Google Analytics

Google Analytics is the most widely used web analytics service in the world, installed on over 55 million websites. The current version, GA4 (Google Analytics 4), replaced Universal Analytics in July 2023, representing the biggest change in Google Analytics history. GA4 moved from a session-based, pageview-centric model to an event-based model where every user interaction — page views, clicks, scrolls, form submissions, video plays — is tracked as an event. This fundamental shift better reflects how users interact with modern websites and apps but required every GA user to re-learn the platform.

Event-Based Data Model

In GA4, everything is an event. A page view is an event. A scroll is an event. A purchase is an event. Each event can have parameters that provide context: the page URL, the scroll depth percentage, the transaction value. This unified model eliminates the artificial distinction between pageviews, events, and goals that existed in Universal Analytics. You define custom events for any interaction that matters to your business: button clicks, form submissions, video completions, file downloads. Enhanced Measurement automatically tracks common events (scrolls, outbound clicks, site search, video engagement, file downloads) without any custom code — just toggle them on in settings.

Explorations and Reporting

GA4's reporting is split into two areas: pre-built Reports and custom Explorations. Reports provide a dashboard-like view of key metrics: user acquisition, engagement, monetization, and retention. They're good for quick overviews but less customizable than Universal Analytics reports. Explorations are GA4's power tool — free-form analysis, funnel exploration, path exploration, segment overlap, and cohort analysis. Funnel exploration lets you define multi-step conversion paths and see where users drop off. Path exploration visualizes the journeys users take through your site. These advanced analysis tools are genuinely powerful for understanding user behavior, but they require analytical skill to use effectively.

Audiences and Predictive Metrics

GA4 uses machine learning to generate predictive metrics: purchase probability (likelihood a user will purchase in the next 7 days), churn probability (likelihood a user won't return), and predicted revenue. These predictions power Predictive Audiences — segments of users likely to convert or churn — that can be exported to Google Ads for targeted campaigns. For example, you can create a Google Ads remarketing audience of users GA4 predicts will purchase soon, or suppress ads for users likely to buy anyway. This integration between analytics and advertising is Google's strategic moat — no competing analytics platform can feed audience segments directly into Google Ads with the same depth.

BigQuery Integration

GA4 offers free BigQuery export, which sends raw event-level data to Google's cloud data warehouse. This is transformative for data teams: instead of being limited to GA4's interface and sampling, you can run SQL queries against every single event from every user. BigQuery export enables custom attribution models, advanced cohort analysis, data blending with CRM or product data, and retention calculations that GA4's UI can't perform. The free export (available on all GA4 properties, not just GA360) generates approximately 10GB of data per million monthly events and qualifies for BigQuery's free tier for small-to-medium sites.

Privacy and Consent

GA4 was designed with privacy regulations in mind. Consent Mode lets GA4 adjust data collection based on user consent: if a user declines cookies, GA4 collects anonymized data and uses machine learning to model the behavior of non-consenting users. IP anonymization is on by default. Data retention can be set to 2 or 14 months for user-level data. Server-side tagging via Google Tag Manager reduces client-side data exposure. Despite these features, GA4 remains controversial in Europe — several EU data protection authorities have ruled Google Analytics non-compliant with GDPR because data is transferred to US servers. Many European companies are migrating to Matomo, Plausible, or Fathom for GDPR compliance.

GA4 vs Universal Analytics

The transition from Universal Analytics to GA4 frustrated millions of users. GA4's interface is less intuitive, standard reports are harder to find, and many features that were simple in Universal Analytics (like bounce rate, which GA4 replaced with engagement rate) changed conceptually. The learning curve is substantial even for experienced analytics users. However, GA4's event-based model is objectively more flexible, the BigQuery export is a massive upgrade, and predictive audiences provide capabilities Universal Analytics never had. GA4 is a better analytics platform — it's just a harder one to learn.