Google Analytics 4: complete guide

GA4 drops the session-and-pageview model of Universal Analytics in favour of an event-based model. Every interaction — a page view, a click, a download — is recorded as an event with associated parameters. This enables more granular and flexible measurement.

Another major difference is native Google Ads integration and the ability to combine web and app data in a single property. GA4 also incorporates machine learning to fill data gaps caused by privacy and consent restrictions.

GA4 setup starts by creating a property in Google Analytics and connecting a data stream for your website or app. The measurement code can be installed directly or, ideally, through Google Tag Manager for greater control and flexibility.

From day one, configure conversion events, enable Google Signals for demographic data, and link GA4 with Google Ads and Search Console. Also extend data retention to the maximum available (14 months) and exclude internal traffic.

GA4 automatically collects basic events (page_view, scroll, outbound link clicks, file downloads) without additional configuration. But the real value lies in custom events and recommended events you define based on your business model.

To mark an event as a conversion, simply toggle it in the GA4 interface. Every business should identify its key conversions: form submissions, purchases, sign-ups, content downloads. These feed attribution reports and Google Ads campaign optimisation.

GA4 lets you build audiences based on behaviour, demographics and time-based conditions. These audiences sync automatically with Google Ads for remarketing and can trigger events when a user enters or leaves a segment.

Predictive audiences are a GA4-exclusive feature: using machine learning, they identify users with a high probability of purchase or churn within the next 7 days. This enables proactive action before losing the user.

GA4 splits its analysis into standard reports and explorations. Standard reports cover acquisition, engagement, monetisation and retention with aggregated data. Explorations enable ad-hoc analysis with free-form tables, funnels, path analysis and cohort breakdowns.

Explorations are GA4’s most powerful tool for analysts. They allow you to cross-reference dimensions and metrics with full flexibility, apply comparative segments and uncover patterns that standard reports miss.

One of GA4’s most significant advantages is free raw data export to BigQuery. This lets you run SQL queries against the full event dataset without the sampling limitations of the interface.

With BigQuery you can build custom attribution models, combine GA4 data with CRM or sales systems, and create advanced dashboards in tools like Looker Studio. It is the foundation for a truly advanced data strategy.

A solid GA4 implementation requires planning. Define a measurement plan before touching the configuration: what business questions you need to answer, which events and parameters cover them, and how they connect to your KPIs.

Maintain consistent naming conventions for events and parameters (snake_case, no spaces), document the entire implementation, and periodically audit that data is being collected correctly. A messy implementation produces useless data.