# Metrics Frameworks — Reference Data # Canonical models used by all Data Squad agents # Version: 1.0.0 see_think_do_care: name: "See-Think-Do-Care Framework" author: "Avinash Kaushik" description: "Audience-intent framework that maps marketing to customer journey stages" stages: - id: see name: "See" audience: "Largest addressable qualified audience" intent: "No commercial intent" content: "Awareness, inspiration, education" metrics: ["impressions", "reach", "new visitors", "brand recall"] channels: ["display", "social organic", "video", "PR"] - id: think name: "Think" audience: "Audience with some commercial intent" intent: "Considering options" content: "Comparison, reviews, how-to" metrics: ["engagement rate", "time on site", "pages per session", "email signups"] channels: ["search", "content marketing", "retargeting", "email"] - id: do name: "Do" audience: "Audience with strong commercial intent" intent: "Ready to act" content: "Product pages, pricing, trials, demos" metrics: ["conversion rate", "cost per acquisition", "revenue", "ROAS"] channels: ["paid search", "remarketing", "email nurture", "sales"] - id: care name: "Care" audience: "Existing customers (2+ transactions)" intent: "Loyal, repeat behavior" content: "Support, loyalty, upsell, community" metrics: ["NPS", "retention rate", "CLV", "referral rate"] channels: ["email", "community", "support", "loyalty programs"] aarrr_pirate_metrics: name: "AARRR Pirate Metrics" author: "Dave McClure / Sean Ellis" description: "Startup growth funnel framework" stages: - id: acquisition name: "Acquisition" question: "How do users find us?" metrics: ["traffic by channel", "CAC", "signup rate"] - id: activation name: "Activation" question: "Do users have a great first experience?" metrics: ["onboarding completion", "time to value", "activation rate"] - id: retention name: "Retention" question: "Do users come back?" metrics: ["DAU/MAU", "cohort retention", "churn rate"] - id: referral name: "Referral" question: "Do users tell others?" metrics: ["viral coefficient", "NPS", "referral rate"] - id: revenue name: "Revenue" question: "Do users pay?" metrics: ["ARPU", "MRR", "LTV/CAC ratio"] north_star_metric: name: "North Star Metric Framework" author: "Sean Ellis" description: "The single metric that best captures the core value delivered to customers" properties: - "Reflects customer value (not just revenue)" - "Is leadable (team can influence it)" - "Is measurable (can track weekly)" - "Predicts long-term business success" input_dimensions: breadth: "How many users experience the value" depth: "How much value each user gets" frequency: "How often users get value" examples: - product: "Airbnb" nsm: "Nights booked" - product: "Spotify" nsm: "Time spent listening" - product: "Slack" nsm: "Messages sent within a team" - product: "Shopify" nsm: "GMV (Gross Merchandise Volume)" ice_scoring: name: "ICE Scoring Framework" author: "Sean Ellis" description: "Prioritization framework for growth experiments" dimensions: impact: range: "1-10" question: "How much will this move the North Star Metric?" anchors: 1: "Negligible impact" 5: "Moderate impact on a subset of users" 10: "Massive impact on all users" confidence: range: "1-10" question: "How sure are we this will work?" anchors: 1: "Pure guess" 5: "Some supporting data" 10: "Proven in similar context" ease: range: "1-10" question: "How easy is this to implement?" anchors: 1: "Months of work, many dependencies" 5: "1-2 weeks, some dependencies" 10: "Hours, no dependencies" formula: "ICE = (Impact + Confidence + Ease) / 3" bg_nbd_clv_model: name: "BG/NBD Customer Lifetime Value Model" author: "Peter Fader & Bruce Hardie" description: "Probabilistic model for predicting customer lifetime value in non-contractual settings" principles: - "Customers have heterogeneous transaction rates" - "Customers have heterogeneous dropout rates" - "After any transaction, a customer may become permanently inactive" - "Transaction rates and dropout rates vary independently" inputs: recency: "Time since last transaction" frequency: "Number of repeat transactions" monetary: "Average transaction value" tenure: "Time since first transaction" outputs: - "P(alive) — probability customer is still active" - "Expected future transactions" - "Expected CLV over horizon" segmentation: platinum: "Top 5% by CLV" gold: "Next 15% by CLV" silver: "Next 30% by CLV" bronze: "Bottom 50% by CLV" spaces_model: name: "SPACES Model" author: "David Spinks" description: "Framework for defining community business value" dimensions: - id: support name: "Support" description: "Members help each other, reducing support costs" metrics: ["ticket deflection rate", "community resolution rate", "time to resolution"] - id: product name: "Product" description: "Members provide feedback, improving the product" metrics: ["ideas submitted", "beta testers", "feature adoption from feedback"] - id: acquisition name: "Acquisition" description: "Community attracts new customers" metrics: ["community-sourced leads", "referral rate", "SEO from UGC"] - id: contribution name: "Contribution" description: "Members create content or code" metrics: ["UGC volume", "contributor count", "content quality score"] - id: engagement name: "Engagement" description: "Members connect, increasing retention" metrics: ["DAU/MAU", "response rate", "returning visitor rate"] - id: success name: "Success" description: "Members achieve outcomes, reducing churn" metrics: ["goal completion rate", "NPS from community members", "churn diff vs non-members"] ten_laws_customer_success: name: "10 Laws of Customer Success" author: "Nick Mehta" description: "Foundational principles for building a customer success practice" laws: - number: 1 title: "Sell to the Right Customer" principle: "CS starts in sales — wrong-fit customers cannot be saved" - number: 2 title: "The Natural Tendency for Customers is to Churn" principle: "Without proactive engagement, customers drift away" - number: 3 title: "Customers Expect You to Make Them Wildly Successful" principle: "The bar is not satisfaction — it is transformation" - number: 4 title: "Relentlessly Monitor and Manage Customer Health" principle: "Health scores must be multi-dimensional and predictive" - number: 5 title: "You Can No Longer Build Loyalty Through Personal Relationships Alone" principle: "Product and process must scale beyond individual relationships" - number: 6 title: "Product Is Your Only Scalable Differentiator" principle: "Long-term retention is driven by product value" - number: 7 title: "Obsessively Improve Time-to-Value" principle: "The faster customers see value, the higher retention" - number: 8 title: "Deeply Understand Your Customer Metrics" principle: "Data-driven CS outperforms gut-driven CS" - number: 9 title: "Drive Customer Success via Hard Metrics" principle: "NRR, expansion, and health score — not feelings" - number: 10 title: "It Is a Top-Down, Company-Wide Commitment" principle: "CS is not a department — it is a company strategy" health_score: name: "Customer Health Score" description: "Multi-dimensional health assessment for churn prediction" dimensions: - name: "Product Usage" weight: 25 signals: ["login frequency", "feature adoption", "time in app"] - name: "Support Interaction" weight: 15 signals: ["ticket volume", "sentiment", "escalation rate"] - name: "Engagement" weight: 20 signals: ["email opens", "event attendance", "community activity"] - name: "Business Outcome" weight: 20 signals: ["goal achievement", "ROI reported", "success milestones"] - name: "Relationship" weight: 10 signals: ["NPS score", "executive sponsor engagement", "QBR attendance"] - name: "Financial" weight: 10 signals: ["payment timeliness", "contract value trend", "expansion signals"] thresholds: healthy: "0-25" monitor: "26-50" at_risk: "51-75" critical: "76-100" spiky_pov_framework: name: "Spiky Point of View Framework" author: "Wes Kao" description: "Framework for developing a differentiated perspective that builds audience" criteria: specific: "Not a broad truism — a precise, narrow claim" provable: "Backed by data, experience, or results" interesting: "Makes people stop scrolling" contrarian: "Challenges conventional wisdom" your_terrain: "You have more credibility here than most" validation_test: "Would someone knowledgeable argue with this?" content_generation: minimum_pieces: 50 description: "A valid spiky POV should generate at least 50 content pieces" platform_selection: factors: - "Audience presence (where they already are)" - "Content format fit (matches your creation style)" - "Distribution mechanics (organic reach potential)" - "Monetization path (leads, courses, consulting)" rule: "Max 2 platforms simultaneously"