Implement Product Analytics Tracking Systems

Product Analytics Setup guides the design of tracking systems, meaningful metrics, and measurement frameworks for product decisions. Events are organized around user actions and business outcomes, with consistent object_action names and a page-to-section-to-element taxonomy. Contextual properties support segmentation rather than recording isolated technical events without business meaning.

The event-schema example combines product identifiers, category, revenue, and currency with user identity and tier, session and referral context, platform, app version, and an ISO timestamp. A web wrapper merges global context with event-specific properties and illustrates dispatch to several analytics platforms. Critical events receive client-side and server-side tracking, schemas, development validation, and ongoing data-quality monitoring.

The metric framework includes acquisition and activation queries alongside engagement, retention, revenue, and product-health measures. Measures include new users by source, trial conversions, onboarding activation, DAU/MAU, session frequency, feature adoption, retention cohorts, churn scores, ARPU, LTV, time to value, stickiness, and satisfaction. Cohort-analysis code groups unique users by cohort and period, then divides by cohort size to produce retention rates.

Experiment patterns record exposure with experiment name, variant, and user identity, then attach active experiment context to conversion events. A consent-aware wrapper suppresses collection without consent, distinguishes essential from full tracking, and removes email and phone fields from the properties it sends. These are collection-control examples rather than a complete compliance assessment.

Use this skill before product instrumentation or when aligning existing events with decisions, funnels, retention, and experiments. The development sequence starts with a tracking specification, proceeds through a consistent wrapper and event-verification checklist, and uses feature flags for gradual rollout.

Avoid designing the taxonomy only around implementation details. The skill emphasizes meaningful user actions, validated schemas, and quality dashboards so that reported measures have a consistent interpretation. Platform examples need real identifiers, configuration, and application-specific events before they become a working deployment.

Inputs include product actions, business outcomes, segmentation properties, user and session context, experiment assignments, consent levels, and reporting requirements. Outputs include tracking specifications, event schemas, wrapper patterns, metric queries, retention tables, and dashboard plans. Configure GA4 custom dimensions with this structure:

gtag('config', 'GA_MEASUREMENT_ID', {
  custom_map: {
    'custom_parameter_1': 'user_tier',
    'custom_parameter_2': 'feature_flag'
  }
});

Operational guidance covers asynchronous collection, batching, offline queues, and sampling for high-volume events. Reporting spans executive and operational dashboards, anomaly alerts, regular cohort and funnel reports, and self-service access for product teams.

The examples name Amplitude, Mixpanel, and Google Analytics 4. Amplitude setup covers default tracking and user properties; GA4 configuration addresses purchases and custom dimensions. Python cohort analysis uses pandas, with NumPy also imported in the source.