Designing Event Data for Product Analytics
Design event contracts that keep product metrics interpretable as identifiers, client clocks, delivery retries, and application versions change.
-
-
Conversion Window
Set the elapsed-time limit for a funnel, define entry and boundary rules, and avoid treating unfinished observation windows as failed conversions.
-
Transactional Outbox
Commit a business change with an outbox record, relay the event afterward, and handle duplicate deliveries without assuming an atomic broker send.
-
Schema Evolution and Compatibility
Distinguish backward and forward compatibility, test supported readers and writers, and keep structural changes separate from changes in meaning.
-
Identity Stitching
Link observations across identifiers with explicit evidence and time scope, handle shared devices, and distinguish accounts from individual people.
-
Event Deduplication
Recognize repeated deliveries with stable event IDs, detect conflicting payloads, and keep delivery deduplication separate from business uniqueness.
-
Event Envelope
Separate shared event metadata from business properties, define identifier and timestamp scopes, and distinguish supplied fields from trusted evidence.
-
Tracking Plan
Define event triggers, properties, units, owners, and consumers in a tracking plan, then test whether deployed instrumentation follows it.