What is the Difference Between GA3 and Google Analytics 4?

If you've been working with Google Analytics for a while, you know the announcements felt like a major shift: Universal Analytics, or GA3, stopped processing new data in July 2023. The replacement, Google Analytics 4, isn't just an update - it's a completely new way of thinking about website and app measurement. This article breaks down the fundamental differences between GA3 and GA4, from the data model to the reports you see every day.

Graphed

Still Building Reports Manually?

Watch how growth teams are getting answers in seconds — not days.

Watch the Demo

From Sessions and Pageviews to Events and Parameters

The most important difference to understand is the change in the underlying data model. Everything else about GA4 branches from this one central idea.

Universal Analytics was built around the concept of sessions and pageviews. A session was like a container that held all the hits a user took on your website within a specific timeframe (typically 30 minutes of inactivity). Inside that container, you had different hit types: pageviews, event hits (which you had to manually set up), transaction hits, and so on. This model worked well for a world of simple websites, but it struggled to accurately represent the user journey across different devices or on modern single-page applications.

GA4 throws that model out. In GA4, everything is an event.

When someone views a page, it triggers a page_view event. When they first arrive, a first_visit event fires. When they start a session, a session_start event is recorded. Clicking a button, filling out a form, scrolling down a page - these are all just events. This might seem like a small detail, but it's a huge shift in perspective. Instead of focusing on "sessions," GA4 focuses on the user and the actions they take.

With this event measurement model also comes the use of parameters - additional pieces of information that add context to an event. In Universal Analytics, setting up custom event tracking involved rigid fields: Event Category, Event Action, and Event Label. In GA4, an event can have multiple custom parameters with descriptive names.

Here’s a comparison:

• GA3 Event Tracking: To track a click on a "Request a Demo" button on your homepage, you might set it up like this:
• GA4 Event Tracking: The same action would be an event named something like cta_click. You would then send relevant parameters with it:

This approach is infinitely more flexible and allows you to capture richer, more descriptive data about every single user interaction, not just the ones that load a new page.

Graphed

Still Building Reports Manually?

Watch how growth teams are getting answers in seconds — not days.

Watch the Demo

Unifying Users Across Devices and Platforms

Another weakness of Universal Analytics was its reliance on cookies to identify users. This made tracking a single user's journey across their laptop, work computer, and mobile phone nearly impossible. The result was often inflated "user" counts and a fragmented view of the customer journey.

GA4 was built to address this. Instead of separate properties for your website and mobile app (something you had to manage with Google Analytics for Firebase), GA4 uses a single property with multiple "Data Streams." You can have one data stream for your website, another for your iOS app, and a third for your Android app, all feeding data into one GA4 property.

To unify users across these streams, GA4 uses a more sophisticated approach to identity called identity spaces. It attempts to "stitch" together sessions by looking for signals in a specific order:

1. User-ID: This is a unique, anonymous ID that you provide when a user logs into your site or app. This is the most accurate signal.
2. Google Signals: If a user is logged into their Google account and has ad personalization enabled, Google can use this to recognize them across devices.
3. Device ID: This is the classic method using first-party cookies for browsers or App Instance ID for mobile apps.

By layering these methods, GA4 can provide a much more accurate, cross-platform picture of how a single user interacts with your business over time, even if they're switching between your app and website.

Major Changes to Reporting and Dashboards

For most day-to-day users, the most obvious difference is the reporting interface. Many of the familiar pre-built reports from GA3 have been removed or dramatically altered in GA4. This has been a big point of friction for many teams, but the changes reflect GA4's new focus on custom analysis.

Graphed

Still Building Reports Manually?

Watch how growth teams are getting answers in seconds — not days.

Watch the Demo

Engagement Metrics Have Replaced Bounce Rate

One of the metrics long-time marketers miss most is Bounce Rate. A "bounce" in GA3 was a session with a single "hit," meaning the user visited one page and then left. However, this metric had serious flaws. A user could land on your blog, read an entire 2,000-word article, find exactly the answer they needed, and leave completely satisfied. In GA3, that would be counted as a bounce, signaling a "bad" session.

GA4 replaces this with a much more useful set of metrics centered around the idea of an engaged session. An engaged session is one that meets at least one of these criteria:

• Lasts longer than 10 seconds (this is customizable).
• Includes a conversion event.
• Has at least 2 pageviews.

From this, you can now measure Engagement Rate - the percentage of total sessions that were engaged sessions. It’s essentially the inverse of Bounce Rate and a far better indicator of whether users are finding your content valuable and interacting with your site in a meaningful way.

The Rise of the "Explore" Hub

While GA4 has fewer standard "Reports," it gives you a much more powerful toolkit for custom analysis inside the Explore section. Where the GA3 standard reports gave you answers to common questions, Explore gives you the tools to ask your own specific questions.

Many of the features in Explore, like advanced funnel builders and path analysis, were only available in the paid enterprise version, GA360, which cost upwards of $150,000 per year. Now, they are free for everyone in GA4.

Inside the Explore hub, you can create analyses like:

• Free-form Exploration: Similar to a pivot table, this lets you drag and drop dimensions and metrics to build custom tables and charts on the fly.
• Funnel Exploration: Visualize the steps users take to complete a conversion, and importantly, where they drop off in the process.
• Path Exploration: See the most common paths users take after opening your app or visiting a certain page on your site.
• Segment Overlap: Compare how different user segments (e.g., mobile users vs. desktop users) overlap and interact.

Other Key Differences You Need to Know

Beyond the philosophical shifts, there are several key feature updates and changes that are crucial to understand.

Privacy Controls First

GA4 was built with user privacy and the future of tracking (or lack thereof) in mind. IP anonymization, which was an optional setting in GA3, is now enabled by default and cannot be turned off. GA4 also provides more granular controls for managing personal data and is better prepared for a world with fewer cookies.

Graphed

Still Building Reports Manually?

Watch how growth teams are getting answers in seconds — not days.

Watch the Demo

"Conversions" Instead of "Goals"

Setting up tracking for key actions is much simpler in GA4. In GA3, you had to configure "Goals," which had different types (Destination URL, Duration, Event, etc.) and were limited to 20 per property view.

In GA4, any event can be marked as a conversion. All you do is go to your list of collected events and flip a toggle switch for the ones that are important to your business, like generate_lead or purchase. It's that simple.

Free BigQuery Integration

This is a game-changer for data-driven companies. In GA3, the only way to export your raw, unsampled data was to pay for GA360. GA4 makes a streaming export of your data to Google BigQuery available to all users for free. This unlocks advanced analysis capabilities, allowing you to join your analytics data with CRM data, ad platform spend, and other business sources to create a complete picture of your performance.

Predictive Metrics and Audiences

By leveraging Google's machine learning, GA4 can automatically generate predictive metrics based on your data. Out of the box, it can generate metrics like:

• Purchase Probability: The likelihood that an active user will make a purchase in the next 7 days.
• Churn Probability: The likelihood that a recently active user will not visit your site or app in the next 7 days.

You can use these metrics to create "Predictive Audiences" that can be used for remarketing in Google Ads, like targeting ads to users who are likely to convert soon.

Final Thoughts

The move from Universal Analytics to Google Analytics 4 isn't just an interface facelift, it's a fundamental reimagining of what web and app analytics should be. GA4 leaves behind the session-based model for a more flexible, user-centric approach built on events and parameters, giving you a truer understanding of how people interact with your business across all their devices.

Because GA4 is so different, building the exact reports you need can feel like a new challenge. We built Graphed to simplify this process by connecting directly to your marketing and sales data sources, including GA4. You can ask questions in plain English like, "create a dashboard showing a funnel from sessions to purchases for last month, broken down by traffic source," and get a live, customizable dashboard in seconds without having to master the new Explore interface.