Event-Tracking Schema

What It Is

The tracking plan that specifies, before any tool is wired, exactly what the business records:
- Events — meaningful actions (Lead Submitted, Quote Requested, Trial Started), named by one enforced convention.
- Properties — typed parameters on each event (value, channel, plan), with required/optional status.
- Identity keys — the IDs that stitch a user together (anonymous_id, user_id, hashed email).
- Data-quality rules — required fields, valid ranges, rejection criteria.

Why It's Load-Bearing

The schema is a contract enforced at ingestion. Without it, teams invent inconsistent event names and the warehouse becomes unqueryable — and every downstream metric inherits the mess. Clean inputs here are the difference between trustworthy and misleading analytics.

How It Applies to Marketing Factory

The schema is the first foundation of the data layer: it feeds marketing-attribution and every measurement concept, and pairs with identity-resolution (the identity keys) and privacy-resilient-tracking (how events are captured under consent). It is the structural defense against the GIGO agentic-failure-modes risk — agents optimizing on inconsistent events will confidently chase noise. Schema definition and ingestion-time validation are fully agent-ownable.

Referenced from: marketing-data-layer