Measurement Data Dictionary

A measurement data dictionary gives analysts, developers and decision-makers a shared language for digital reporting. A useful entry records a field name, plain-English meaning, data type, accepted values, collection trigger, system of record, responsible owner and known caveats. For example, a lead_submit event should distinguish a genuinely completed enquiry from a mere click on the submit button. Without that distinction, different dashboards may report incompatible conversion totals.

For event-based analytics, link dictionary entries to the tracking plan and implementation specifications. Explain whether an event is collected in a browser or server environment, how consent affects it, which identifiers are prohibited and how retries or duplicate sends are handled. Maintain a change history when a site redesign or tag-manager release changes definitions. Share documentation with analysts and developers, not only the person who set up tracking.

A dictionary differs from a measurement plan. The plan starts with business questions and identifies what should be measured; the dictionary precisely documents what each collected field means and how to interpret it. It also differs from a raw export schema, which lists columns but may not explain their business purpose. Documentation improves comparability, but does not by itself prove that data collection is accurate.