Why GA4 setup matters more than most teams realise
Google Analytics 4 replaced Universal Analytics in July 2023, and the transition has been painful for most marketing teams. GA4 is a fundamentally different measurement system: event-based rather than session-based, with a different data model, different attribution logic, and different reporting interfaces. Teams that simply accepted the auto-migration from UA to GA4 are typically working with poorly configured data that cannot support meaningful marketing decisions.
The most common GA4 configuration problems we encounter: events that were not properly migrated from UA goals, conversions that are not correctly defined, cross-domain tracking that breaks attribution across brand properties, and audience definitions that exclude significant user segments. Each of these problems silently corrupts the data that marketing decisions are based on.
Where GA4 setups go wrong
Events and conversions: the configuration foundation
GA4 measures everything as events. Unlike Universal Analytics, where page views, goals, and events were separate constructs, GA4 treats all user interactions as events with parameters. This architecture is more flexible but requires deliberate configuration to produce useful data.
Start by auditing which GA4 events are automatically collected (page_view, session_start, first_visit, user_engagement, scroll, click, file_download, video events) and which require custom implementation. Then define your conversion events (the actions that have genuine business value) and mark them as conversions in GA4. Avoid the common mistake of marking too many events as conversions, which inflates conversion counts and makes attribution analysis unreliable.
Attribution settings: the most consequential configuration choice
GA4's default attribution model is data-driven attribution, which uses machine learning to distribute conversion credit across touchpoints based on observed patterns (Google Analytics Help). Google's documentation lists no minimum conversion volume for it in GA4. With few conversions the model has less data to learn from, so low-volume accounts should compare its results with the paid and organic last click model before acting on them.
The attribution window (how far back GA4 looks for touchpoints to credit) defaults to 30 days for acquisition key events (first_visit, first_open) and 90 days for all other key events (Google Analytics Help). These defaults are appropriate for most B2C businesses but may need adjustment for longer B2B sales cycles.
Custom reports and explorations
GA4's standard reports are designed for general audiences and rarely provide the specific cut of data that marketing teams need for optimisation decisions. The Explore section allows construction of custom analyses: funnel explorations to identify where users drop out of conversion flows, path explorations to understand common user paths, segment overlap reports to understand audience composition.
Magic Pulpo builds custom GA4 Explore templates for every retainer client at the start of an engagement. The templates standardise the data views used for weekly performance reviews, so reporting stays consistent and comparable over time.
For help with your GA4 configuration or analytics setup, speak to the Magic Pulpo analytics team.
Magic Pulpo
Agencia de crecimiento × startup studio, Budapest
