quality

  • ✅ Include frustration levels (1-5) for strategic depth
  • ✅ Include MRR data to prioritize by revenue
  • ✅ Use consistent date formats (YYYY-MM-DD)
  • ✅ Categorize issues (Product Area > Sub Category)

Analysis Scope

  • 🎯 60 days: Best for recent trends and sprint planning
  • 🎯 All-time: Best for big-picture strategy and Ideas Forum validation
  • 🎯 Focused: Best for deep-dive into specific product area

Interpreting Frustration

  • 1-2: Annoyance, low priority
  • 3: Moderate pain, workflow friction
  • 4: High pain, actively seeking workarounds
  • 5: Critical blocker, considering churn

Customization

Adjusting Priority Thresholds

Edit resources/PRIORITIZATION_FRAMEWORK.md to change:

  • Volume thresholds for P0-P3
  • Frustration weight in priority calculation
  • Revenue impact thresholds

Changing Report Structure

Edit resources/REPORT_TEMPLATE.md to:

  • Add/remove sections
  • Change formatting preferences
  • Adjust example count per theme

Dependencies

  • Python 3.8+ (for Excel processing)
  • openpyxl (for .xlsx file reading)

Install dependencies:

pip install openpyxl

Troubleshooting

"Cannot read Excel file"

  • Convert to CSV manually, or
  • Check file is not corrupted
  • Verify file has expected sheet names

"Date filtering returns 0 results"

  • Check date column format (should be YYYY-MM-DD HH:MM:SS)
  • Verify date column name matches expected ("Created UTC Date", "Submitted At Date")

"Missing frustration data"

  • Frustration levels optional, will prioritize by volume only
  • Note: Support tickets typically don't have frustration, only PIT data

"MRR totals seem low"

  • Not all issues have MRR data (support tickets often don't)
  • "Revenue at risk" is from PIT data only, not complete picture

Version History

  • v1.0.0 (2025-10-31): Initial release
    • Supports PIT, Support Tickets, CSAT, Ideas Forum
    • P0-P3 prioritization framework
    • 60-day and all-time analysis modes
    • Trend comparison
    • Customer example extraction

Support

For issues or questions about this skill:

  1. Check resources/EXAMPLES.md for sample analyses
  2. Review resources/PRIORITIZATION_FRAMEWORK.md for priority logic
  3. Verify your data format matches expected columns

License

Internal use only. Do not distribute support data externally.

下载 Skill