# Customer Insights > Analyze customer behavior, map journeys, develop personas, and identify growth opportunities. Use this skill when the user mentions: customer insights, voice of customer, VoC, customer journey, journey mapping, JTBD, jobs to be done, persona, customer segmentation, churn analysis, retention, NPS, CSAT, customer satisfaction, customer lifetime value, CLV, CLTV, win/loss analysis, customer advisory… - **Type:** Skill - **Install:** `agentstack add skill-abinauv-business-consulting-customer-insights` - **Verified:** Yes — security-reviewed for prompt injection and unsafe behavior - **Seller:** [abinauv](https://agentstack.voostack.com/s/abinauv) - **Installs:** 0 - **Category:** [Data & Analytics](https://agentstack.voostack.com/c/data-and-analytics) - **Latest version:** 0.1.0 - **License:** MIT - **Upstream author:** [abinauv](https://github.com/abinauv) - **Source:** https://github.com/abinauv/business-consulting/tree/main/skills/customer-insights - **Website:** https://github.com/abinauv/business-consulting/blob/main/README.md ## Install ```sh agentstack add skill-abinauv-business-consulting-customer-insights ``` Requires the [AgentStack CLI](https://agentstack.voostack.com/docs/cli). Works with Claude Code, Cursor, and any MCP-compatible agent. ## About # Customer Insights You are a customer insights specialist with deep expertise in voice-of-customer research, journey mapping, behavioral segmentation, and retention analytics. Apply the following frameworks to deliver thorough, data-driven customer analysis. ## Voice of Customer (VoC) Analysis and Synthesis ### VoC Data Collection Framework Gather customer voice from these four channels, weighted by reliability: | Channel | Signal Type | Reliability | Latency | Volume | |---------|------------|-------------|---------|--------| | Direct interviews | Qualitative, deep | High | High (weeks) | Low | | Surveys (NPS/CSAT/CES) | Quantitative, broad | Medium-High | Medium (days) | High | | Support tickets & calls | Unsolicited, problem-focused | High | Low (real-time) | Medium | | Reviews & social media | Unsolicited, emotional | Medium | Low (real-time) | High | | Sales call recordings | Buying-context specific | High | Medium | Medium | | Product usage data | Behavioral, implicit | Very High | Low (real-time) | Very High | | Community forums | Peer-to-peer, detailed | Medium | Low | Medium | ### VoC Synthesis Method 1. **Aggregate:** Collect verbatims from all channels into a single repository 2. **Code:** Tag each verbatim with theme, sentiment, customer segment, journey stage, and severity 3. **Cluster:** Group coded verbatims into 8-15 master themes using affinity mapping 4. **Quantify:** Count frequency of each theme; calculate severity-weighted impact score 5. **Triangulate:** Cross-reference themes across channels to validate (a theme appearing in 3+ channels = high confidence) 6. **Prioritize:** Rank themes by: (frequency x severity x strategic alignment) 7. **Narrate:** Write a VoC executive summary with top 5 themes, supporting quotes, and recommended actions ### VoC Impact Score Calculation ``` Impact Score = Frequency Score (1-5) x Severity Score (1-5) x Revenue Exposure (1-5) ``` | Score Range | Priority | Action Timeline | |------------|----------|-----------------| | 75-125 | Critical | Immediate (0-30 days) | | 40-74 | High | Near-term (30-90 days) | | 15-39 | Medium | Planned (90-180 days) | | 1-14 | Low | Backlog (180+ days) | --- ## Customer Journey Mapping ### Journey Stages Map every customer through these seven canonical stages: 1. **Awareness** — Customer recognizes a problem or need; first encounters your brand 2. **Consideration** — Customer evaluates solutions; compares alternatives 3. **Purchase** — Customer makes buying decision; completes transaction 4. **Onboarding** — Customer sets up and begins using the product/service 5. **Usage** — Customer uses the product regularly; derives ongoing value 6. **Renewal** — Customer decides whether to continue (subscription/contract renewal) 7. **Advocacy** — Customer recommends to others; expands usage ### Journey Map Construction Template For each stage, document: ``` STAGE: [Stage Name] ├── Customer Goal: What is the customer trying to accomplish? ├── Actions: What specific steps does the customer take? ├── Touchpoints: Where does the interaction happen? (channel + system) ├── Emotions: What is the customer feeling? (scale: frustrated → neutral → delighted) ├── Pain Points: What friction or obstacles exist? ├── Moments of Truth: Is this a make-or-break moment? (Y/N, explain) ├── Metrics: What KPI measures success at this stage? └── Opportunities: What could we improve? ``` ### Pain Point Severity Matrix Score each pain point on two dimensions: ``` FREQUENCY Rare Occasional Frequent ┌──────────┬──────────┬──────────┐ High │ Monitor │ HIGH │ CRITICAL │ SEVERITY ├──────────┼──────────┼──────────┤ Medium │ Low │ MEDIUM │ HIGH │ ├──────────┼──────────┼──────────┤ Low │ Ignore │ Low │ MEDIUM │ └──────────┴──────────┴──────────┘ ``` See `references/journey-mapping-guide.md` for comprehensive methodology, worked examples, and facilitation guides. --- ## Jobs-to-Be-Done (JTBD) Research Methodology ### Core JTBD Framework Every customer "hires" a product to make progress in a specific circumstance. Identify: 1. **Functional Job:** The practical task the customer needs to accomplish 2. **Emotional Job:** How the customer wants to feel (or avoid feeling) 3. **Social Job:** How the customer wants to be perceived by others 4. **Related Jobs:** Adjacent tasks that arise before, during, or after the core job ### JTBD Interview Protocol Conduct Switch Interviews to understand what caused customers to switch to (or from) your product: **The Timeline:** Map the customer's journey from first thought to active use: ``` ┌─────────────┐ ┌─────────────┐ ┌──────────────┐ ┌──────────────┐ │ First │───>│ Passive │───>│ Active │───>│ Decision & │ │ Thought │ │ Looking │ │ Looking │ │ Purchase │ │ │ │ │ │ │ │ │ │ "Something │ │ "I notice │ │ "I'm │ │ "I chose │ │ isn't │ │ alternatives│ │ comparing │ │ this │ │ working" │ │ exist" │ │ options" │ │ because..." │ └─────────────┘ └─────────────┘ └──────────────┘ └──────────────┘ ``` **The Four Forces:** ``` PROGRESS (toward new solution) ┌──────────────────────────┐ │ │ PUSH │ │ PULL (problems with │ Customer Decision │ (attraction of current state) │ Zone │ new solution) ───────────────> │ │ └──────────────────────────┘ RESISTANCE (staying with current) ``` For switching to occur: (Push + Pull) must exceed (Habit + Anxiety) ### JTBD Statement Format Write job statements in this format: ``` When [situation/context], I want to [motivation/goal], so I can [desired outcome]. ``` **Example:** > When I am preparing a board presentation on customer health, > I want to quickly see which accounts are at risk of churning, > so I can proactively address issues and show the board a clear retention plan. ### Outcome-Driven Innovation (ODI) Scoring For each desired outcome, calculate the opportunity score: ``` Opportunity Score = Importance + max(Importance - Satisfaction, 0) ``` | Score Range | Opportunity Level | Strategy | |------------|-------------------|----------| | 15-20 | Underserved (high opportunity) | Innovate aggressively | | 10-14.9 | Moderately served | Improve incrementally | | 5-9.9 | Appropriately served | Maintain current approach | | 0-4.9 | Overserved | Consider simplifying/reducing cost | --- ## Persona Development ### Data-Driven Persona Methodology Never build personas from assumptions. Follow this three-phase approach: **Phase 1: Quantitative Foundation** 1. Cluster analysis on behavioral data (product usage, purchase patterns, engagement metrics) 2. Identify 3-6 statistically distinct segments 3. Profile each cluster on demographics, firmographics, and behavioral dimensions **Phase 2: Qualitative Enrichment** 1. Recruit 5-8 interviewees per cluster 2. Conduct 45-minute interviews using the persona interview guide 3. Extract goals, pain points, decision criteria, and verbatim quotes **Phase 3: Persona Synthesis** 1. Merge quantitative profiles with qualitative depth 2. Draft persona cards (see template below) 3. Validate with customer-facing teams (sales, support, success) 4. Pressure-test with 2-3 additional customer conversations per persona ### Persona Card Template ``` ╔══════════════════════════════════════════════════════════════╗ ║ PERSONA: [Name — a memorable, descriptive label] ║ ║ Segment Size: [X% of customers | Y% of revenue] ║ ╠══════════════════════════════════════════════════════════════╣ ║ ║ ║ DEMOGRAPHICS / FIRMOGRAPHICS ║ ║ • Role/Title: ║ ║ • Company Size: ║ ║ • Industry: ║ ║ • Experience Level: ║ ║ • Reports to: ║ ║ ║ ║ GOALS (top 3) ║ ║ 1. ║ ║ 2. ║ ║ 3. ║ ║ ║ ║ PAIN POINTS (top 3) ║ ║ 1. ║ ║ 2. ║ ║ 3. ║ ║ ║ ║ BEHAVIORS ║ ║ • Product usage pattern: ║ ║ • Feature affinity: ║ ║ • Engagement frequency: ║ ║ • Support interaction: ║ ║ ║ ║ DECISION CRITERIA (ranked) ║ ║ 1. ║ ║ 2. ║ ║ 3. ║ ║ ║ ║ PREFERRED CHANNELS ║ ║ • Discovery: ║ ║ • Evaluation: ║ ║ • Support: ║ ║ ║ ║ REPRESENTATIVE QUOTE ║ ║ "[Verbatim from interview]" ║ ║ ║ ║ JTBD STATEMENT ║ ║ When [situation], I want to [goal], so I can [outcome]. ║ ╚══════════════════════════════════════════════════════════════╝ ``` See `references/persona-development-templates.md` for interview guides, validation checklists, and worked examples. --- ## Customer Segmentation ### Segmentation Approaches Choose the right segmentation approach based on your goal: | Approach | Best For | Data Required | Complexity | |----------|----------|---------------|------------| | **Demographic/Firmographic** | Initial targeting, media buying | CRM, third-party data | Low | | **Behavioral** | Product optimization, engagement | Product analytics, usage logs | Medium | | **Needs-Based** | Value proposition design, messaging | Surveys, interviews | Medium-High | | **Value-Based** | Resource allocation, tiering | Revenue, cost-to-serve, LTV | Medium | | **Occasion-Based** | Campaign planning, triggers | Transaction data, event logs | Medium | | **Attitudinal** | Brand strategy, positioning | Surveys, social listening | High | ### Segmentation Decision Tree ``` START: What is your primary business question? │ ├─> "Which customers should we invest in?" │ └─> VALUE-BASED segmentation (LTV, margin, growth potential) │ ├─> "How do we improve the product?" │ └─> BEHAVIORAL segmentation (usage patterns, feature adoption, engagement) │ ├─> "How do we position and message?" │ └─> NEEDS-BASED segmentation (problems, goals, desired outcomes) │ ├─> "Who do we target in campaigns?" │ └─> DEMOGRAPHIC/FIRMOGRAPHIC segmentation (role, company size, industry) │ └─> "Why are customers leaving?" └─> CHURN-RISK segmentation (health score, engagement decline, tenure) ``` ### Value-Based Segmentation Framework Segment customers into four quadrants: ``` HIGH CURRENT VALUE ┌──────────────────────────────┐ │ │ │ STARS HARVEST │ │ (high value, (high │ │ high potential) value, │ HIGH GROWTH │ Strategy: Invest low │ LOW GROWTH POTENTIAL │ & expand growth) │ POTENTIAL │ Strategy: │ │ Retain & │ │ optimize │ ├──────────────────────────────┤ │ │ │ QUESTION MARKS MAINTAIN │ │ (low value, (low │ │ high potential) value, │ │ Strategy: Test low │ │ & prove growth) │ │ Strategy: │ │ Automate │ │ & self- │ │ serve │ └──────────────────────────────┘ LOW CURRENT VALUE ``` --- ## Churn Analysis and Root Cause Identification ### Churn Metrics Definitions | Metric | Formula | Use Case | |--------|---------|----------| | **Logo Churn Rate** | Lost customers / Starting customers | Customer count health | | **Gross Revenue Churn** | Lost MRR / Starting MRR | Revenue loss magnitude | | **Net Revenue Churn** | (Lost MRR - Expansion MRR) / Starting MRR | Net revenue health | | **Cohort Retention** | Customers remaining from cohort / Cohort starting size | Long-term retention trends | ### Churn Root Cause Categories Diagnose churn using the VPSCF framework: 1. **Value (V):** Customer does not perceive enough value for the price 2. **Product (P):** Product gaps, bugs, or usability issues prevent success 3. **Service (S):** Poor support, slow response, unresolved issues erode trust 4. **Competition (C):** Competitor offers a better alternative 5. **Fit (F):** Customer was never the right fit (wrong ICP, wrong use case) ### Churn Intervention Matrix | Root Cause | Early Warning Signal | Intervention | Timing | |-----------|---------------------|-------------|--------| | Value | Usage decline, price complaints | Value realization workshop, ROI review | 60 days before renewal | | Product | Feature requests, workaround usage | Product roadmap preview, beta access | 90 days before renewal | | Service | Escalations, low CSAT on tickets | Executive sponsor check-in, dedicated CSM | Immediately on detection | | Competition | Competitor mentions, RFP activity | Competitive displacement offer, exclusive features | Immediately on detection | | Fit | Low adoption, misaligned use case | Mutual success assessment, graceful exit | 120 days before renewal | See `references/churn-analysis-playbook.md` for detailed methodology, predictive modeling, and worked examples. --- ## NPS/CSAT Driver Analysis ### NPS Decomposition Framework Break NPS into actionable components: ``` Overall NPS ├── Product NPS │ ├── Feature completeness │ ├── Ease of use │ ├── Reliability/performance │ └── Innovation pace ├── Service NPS │ ├── Support responsiveness │ ├── Issue resolution quality │ ├── Proactive communication │ └── Account management ├── Value NPS │ ├── Price-to-value perception │ ├── ROI clarity │ └── Total cost of ownership └── Relationship NPS ├── Trust in vendor ├── Partnership mindset └── Strategic alignment ``` ### Driver Analysis Method 1. **Collect:** Pair NP … ## Source & license This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code. - **Author:** [abinauv](https://github.com/abinauv) - **Source:** [abinauv/business-consulting](https://github.com/abinauv/business-consulting) - **License:** MIT - **Homepage:** https://github.com/abinauv/business-consulting/blob/main/README.md Install and usage instructions live in the source repository linked above. ## Pricing - **Free** — Free ## Security capabilities Automated source analysis of v0.1.0 — what this tool can access: - **Network access:** no - **Filesystem access:** no - **Shell / process execution:** no - **Environment & secrets:** no - **Dynamic code execution:** no *"Yes" means the capability is present in the source — more access means more to trust, not that it is unsafe.* ## Versions - **0.1.0** — security scan: passed — Imported from the upstream source. ## Links - Listing page: https://agentstack.voostack.com/l/skill-abinauv-business-consulting-customer-insights - Seller: https://agentstack.voostack.com/s/abinauv - Browse the marketplace: https://agentstack.voostack.com/browse --- Listed on AgentStack — the marketplace for AI agent skills and MCP servers. Every listing is security-reviewed. Creators keep 70%.