GA4 migration: a practical setup and transition guide

Google Analytics 4 is not a new version of Universal Analytics — it is a different measurement model. Sessions give way to events, the data is user- and event-scoped, and the most valuable reporting now lives in BigQuery rather than the UI. Treating the move as a like-for-like swap is the single most common reason migrations produce numbers teams don't trust.

Start from a measurement plan, not the interface

Before touching a tag, write down the decisions the business needs to make and the events that inform them: which campaigns drove revenue, where users drop out of the funnel, what a qualified lead looks like. A short measurement plan keeps your event taxonomy consistent and prevents the sprawl of half-named events that makes GA4 reporting unusable six months later.

Model events and parameters deliberately

• —Use a small, consistent set of events with descriptive names (purchase, generate_lead, view_item) rather than one event per interaction.
• —Put the detail in parameters, and register the ones you'll report on as custom dimensions.
• —Mark revenue and lead events as key events (conversions) so they're usable in Ads and explorations.

Turn on BigQuery export from day one

GA4's free BigQuery export is the feature that makes it worth the migration. It gives you unsampled, row-level data you can join to CRM, cost and offline data — the foundation for attribution, LTV and dashboards that the standard interface can't produce. Enable it before you start collecting in earnest so you have history when you need it.

Run both systems in parallel, then validate

Collect GA4 alongside your existing setup for a few weeks and reconcile the deltas: traffic by channel, key conversions, and ecommerce totals. Discrepancies are expected — the models differ — but they should be explainable. Document the expected gaps so stakeholders aren't surprised when the numbers don't match to the decimal.

Common pitfalls

• —Leaving consent and regional rules unconfigured, which silently drops data.
• —No spam/bot filtering, inflating event counts.
• —Skipping server-side collection, so iOS, Safari and ad-blockers erode conversion data.

Done well, GA4 becomes the trustworthy backbone for marketing decisions. Done as a rushed swap, it becomes another dashboard nobody believes. If you'd like a second pair of eyes, we offer a free audit of your GA4 setup.