Sentiment Analysis
Turn messy reviews, surveys, and feedback exports into segment-level sentiment scores and JTBD themes before you prioritize roadmap work.
Overview
Sentiment Analysis is an agent skill most often used in Grow (also Idea, Validate) that synthesizes large-scale user feedback into segments, sentiment scores, and JTBD-style satisfaction insights.
Install
npx skills add https://github.com/phuryn/pm-skills --skill sentiment-analysisWhat is this skill?
- Stepwise pipeline: ingest sources, name ≥3 user segments, theme pain and praise per segment
- Assigns per-segment sentiment scores on a -1 to +1 satisfaction scale
- Reads CSV, PDF, survey, review, and social-listening inputs directly in the agent session
- Maps recurring themes to JTBD-style insights for product improvement bets
- Designed for large-scale feedback batches, not one-off anecdotal quotes
- At least 3 distinct user segments identified in the workflow
- Sentiment scores use a -1 to +1 scale per segment
Adoption & trust: 1.2k installs on skills.sh; 12.3k GitHub stars; 2/3 security scanners passed (skills.sh audits).
What problem does it solve?
You have piles of reviews and survey rows but no clear picture of which user segments are delighted versus frustrated or what to fix first.
Who is it for?
Solo builders with exported reviews, NPS comments, or multi-file feedback who need a structured research pass before roadmap or messaging changes.
Skip if: Teams that only need a one-line summary of ten comments, formal statistical NLP pipelines, or decisions when you have no raw feedback files to analyze.
When should I use this skill?
Analyzing user feedback at scale, running sentiment on reviews or surveys, or identifying satisfaction patterns across segments.
What do I get? / Deliverables
You get segment-level sentiment scores, themed pain and praise, and prioritized product-improvement angles grounded in the feedback you supplied.
- Segment inventory with themed pain and praise
- Per-segment sentiment scores (-1 to +1)
- Actionable product-improvement insights tied to JTBD themes
Recommended Skills
Journey fit
Spans multiple journey phases - primary shelf plus alternate fits below.
Canonical shelf is Grow because the skill’s output is satisfaction measurement and feedback synthesis—the work you do once users are talking about the product. Analytics is the best fit for scaled qualitative synthesis, sentiment scoring per segment, and pattern detection across CSVs, PDFs, and survey dumps.
Where it fits
Cluster early interview notes and forum quotes into segments before you commit to a niche.
Score beta-tester sentiment by persona so you cut features that only one angry segment cares about.
Refresh quarterly review dumps to see which segments slipped negative after a pricing change.
Link churn-survey themes to onboarding gaps for lifecycle email or in-app fixes.
How it compares
Use instead of hand-rolling pivot tables in Sheets when you want JTBD-linked themes and scored segments in one agent workflow.
Common Questions / FAQ
Who is sentiment-analysis for?
Indie and solo product owners, PM-minded builders, and small teams who ship SaaS or content products and need repeatable voice-of-customer synthesis without a dedicated researcher.
When should I use sentiment-analysis?
Use it during Grow analytics on reviews and churn reasons, during Idea audience research on early qualitative signals, and during Validate scope when survey or beta feedback should constrain what you build next.
Is sentiment-analysis safe to install?
It is designed to read feedback you provide in-session; review the Security Audits panel on this Prism page and avoid pasting secrets or PII you are not allowed to process.
SKILL.md
READMESKILL.md - Sentiment Analysis
# Sentiment Analysis ## Purpose Analyze large-scale user feedback data to identify market segments, measure satisfaction, and uncover product improvement opportunities. This skill synthesizes feedback into actionable insights organized by user segment, sentiment, and impact. ## Instructions You are an expert user researcher and feedback analyst specializing in qualitative data synthesis and sentiment analysis at scale. ### Input Your task is to analyze user feedback data for **$ARGUMENTS** and identify market segments with associated sentiment insights. If the user provides CSV files, PDFs, survey responses, review data, social listening reports, or other feedback sources, read and analyze them directly. Extract patterns, themes, and sentiment signals from the data. ### Analysis Steps (Think Step by Step) 1. **Data Ingestion**: Read all feedback sources and create a working inventory 2. **Segment Identification**: Identify at least 3 distinct user segments or personas from the feedback 3. **Thematic Analysis**: Extract recurring themes, pain points, and positive feedback per segment 4. **Sentiment Scoring**: Assign sentiment scores (-1 to +1) for overall satisfaction per segment 5. **Impact Assessment**: Prioritize insights by frequency, severity, and business impact 6. **Synthesis**: Create segment profiles with consolidated insights ### Output Structure For each identified segment: **Segment Profile** - Name/identifier and common characteristics - User count or proportion in feedback dataset - Primary use case or context **Jobs-to-be-Done** - Core job this segment is trying to accomplish - Associated desired outcomes **Sentiment Score & Satisfaction Level** - Overall sentiment score (-1 to +1) - Key satisfaction drivers and detractors - Net Promoter Score (NPS) proxy if applicable **Top Positive Feedback Themes** - What this segment loves about $ARGUMENTS - Key strengths from user perspective - Examples of successful use cases **Top Pain Points & Criticism** - Most frequent complaints or frustrations - Unmet needs or missing features - Friction points in user journey - Direct quotes from feedback when available **Product-Segment Fit Assessment** - How well $ARGUMENTS serves this segment's needs - Potential to improve fit through product changes - Risk of churn or dissatisfaction **Actionable Recommendations** - 2-3 highest-impact improvements per segment - Quick wins vs. strategic initiatives - Segments to prioritize or de-prioritize ## Best Practices - Ground all findings in actual user feedback; cite sources - Identify both majority and minority perspectives within segments - Distinguish between feature requests and fundamental pain points - Consider context and constraints users face - Flag segments with small sample sizes or uncertain sentiment - Look for cross-segment patterns and universal pain points - Provide balanced view of product strengths and weaknesses --- ### Further Reading - [Market Research: Advanced Techniques](https://www.productcompass.pm/p/market-research-advanced-techniques) - [User Interviews: The Ultimate Guide to Research Interviews](https://www.productcompass.pm/p/interviewing-customers-the-ultimate)