Context Slice Analysis Types
Based on detected columns, the skill can perform:
| Data Type | Analysis | Key Outputs |
|---|---|---|
| Feedback text | Theme analysis | Top themes, sentiment, quotes |
| Usage metrics | Adoption analysis | Trends, segments, risk signals |
| Experiment data | A/B interpretation | Significance, lift, recommendations |
| Segment data | Comparison analysis | Differences, correlations |
General Output Principles
See Data Analytics Guidelines for complete analysis principles. Key points for product data:
- Lead with insights — Key findings first, methodology second
- Quantify claims — "23% of feedback mentions onboarding" not "many users"
- Include evidence — Quotes, data points, examples
- Connect to roadmap — Link findings to product decisions (ship/cut, prioritize, iterate)
Data Quality Checks
Before analysis, verify:
- Sample size — Enough data for meaningful conclusions?
- Time range — Representative period or anomalous?
- Segment coverage — All relevant segments represented?
- Missing data — Significant gaps that affect interpretation?
Always note limitations in the output.
Visualization Guidelines
When presenting data patterns:
- Trends — Use time-series descriptions or ASCII charts for trajectories
- Distributions — Describe shape (normal, skewed, bimodal)
- Comparisons — Use tables for segment-by-segment analysis
- Proportions — Use percentages with raw counts for context
Connecting to Strategy
Analysis should connect to product decisions:
- Feedback themes → Feature prioritization, roadmap input
- Adoption patterns → Onboarding improvements, churn prevention
- A/B results → Ship/no-ship decisions, iteration ideas
- Segment differences → Targeting, personalization, pricing
End every analysis with "So what?" — What should the PM do with this information?