Google Analytics 4 (GA4) Explained for Builders

GA4 is the current Google Analytics property type. It is built on an event-based data model and supports unified measurement across web and apps. If your marketing stack depends on Google Ads or the wider Google ecosystem, GA4 is still the default choice for many teams. This guide focuses on the parts that matter most for builders: events, server-side measurement, and data retention.

Event-based by default

GA4 replaced session-first reporting with an event-based model. That means every interaction is an event with parameters, which makes it easier to track product behavior and unify reporting across platforms. Google introduced GA4 specifically to unify app and website measurement in a single property.

Web and app in one property

GA4 lets you collect data from both websites and mobile apps in the same property. This is useful if you want cross-platform reporting, shared audiences, or unified attribution across web and app.

Server-side and offline events

GA4 supports the Measurement Protocol for server-to-server event collection. This is useful when you want to send backend events, offline conversions, or data that is more reliable than client-side tracking alone.

Data retention basics

GA4 data retention settings control how long user and event data is retained for analysis. The official data retention settings include 2 months and 14 months, with longer retention options reserved for GA4 360. Retention settings matter for exploratory reports and should be configured early if you need longer analysis windows.

When GA4 is a good fit

You rely on Google Ads or Google Marketing Platform integrations.
You need unified web and app measurement in one property.
You want a widely adopted baseline tool that stakeholders recognize.

Common friction points

GA4 is flexible but can feel complex for day-to-day product insight. Event naming, key event setup, and report configuration matter more than in older analytics models. Many teams keep GA4 for marketing integration and add a simpler product analytics or privacy-first tool for faster iteration.