Skip to content

Collect: Data Collection

Overview of the Collect phase in the CORE framework

Data collection is the foundation of any analytics strategy. This phase covers best practices for collecting high-quality data that powers your analytics and decision-making.

The Collect phase focuses on gathering reliable, well-structured data from all relevant sources. A well-designed collection strategy ensures consistency, reliability, and privacy compliance across your entire analytics system.

A well-designed schema ensures consistency and reliability across your data collection:

  • Event naming conventions: Use clear, descriptive names (e.g., purchase_completed not evt_123)
  • Parameter standardization: Define required and optional parameters upfront
  • Versioning strategy: Plan for schema evolution as your needs grow

Server-side tagging provides several advantages:

  • Reduced client-side load: Move processing to your servers
  • Enhanced privacy: Better control over data collection
  • Improved reliability: Less dependency on client-side JavaScript
  • Event schema defined and documented
  • Tracking plan created and approved
  • Server-side tagging infrastructure deployed
  • Consent management integrated
  • Data quality validation in place
  • Privacy and compliance requirements met
  • Tracking Plan: Document outlining all events, parameters, and data collection requirements
  • Event Schema: Structured definition of events and their properties
  • Tagging Implementation: Server-side or client-side tagging setup
  • Consent Management: Integration with consent management platform
  • Data Validation Rules: Automated checks to ensure data quality
  • Over-collection: Collecting too much data without clear purpose leads to noise and privacy concerns
  • Inconsistent naming: Poor event naming makes analysis difficult and error-prone
  • Ignoring privacy: Failing to implement consent management can lead to compliance issues
  • Client-side only: Relying solely on client-side tracking creates reliability and privacy risks
  • No validation: Missing data quality checks allows bad data to pollute your analytics