Foundation · 2 min
What you need to know
How should ecommerce analytics be migrated?
Rebuild analytics from a versioned measurement plan that defines each event, parameter, identifier and owner. Validate revenue and customer journeys before launch instead of copying tags blindly.
How do I preserve historical analytics?
Keep the old analytics property, exports or warehouse records accessible and annotate the cutover date. Do not rewrite historical events to imitate the new implementation.
Should event names remain the same after migration?
Keep an event name only when its trigger and meaning remain unchanged. Version or document the event when scope, currency, identity, attribution or timing changes.
How do I validate purchase tracking before launch?
Place controlled orders covering products, discounts, tax, shipping, refunds, and payment outcomes. Reconcile transaction ID, value, currency and items across the store, tag debugger and analytics destination.
How do I compare analytics before and after migration?
Use matched periods and segment by device, channel, market, and customer type. Separate real behavior changes from tracking-definition, consent and traffic-mix changes.
Why did analytics suddenly change after migration?
Analytics can change suddenly when event definitions, identifiers, consent behavior, currency, time zone, attribution or deduplication changes. Reconcile known orders end to end before interpreting the trend.
A quick store check
Trace one purchase from browser events to the analytics report and reconcile it with the actual order and refund state.
Foundation · 2 min
Frequently asked questions
Should old and new analytics be merged?
Keep historical data accessible but document definition changes instead of forcing incomparable events into one series.
When can performance comparison begin?
After the new implementation is validated and differences in consent, attribution and event definitions are understood.