Google Analytics 4: A Complete Guide to Understanding Your Data

On 1 July 2023 Google switched off Universal Analytics and GA4 became the only standard. Since then millions of websites have been collecting data with GA4 — but most read it wrongly, configure it halfway or ignore it altogether. The result is that important decisions about where to invest in marketing are made without a reliable data foundation.

This guide takes you from understanding the data model of Google Analytics 4 to the practical setup of key events, the reports professionals actually use and the integrations with Google Ads and Search Console. Whether you're starting from scratch or trying to make sense of the data you already collect, you'll find everything you need here.

Sources: W3Techs (2026); Google Analytics Help

GA4 vs Universal Analytics: a paradigm shift

To understand GA4 you need to understand why it's so different from Universal Analytics. It isn't an incremental update: it's a complete rewrite based on a different data model.

Universal Analytics was based on sessions and page views: every visit was a session, and within the session page views and goals were counted. It was a model designed for the static websites of the 2000s, before mobile apps, video and complex JavaScript interactions became the norm.

GA4 is based on events: every interaction a user has with the site or app is an event — viewing a page, clicking a button, playing a video, scrolling to 90% of the page, making a purchase. This model is much more flexible and lets you track any user behaviour, not just the pages visited.

GA4's structure: properties, streams and events

Before configuring GA4, it helps to understand its hierarchical structure. A GA4 property collects data from one or more streams. A data stream is a data source: a website, an iOS app, an Android app. A single GA4 property can have several streams — for example a company's website and mobile app, sharing the same property to get a unified view of user behaviour.

Every interaction recorded in GA4 is an event. Events have a name (e.g. page_view, click, purchase) and can have additional parameters that provide context (e.g. page_title, link_url, value). This flexible structure lets you capture any type of interaction at the level of detail you need.

Basic GA4 setup: the essential steps

Creating a GA4 property is only the first step. Without the right configuration, the data collected is incomplete or misleading. These are the mandatory steps after creating the property.

The GA4 reports that matter: where to find useful information

GA4's interface is disorienting at first compared with Universal Analytics. Reports are organised differently and some familiar data seems to have disappeared. In reality it's all there — in different places or under different names.

Acquisition reports

The Acquisition section shows where users come from: channel (Organic Search, Paid Search, Direct, Referral, Social, Email), specific source/medium (google/organic, facebook/cpc) and campaign (for UTM-tagged traffic). The User acquisition report shows the first channel the user interacted with; the Traffic acquisition report shows the specific session.

Engagement reports

The Engagement section has replaced UA's Behaviour report. It shows the most visited pages and screens (with sessions, views and average engagement time), events collected with counts, and conversions with value. Average engagement time is GA4's equivalent of UA's time on page — but more accurate, because it only measures the time during which the browser tab was actually active.

Monetisation reports

For e-commerce, the Monetisation section shows an overview of purchases with total revenue, best-selling products with quantities and revenue, and the purchase journey showing where users drop out of the funnel. It requires implementing the purchase event with the correct item parameters.

Explore: the advanced report builder

The Explore section is GA4's advanced analysis module. It lets you build custom reports with any combination of dimensions and metrics, create funnel explorations, path explorations (how users move between pages) and cohort explorations (the behaviour of groups of users over time). It's GA4's most powerful tool for anyone who wants to go beyond the standard reports.

Debugging and verifying tracking: how to know whether GA4 is working

One of the most common problems with GA4 is tracking that seems to work but actually records incomplete or duplicate data. Before trusting the reports, verify the tracking.

Common GA4 configuration mistakes

GA4 for e-commerce: advanced purchase tracking

For an e-commerce store, GA4 offers far more detailed purchase tracking than Universal Analytics, but it requires a more precise implementation. GA4's enhanced e-commerce tracking records every step of the purchase funnel.

The GA4 e-commerce funnel events to implement, in order: view_item_list (viewing a product list), view_item (viewing a product page), add_to_cart (adding to cart), begin_checkout (starting checkout), add_payment_info (entering payment details), purchase (completing the purchase, with parameters: transaction_id, value, currency, items).

With all these events implemented, the Purchase journey report shows exactly where users drop out of the funnel: how many reach the product page, how many add to cart, how many start checkout and how many complete the purchase. This funnel is the basis for identifying your e-commerce optimisation priorities.

On WooCommerce, the official WooCommerce Google Analytics Integration plugin or third-party plugins such as Pixel Manager implement the GA4 e-commerce events automatically, with no custom code required.

GA4 and privacy: GDPR, consent and data

GA4 collects data about users — and in Europe that requires GDPR compliance. Collecting analytics data through GA4 requires the user's explicit consent under most interpretations of the GDPR adopted by European data protection authorities, including the Italian Garante.

The practical solution is to implement a Consent Management Platform (CMP) with a compliant cookie banner — Cookiebot, Iubenda, Usercentrics — and configure GA4 in Consent Mode v2. With Consent Mode v2, GA4 collects partial, aggregated data even for users who don't accept, letting you keep a view of your traffic without violating the privacy of those who refuse cookies.

Without Consent Mode, every user who refuses cookies is invisible to GA4. On sites where the cookie refusal rate exceeds 30–40% (common in Italy), GA4 data without Consent Mode severely underestimates real traffic.

GA4 and Looker Studio: custom, shareable dashboards

GA4's native reports are powerful but not always presentable to clients or non-technical stakeholders. Looker Studio (formerly Google Data Studio) is Google's free tool for creating custom visual dashboards with GA4 data.

With Looker Studio you can combine GA4 data with Google Ads, Search Console and Google Sheets data in a single dashboard, create custom charts showing exactly the metrics that matter to your business, update the data automatically and share it via a link with anyone, without them needing access to GA4. It's the standard tool for monthly performance reports sent to clients.

GA4 audiences: advanced segmentation for remarketing and analysis

One of GA4's most underused features is its advanced audience system. Audiences in GA4 are segments of users defined by behavioural conditions — and they can be used both for internal analysis and for remarketing in Google Ads campaigns.

Creating an audience in GA4 is simple: go to Admin → Audiences → New audience. You can define criteria such as: users who visited a specific product page but didn't complete the purchase, users who spent more than 3 minutes on the site, users who had at least 2 sessions in the last month, users who made a purchase worth more than €100.

Audiences are automatically shared with Google Ads if the properties are linked. This lets you create highly specific remarketing campaigns — showing an ad only to people who abandoned their cart in the last 7 days, or a special offer to people who visited the pricing page more than 3 times without converting.

For internal analysis, audiences can be applied as filters to GA4 reports to understand how specific segments of users behave compared with the average. How do users arriving from the newsletter behave compared with those from organic search? Which channel brings users with the highest lifetime value? Audience analysis answers these questions.

GA4 predictive metrics: artificial intelligence serving analysis

GA4 includes machine-learning-based metrics that no previous version of Analytics had: predictive metrics. These metrics use AI models trained on your historical data to predict users' future behaviour.

GA4's three main predictive metrics are: purchase probability (the probability that a user active in the last 28 days will make a purchase in the next 7 days), churn probability (the probability that a user active in the previous 7 days won't be active in the next 7) and predicted revenue (an estimate of the revenue expected over the next 28 days from an active user).

These predictive metrics need a minimum volume of data to work: at least 1,000 purchase transactions in the last 28 days for purchase probability. For most small and medium-sized sites they won't be available straight away, but they represent the direction modern analytics is moving in — from describing the past to predicting the future.

An immediate practical application: create an audience based on "likely purchasers" in GA4 and use it in Google Ads campaigns with specific offers for people who are already close to converting. This kind of predictive targeting beats any traditional demographic segmentation because it's based on real behavioural signals.

GA4 for multilingual and international sites: configuration and geographic reports

For sites with versions in different languages or with traffic from several countries, GA4 offers geographic and language dimensions built into the reports. In the Demographics section of the reports you'll find the user's country, region, city and browser language.

For e-commerce stores with prices in different currencies or sites with localised versions, it's important to configure the currency parameters correctly in purchase events: the currency parameter in the purchase event tells GA4 which currency the transaction value is expressed in. GA4 can automatically convert to a reference currency if the property is configured with the correct currency.

Geographic acquisition reports let you compare performance by country: which market generates the most revenue, which has the highest conversion rate and which has the lowest cost per acquisition when linked to your Ads campaigns. For anyone with ambitions of international expansion, this data shows where to invest.

BigQuery export from GA4: advanced data analysis for those who want more

GA4's free plan includes an integration with Google BigQuery that few users take advantage of: the ability to export all raw event data to BigQuery, Google's cloud data warehouse. This feature, available for free since 2023, turns GA4 from an analytics tool into a data source for advanced analysis.

With GA4 data in BigQuery you can do things that are impossible in the standard interface: analyse the complete user journey at individual session level, join with CRM or e-commerce data for lifetime value analysis, build custom attribution models and create Looker Studio dashboards that update almost in real time without the sampling limits of the GA4 interface.

You activate it under Admin → Product links → BigQuery links. It requires a Google Cloud project (BigQuery's free tier includes 1 TB of free queries a month, enough for small and medium-sized sites). Data is exported daily into tables structured by date. For anyone with SQL skills or working with a data analyst, this integration turns GA4 into an enterprise-level source of insight at no significant extra cost.

Conclusion: GA4 is a tool, the data is the value

A well-configured GA4 is the difference between making marketing decisions based on gut feeling and making them based on data. However sophisticated your SEO or advertising strategy, if you don't measure what works and what doesn't, you're operating blind.

Start with the basic setup — the tag installed, internal traffic filtered, key events configured, data retention at 14 months — and build from there. Every month that passes without these elements is a month of data lost that you can't recover.

If you need help configuring GA4 for your website or e-commerce store, explore the services available or contact me for an assessment of your current setup.

Data doesn't lie — but a wrong configuration does. Take the time to get it right once.