customer-feedback-analysis
Analyze NPS, CSAT, and qualitative customer feedback to extract themes, identify trends, and generate actionable insight reports. Use when the user requests customer feedback analysis or provides relevant inputs for this workflow.
git clone --depth 1 https://github.com/seb1n/awesome-ai-agent-skills /tmp/customer-feedback-analysis && cp -r /tmp/customer-feedback-analysis/customer-success/customer-feedback-analysis ~/.claude/skills/customer-feedback-analysisSKILL.md
# Customer Feedback Analysis Transform raw customer feedback from NPS surveys, CSAT responses, support interactions, and app store reviews into structured insights. This skill extracts recurring themes from open-text responses, calculates quantitative score distributions, identifies emerging trends over time, and produces reports that connect customer sentiment to specific product areas and business outcomes. ## Workflow 1. **Collect feedback data** — Aggregate feedback from all available sources: NPS survey responses (score + open text), CSAT ratings from support interactions, in-app feedback widgets, app store reviews, social media mentions, G2/Capterra reviews, and sales call notes. Tag each response with metadata: date, customer segment, plan tier, account tenure, and source channel. Ensure consistent schema across all sources. 2. **Clean and normalize** — Deduplicate responses from the same customer across channels. Standardize rating scales (convert 1-5 CSAT to 1-10 for cross-comparison). Strip PII from open-text responses. Handle multilingual responses by detecting language and translating to English while preserving the original. Remove bot/spam responses using pattern detection (identical text, suspicious timing, single-word noise). 3. **Extract themes from open-text responses** — Apply topic modeling to cluster open-text feedback into coherent themes. Common theme categories include: product reliability, ease of use, specific feature feedback, pricing/value perception, support quality, onboarding experience, and competitive comparison. Assign each response to one or more themes with a confidence score. Pull representative verbatim quotes for each theme. 4. **Calculate quantitative scores** — Compute aggregate metrics: NPS (% Promoters minus % Detractors), CSAT average, and theme frequency distribution. Break scores down by segment (plan tier, industry, account size, tenure) to identify which cohorts are most and least satisfied. Calculate statistical significance for segment differences to avoid acting on noise. 5. **Identify trends** — Compare current period metrics against previous periods (month-over-month, quarter-over-quarter). Flag themes with significant volume changes (up or down 20%+ from baseline). Detect emerging themes that appear for the first time or cross a frequency threshold. Correlate sentiment shifts with product releases, pricing changes, or market events. 6. **Generate insight report** — Produce a structured report with: executive summary (3-5 key takeaways), quantitative scorecard, theme breakdown with representative quotes, trend analysis, segment comparison, and recommended actions. Each recommendation should be tied to a specific theme and prioritized by frequency and business impact. ## Usage Provide feedback data (CSV, JSON, or describe the dataset) and specify the analysis scope. The agent returns a structured insight report. ``` Analyze our Q4 NPS survey results (850 responses). Break down scores by plan tier, extract the top themes from detractor comments, and identify any new themes that weren't present in Q3. ``` ## Examples ### Example 1: NPS survey analysis **Input:** 850 NPS survey responses from Q4, segmented by plan tier. **Output:** **Executive Summary:** - Overall NPS: +32 (up from +28 in Q3, statistically significant at p<0.05) - Enterprise NPS: +52 | Pro NPS: +31 | Starter NPS: +14 - Top detractor theme: "Reporting is too slow" (38% of detractor comments, up from 22% in Q3) - New emerging theme: "Need better mobile experience" (appeared in 12% of Q4 responses, absent in Q3) - Promoter loyalty driver: "Customer support is exceptional" (45% of promoter comments) **Score Distribution:** | Segment | Promoters (9-10) | Passives (7-8) | Detractors (0-6) | NPS | Responses | |---------|-----------------|----------------|-------------------|-----|-----------| | Enterprise | 68% | 16% | 16% | +52 | 180 | | Pro | 52% | 27% | 21% | +31 | 420 | | Starter | 38% | 38% | 24% | +14 | 250 | | **Overall** | **51%** | **28%** | **21%** | **+32** | **850** | **Theme Breakdown (Detractor Comments, n=179):** | Theme | Frequency | Change vs Q3 | Representative Quotes | |-------|-----------|-------------|----------------------| | Slow reporting | 38% | +16pp | "Dashboards take 20+ seconds to load with large datasets. This is killing our team's productivity." | | Pricing concerns | 24% | -3pp | "The price jump from Pro to Enterprise is too steep. We need the features but can't justify 3x the cost." | | Missing integrations | 18% | -5pp | "Still no native Salesforce integration. We've been asking for over a year." | | Mobile experience | 12% | NEW | "I can't review dashboards on my phone during commute. Competitors have solid mobile apps." | | Complex setup | 8% | -8pp | "Initial configuration took our team 3 weeks. Onboarding docs are outdated." | **Recommended Actions:** 1. **P0 — Reporting performance:** Invest in query optimization and caching. 38% of detractors cite this, and it's growing fast. Estimated NPS impact: +5-8 points if resolved. 2. **P1 — Mobile experience:** Commission a competitive analysis of mobile offerings. 12% is a new theme trending upward — get ahead of it before Q1 NPS. 3. **P2 — Pricing tier gap:** Introduce a "Pro Plus" tier between Pro and Enterprise to capture accounts that need select enterprise features without the full price tag. ### Example 2: App store review sentiment analysis **Input:** 1,200 app store reviews from the past 6 months (iOS and Android combined). **Output:** **Overall Sentiment:** 3.8★ average (iOS: 4.1★, Android: 3.4★) **Sentiment by Star Rating:** - 5★ (34%): "Love the design," "Fast and reliable," "Best in category" - 4★ (22%): "Great but needs offline mode," "Almost perfect" - 3★ (18%): "Decent but buggy on Android," "OK for basic use" - 2★ (14%): "Crashes frequently," "Too expensive for what it offers" - 1★ (12%): "Lost my data," "Customer support unresponsive," "App doesn't open"
Design reproducible evaluations for AI agents with representative task sets, explicit rubrics, appropriate graders, baselines, regression gates, and failure analysis. Use when defining agent quality, comparing prompts or models, validating a release, measuring tool-use reliability, investigating regressions, or deciding whether an agent is ready for production.
Design privacy-aware observability for AI agents using traces, spans, structured events, metrics, cost attribution, dashboards, alerts, and investigation workflows. Use when instrumenting an agent, debugging intermittent tool or model failures, defining service-level objectives, analyzing latency or spend, auditing agent decisions, or preparing production monitoring.
Design and verify auditable human oversight, approval gates, escalation paths, and safe state transitions for AI agent workflows. Use when deciding which agent actions require review, adding approve/reject or dual-control flows, preventing unauthorized autonomous effects, creating decision records, reducing rubber-stamping, or recovering safely from rejected, expired, or failed actions.
Design, implement, harden, and verify Model Context Protocol (MCP) servers with precise tool contracts, least-privilege authorization, safe transports, structured errors, and interoperability tests. Use when creating a new MCP server, exposing an API or data source through MCP, reviewing an MCP server design, adding or revising MCP tools, or preparing an MCP server for production.
Design and operate bounded multi-agent workflows with task decomposition, dependency graphs, ownership, handoff contracts, shared-state controls, approvals, recovery, and synthesis. Use when a task contains genuinely independent workstreams, specialized roles, parallel research or implementation, reviewer-worker loops, or coordination problems that one agent should not execute sequentially.
Design and validate model-facing tool definitions with clear names, action-oriented descriptions, bounded JSON Schema parameters, explicit side effects, safe defaults, idempotency, errors, and realistic tests. Use when creating function-calling tools, MCP tools, agent actions, structured tool inputs, or when a model selects the wrong tool, invents arguments, or causes unsafe side effects.
Plan, execute, document, and retest authorized security assessments of AI agents and multi-agent workflows using safe adversarial cases, synthetic identities, canaries, and evidence-based findings. Use when defining red-team rules of engagement, assessing prompt injection or excessive agency, testing tool and identity boundaries, evaluating memory or cross-agent attacks, scoring a campaign, or verifying remediation in an approved environment.
Threat-model and harden AI agents, RAG systems, assistants, and tool-using workflows against direct, indirect, stored, cross-agent, and multimodal prompt injection. Use when reviewing an agent architecture, isolating untrusted content, constraining tools and egress, protecting secrets, adding injection-focused tests, investigating a suspected injection incident, or documenting residual prompt-injection risk.