Token导航 LogoToken导航TokenDH.com
研究检索需要联网github未标认证来源可访问许可证需确认审计通过

data-and-funnel-analytics数据和漏斗分析

Agent Skill

用于辅助数据整理、表格处理、CSV/Excel 分析、指标计算和图表准备。它适合让 Agent 清洗字段、汇总数据、发现异常、生成统计口径或把分析结果转成可读说明。使用时需要确认数据来源、字段含义和时间范围,避免把样本数据当全量事实;涉及敏感数据、导出文件或批量写回时,应先确认权限和脱敏边界。

总安装

4,627

周安装

189

GitHub Stars

42

下载量

1,497
CodexClaudeCursorGemini CLI

安装说明

本站只整理中文说明和来源信息,不托管安装包,也不代用户安装。

GitHub

来源数

2

许可证

unknown

最后核验

2026-05-01

来源状态

来源可访问

安装方式

通过对话安装

复制提示词发给支持本地命令或 Skills 的 AI 助手,先确认命令和权限,再让它执行。

请帮我安装这个 Agent Skill:data-and-funnel-analytics(数据和漏斗分析)
来源仓库:https://github.com/manojbajaj95/claude-gtm-plugin
仓库路径:skills/data-and-funnel-analytics
安装命令:
npx skills add https://github.com/manojbajaj95/claude-gtm-plugin --skill data-and-funnel-analytics
安装前请先检查当前环境是否支持对应 CLI,并向我确认将要执行的命令、安装目录、联网范围和文件读写权限;确认后再执行。

命令行安装

复制命令到本机终端执行。该命令会通过 npx skills 从第三方来源获取 Skill;本站只展示命令,不托管安装包,也不自动执行。

skills.shnpx skills
npx skills add https://github.com/manojbajaj95/claude-gtm-plugin --skill data-and-funnel-analytics

简介

端到端追踪设置与漏斗转化分析,支撑产品 engagement 与 ROI 测算决策。

  • 事件命名采用 object_action 小写蛇形规则,强调具体性与完成态表达。
  • 每个埋点须对应明确业务动作,避免为数据而追踪的低效实践。
  • 支持转化路径验证、流失节点诊断与归因模型搭建等高级分析场景。
  • 需结合真实业务语义设计属性结构,防止过度整理或命名模糊化。

SKILL.md

Data & Funnel Analytics

End-to-end analytics: set up tracking, interpret data, analyze funnels, measure product engagement, validate conversion paths, and calculate ROI.

Principle: Track for decisions, not data — every event should inform an action.


Analytics Tracking

Event Naming Convention

Format: object_action in lowercase snake_case.

signup_completed | cta_hero_clicked | checkout_started | onboarding_step_completed

Rules: Specific over vague (cta_hero_clicked not button_clicked), past tense for completed actions, context in properties not event name.

Tracking Plan

CategoryEventKey Properties
Marketingpage_viewpage_title, page_location, referrer
cta_clickedbutton_text, location, page
form_submittedform_type, page
signup_completedmethod, plan
Productonboarding_step_completedstep_number, step_name
feature_usedfeature_name, context
trial_startedplan, source
purchase_completedplan, value, currency
E-commerceproduct_viewedproduct_id, category, price
product_added_to_cartproduct_id, price, quantity
checkout_startedcart_value, items_count

Standard Properties

  • User context: user_id, user_type (free/paid/admin), plan_type
  • Attribution: source, medium, campaign, content, term (UTM params)
  • Page: page_title, page_location, content_group
  • PII hygiene: Never send email, name, or phone as event properties. Use hashed user IDs only.

GA4 Implementation

// gtag.js custom event
gtag('event', 'signup_completed', {
  'method': 'email',
  'plan': 'free',
  'user_id': userId
});

// GTM dataLayer
dataLayer.push({
  'event': 'signup_completed',
  'method': 'email',
  'plan': 'free'
});

Enhanced Measurement (enable in GA4): page_view, scroll, outbound_click, site_search, video_engagement, file_download.

Conversions: Admin → Events → Toggle "Mark as conversion." Counting: once per session (form submit) or every time (purchase).

UTM Parameters

Convention: utm_source={channel}&utm_medium={cpc|email|organic|social}&utm_campaign={id}&utm_content={variant}&utm_term={keyword}

  • Apply to ALL paid and email links
  • Never use on internal links (breaks session attribution)
  • Lowercase, hyphens not spaces
  • Document in a UTM tracking sheet

Privacy & Compliance

  • GDPR/CCPA: Implement consent management, block GA4 until consent granted
  • GA4 data retention: 14 months max (Admin → Data Settings)
  • IP anonymization enabled

Analytics Interpretation

GA4 Benchmarks

MetricGoodWarningPoorAction When Poor
Avg Time on Page>3 min1–3 min<1 minImprove content depth
Bounce Rate<40%40–70%>70%Add internal links, improve intro
Engagement Rate>60%30–60%<30%Review content quality
Scroll Depth>75%50–75%<50%Add visual breaks
Pages/Session>2.51.5–2.5<1.5Improve internal linking

Google Search Console Benchmarks

MetricGoodWarningPoorAction When Poor
CTR>5%2–5%<2%Improve title/meta description
Avg Position1–34–10>10Strengthen content, build links
ImpressionsGrowingStableDecliningRefresh content

Traffic Quality Matrix

                    High Engagement
                          │
           ┌──────────────┼──────────────┐
           │  HIDDEN GEM  │   STAR       │
           │  Low traffic  │   High traffic│
           │  → Promote   │   → Maintain  │
Low ───────┼──────────────┼──────────────┼─── High
Traffic    │  UNDERPERFORM│   LEAKY      │   Traffic
           │  Low traffic  │   High traffic│
           │  → Rework    │   → Optimize  │
           └──────────────┼──────────────┘
                          │
                    Low Engagement

Anomaly Detection

MetricSignificant ChangeAlert Level
Traffic±30% WoWHIGH
CTR±1pp WoWMEDIUM
Position±5 positionsHIGH
Bounce Rate±10pp WoWMEDIUM

Product Analytics

North Star Metric

The ONE metric that represents customer value:

CompanyNorth Star
SlackWeekly Active Users
AirbnbNights Booked
SpotifyTime Listening
ShopifyGMV

Criteria: Represents customer value, correlates with revenue, measurable frequently, rallies the team.

Key Metrics by Stage

StageMetrics
AcquisitionTraffic sources, CPC, visitor → signup rate
ActivationSignup → first core action, time to value, onboarding completion
RetentionDAU/MAU (stickiness), D1/D7/D30 retention, churn rate
RevenueMRR/ARR, ARPU, LTV, LTV:CAC ratio
ReferralViral coefficient, referral signups, NPS

Retention Benchmarks

TimeframeGoodBad
D160–80%<40%
D740–60%<10%
D3030–50%<2%

Good = flattening curve. Bad = steep drop-off.

Dashboard Design

  • Executive: North Star Metric (big number), revenue (MRR/ARR), key trends
  • Product: Active users, feature usage, retention cohorts, funnels
  • Marketing: Traffic sources, conversion rates, CPA, ROI by channel

Funnel Analysis

Core Workflow

  1. Load and merge user journey data
  2. Define funnel steps and calculate step-by-step conversion rates
  3. Segment by user attributes (device, cohort, plan)
  4. Visualize bottlenecks
  5. Generate optimization recommendations

Common Funnel Types

FunnelSteps
E-commercePromotion → Search → Product View → Add to Cart → Purchase
SaaS SignupLanding Page → Sign Up → Email Verify → Onboarding Complete
ContentArticle View → Comment → Share → Subscribe

Analysis Patterns

  • Bottleneck identification — Steps with highest drop-off rates
  • Segment comparison — Conversion across user groups
  • Temporal analysis — Conversion over time
  • A/B testing — Compare funnel variations

See examples/ for Python implementations with Plotly visualizations.


Funnel Validation (DotCom Secrets)

Score existing funnels against Russell Brunson's framework: Hook → Story → Offer.

Scoring Dimensions

DimensionWeightWhat It Measures
Hook Strength2xStops the scroll, grabs attention
Story Connection1.5xCreates emotional connection and belief
Offer Clarity2xClear, compelling, irresistible
Value Ladder Fit1xFits the ascension path
Traffic Match1.5xMatched to traffic temperature
Conversion Path1xNext step obvious and frictionless

Rating Scale

ScoreVerdict
85–100Conversion Machine — Ready to scale
70–84Strong Funnel — Fix weak points, then scale
55–69Leaky Funnel — Fix before scaling traffic
40–54Broken Funnel — Rebuild key components
0–39Non-Functional — Start over

Traffic Temperature

TemperatureThey KnowAppropriate Funnel
ColdNothing about youLead funnel, value-first content
WarmProblem + your solutionTripwire, webinar, challenge
HotReady to buySales page, order form, call booking

For complete scoring criteria and examples, see references/full-guide.md.


ROI Analysis

Core Metrics

ROI: (Net Profit / Total Investment) × 100%

  • ✅ INVEST: ROI > 100% (realistic case)
  • ⚠️ REVIEW: ROI 50–100%
  • ❌ REJECT: ROI < 50%

Break-Even: Investment / Monthly Net Profit

  • ✅ INVEST: Break-even < 50% of realistic target
  • ❌ REJECT: Break-even > 70%

Payback Period: Investment / Monthly Net Profit

  • ✅ INVEST: < 12 months
  • ⚠️ REVIEW: 12–24 months
  • ❌ REJECT: > 24 months

3-Scenario Analysis

Always model Best / Realistic / Worst:

CaseAssumptionsRevenueProfitROIAssessment
WorstPessimisticRisk level
RealisticExpectedTarget
BestOptimisticUpside

Decision rule: If worst-case ROI ≥ 0%, investment is low-risk.

Executive Summary Template

[Investment] achieves [ROI%] ROI at [conversion/growth rate].
Break-even occurs at [threshold], with payback in [months].
Investment is [recommended/not recommended] because [reason].

For detailed formulas (NPV, LTV, CAC, sensitivity analysis), see references/roi-reference.md.


Validation & QA

Before Launch

  • Events fire in GA4 DebugView
  • Properties have expected values
  • No duplicate events
  • Conversions marked correctly
  • UTM parameters captured on landing

Ongoing

  • Weekly: Check for sudden drops in key events (>20% change = investigate)
  • Monthly: Audit for new pages/features without tracking
  • Quarterly: Full tracking plan review — remove stale events, add missing ones

Tools

CategoryTools
Event TrackingMixpanel, Amplitude, PostHog (open-source)
Session RecordingFullStory, LogRocket, Hotjar
A/B TestingOptimizely, VWO
Web AnalyticsGA4, Google Search Console
Tag ManagementGoogle Tag Manager

Related Skills

  • ab-test-setup — A/B test measurement and setup
  • seo-and-aeo-strategy — Measuring SEO/AEO performance
  • conversion-rate-optimization — Optimizing conversion after funnel analysis
  • executive-dashboard-generator — Building dashboards from analytics data

适合场景

01

用户想查找某类 Agent Skill 时

02

需要根据任务场景推荐可安装能力包时

03

需要对比不同来源的安装命令和来源信息时

能力概览

能力 1

按任务关键词查找相关 Skills

能力 2

展示可复制的安装命令

能力 3

保留来源站点、仓库和原始说明,方便继续核验

能力 4

展示第三方安全扫描或审计结果

安装后应在对应宿主中按原始 README 的触发条件使用;具体调用方式请以来源页面和 README 为准。

平台分布

Codex

34.02%
按下载量换算509

Claude

32.69%
按下载量换算489

Cursor

17%
按下载量换算254

Gemini CLI

10.07%
按下载量换算151

安全审计

Gen Agent Trust Hub

通过

Socket

通过

Snyk

通过

权限和风险

需要联网

该 Skill 可能需要联网访问来源站点、仓库或外部 API;具体网络访问范围需要结合源码和 README 复核。

安装前确认

本站仅展示第三方公开信息,不托管安装包,不提供自动安装或运行环境。安装前应自行审查源码、依赖和命令行为。当前只有一个来源,正式发布前建议补源仓库或其他目录站核验。

来源信息

继续浏览同类 Skills