Implement Product Analytics Tracking Systems

This skill is expert in product analytics setup, specializing in designing comprehensive tracking systems, defining meaningful metrics, and building measurement frameworks that drive product decisions. Its event-driven architecture principles call for designing events around user actions and business outcomes rather than technical implementation, a consistent object_action naming convention, contextual properties enabling segmentation, and a Page-Section-Element event taxonomy. Its data quality foundation covers validating events pre-production, client- and server-side tracking for critical events, event schemas for consistent structure, and automated data-quality monitoring. It covers a full event schema example separating business, user, session, and technical context, a multi-platform ProductAnalytics class that fans a single track() call out to Amplitude, Mixpanel, and Google Analytics simultaneously, AARRR funnel SQL (acquisition metrics by source, activation rate by cohort week), a product KPI framework (engagement via DAU/MAU and feature adoption, retention via day 1/7/30 cohorts and churn scores, revenue via ARPU/LTV, product health via time-to-value and stickiness), concrete Amplitude and GA4 configuration code, Python cohort-table construction for retention analysis, feature-flag/A-B-test experiment tracking, and GDPR-compliant privacy-aware tracking that respects consent level and strips PII fields before sending events.

Use this skill when setting up or auditing product analytics tracking that needs a real event taxonomy, multi-tool integration, and privacy compliance - not a single ad hoc pageview pixel. It defines a five-step development workflow (write a tracking spec before development, use wrapper libraries for consistency, run a QA checklist for event verification, roll out gradually behind feature flags, monitor with data-quality dashboards) and performance guidance for high-volume tracking (event batching, async tracking to avoid blocking the UI, offline event queuing, sampling for high-volume events). Its privacy-first analytics class explicitly gates tracking by consent level (none, essential, full) and sanitizes PII like email and phone before sending - this is built for GDPR-relevant products, not an afterthought bolted on later.

class ProductAnalytics {
  constructor(config) {
    this.config = config;
    this.context = this.getGlobalContext();
  }
  
  track(eventName, properties = {}) {
    const event = {
      event: eventName,
      properties: {
        ...this.context,
        ...properties,
        timestamp: new Date().toISOString()
      }
    };
    
    this.sendToAmplitude(event);
    this.sendToMixpanel(event);
    this.sendToGoogleAnalytics(event);
  }
  
  getGlobalContext() {
    return {
      user_id: this.getUserId(),
      session_id: this.getSessionId(),
      platform: 'web',
      app_version: this.config.version,
      referrer: document.referrer,
      url: window.location.href
    };
  }
}

Given a product's key user actions, the skill produces an event schema definition, a multi-platform tracking wrapper like the one above, Amplitude and GA4 configuration blocks with custom dimensions, AARRR funnel SQL queries, a Python cohort-retention table function, experiment-exposure and conversion tracking code, and a consent-aware, PII-sanitizing tracking class for GDPR compliance.

Product and growth engineering teams setting up or hardening analytics tracking who need a real event taxonomy, multi-tool integration (Amplitude, Mixpanel, GA4), and cohort/funnel measurement rather than scattered pageview tracking. It suits teams that need privacy compliance built into the tracking layer from the start, and that want executive and operational dashboards, automated anomaly detection, and self-service analytics capabilities for product teams as the end goal, not just raw event collection.