Google Analytics 4 vs. Universal Analytics: what changed?

The most crucial thing to understand from this article is that GA4 isn’t just an update to Universal Analytics; it’s a complete rebuild, offering a brand-new way to measure, report on, and understand how people use your website. Universal Analytics (also called GA3) began in 2012 and was based on page visits. GA4, however, focuses on what people do (events) on both websites and apps, and is designed with user privacy in mind. Because of these many changed definitions and calculations, if you’ve compared numbers from GA4 and UA and found them to be different, you’re not mistaken.

Many marketers are still puzzled by the true differences between GA4 and UA, and that’s why we’re going over them today: the data models, the main metrics, how conversions (key events) are determined, the reports, and how attribution works. We aren’t trying to make GA4 work like UA, but to help you become comfortable using GA4 itself and to answer the questions that really help your business grow.

1) From sessions to events

This is the core change from one framework to another:

• Universal Analytics

  UA grouped users’ actions into sessions (website visits within a given period). Things like pageviews, events, and purchases happened within that session, which is why “bounce rate” and “pages per session” were important; they told you what happened during that one session.

• Google Analytics 4

  GA4 tracks events for everything (page view, scroll, click, purchase, downloads). It doesn’t categorize interactions by “hit type”, so each action someone takes is simply an event and can include extra details (parameters). This gives you a consistent way to analyze data from both your website and app, and more flexibility in tracking what’s important to you.

To truly benefit from GA4, you need to understand this shift in how it works from the beginning. In GA4, each event can include parameters (such as value, currency, or a form’s ID) to provide context. You can send many parameters along with an event, but to use important ones in the usual reports, you have to register them as custom dimensions or metrics. Then, key events (previously conversions/goals) should be flagged.

2) Users, sessions, and pageviews: same words, different math

Let’s clarify other important terms that have changed slightly as you move from UA to GA4:

• Users

  In UA, this meant the total number of users. In GA4, it means active users – people who had an engaged session on your site within the time period you’ve chosen. This small change creates a big difference in the reports, so when you talk about users, remember you aren’t comparing the same thing in both platforms. You can still find “total users” in GA4, if you need to (it’s just not the default).

• Sessions’ length

  In UA, a session would end or restart after 30 minutes of inactivity, at midnight, or when new tracking codes (UTMs) were added. This meant one person could easily have multiple sessions in a short amount of time. In GA4, a session ends after 30 minutes of inactivity, which usually results in fewer sessions. GA4 also removes some duplicate sessions using “session_id”, so totals can be different even when looking at the same timeframe.

• Pageviews / Views

  UA had “pageviews” and “unique pageviews”, but GA4 just reports “views” (which combines pageviews from websites with screen views from apps). If your GA4 property includes data from your app, “views” will be higher than UA’s “pageviews”. Also, GA4’s built-in Enhanced Measurement feature, which automatically tracks certain events, is important here: it can track views on Single-Page Applications (SPAs) without any extra code by detecting changes to the history, resulting in more accurate view numbers for sites built with these frameworks.

3) Key events (conversions) vs. goals

When we talk about real success metrics and results, things are quite a bit different between the two platforms:

• In UA…

  … you set goals based on what a user did, and these actions were counted as conversions. UA had five types of goals: destination (reaching a certain page), session duration, pages per session, smart goals, and event goals. For instance, if you said submitting a particular form was a goal, every time someone submitted it, that was recorded as a conversion. UA counted only a single conversion per session for each goal.

• In GA4…

  … we talk about key events instead of conversions. You decide which events are your key events, and by default, GA4 counts each time a key event occurs, even if it happens multiple times within a single session. Though you can change this to count each key event “once per session” (and this is a good idea for actions that shouldn’t be repeated, like a lead form submission). Some UA goal types (like smart goals or session duration) don’t have a direct equivalent in GA4. You’ll either recreate them with events and additional parameters or realize they don’t translate perfectly.

4) Engagement replaces “bounce”

In UA, “bounce rate” measured the percentage of sessions with only one pageview, which could unfairly penalize a great single-page experience. It was calculated by dividing the number of single-page sessions by the total number of sessions.

GA4 focuses on engagement, using metrics like engaged sessions, engagement time, and engagement rate. This is a better approach, particularly for content-focused sites, product pages, and SPAs that aren’t really about going to multiple pages.

In GA4, if you do choose to use bounce rate, it’s just 100% minus the engagement rate. Also, a session is “engaged” if it lasts over 10 seconds (you can change that number), includes a key event, or has at least two page or screen views. So, “engagement rate” is a more helpful guide than UA’s bounce rate.

5) Cross-platform system

GA4 is a system that works across all platforms. UA had separate property types for your website and your app, and gave you the option to create separate reports for each. GA4 lets you combine your website and app data in one property by adding multiple data streams (for your website, iOS app, and Android app).

You can have as many data streams as you need, each representing a different data source, device, or platform. This allows you to sort and filter your data however you like, and is more flexible than the reporting views in UA.

This structure, and the way GA4 handles events in general, means you get consistent reports and can build audiences that work the same way on all platforms without needing to manage multiple properties.

6) Stricter privacy posture

GA4 is more respectful of people’s privacy, and this is important because of increasingly strict data protection laws (like the General Data Protection Regulation (GDPR) in Europe and the California Consumer Privacy Act (CCPA) in the United States):

• GA4 doesn’t save IP addresses: UA relied on third-party cookies (which are no longer an option now), but GA4 uses your own first-party data and Artificial Intelligence (AI) to compensate. For a complete, non-sampled history of your data, BigQuery is a good option.
• Event data retention: you can only keep user-level event data in GA4 for either two months (which is the default) or 14 months. UA let you keep it from 14 months to forever. This limit applies to the data you use in Explorations – standard reports aren’t limited in the same way.
• Consent Mode: if you implement this feature, it helps GA4 respect people’s choices about their data and still create a model of what’s happening when it’s allowed.
• Reporting identity: GA4 can combine different pieces of information (including a User-ID or Google signals, if you’ve enabled them) to identify the same person using different devices. Again, you’ll likely find that the number of users differs from UA, which used Client IDs to identify users.

7) Reporting and workspace differences

Don’t go looking for the same reports in GA4 as you had in UA, because GA4 measures things in a completely different way. Instead of the ready-made reports you just looked at, GA4 offers a smaller selection of standard reports, but with much more control over how your website or app is set up. Let’s look at how this works in practice:

• Explorations vs. UA standard reports

  Compared to UA’s many pre-built reports, GA4 has fewer views, and all the detailed analysis is done in the Explorations section. Consider it your personal dashboard, where you can put together the analysis you need: drag metrics into free-form tables, create steps in a customer journey, see the paths people take on your site, compare different segments, add breakdowns and filters, save your favourite templates to use again, and export results.

• Views

  GA4 doesn’t have view-level filters anymore. You control what’s included in your data with property settings (like filtering out internal traffic and unwanted referrals), by using collections and topics in the Reports library and in Explorations for quick analysis. If you were using view-level filters in UA, recreate them with data filters and, if necessary, use separate testing and live versions of your GA4 property.

• Attribution

  GA4’s main reporting model is data-driven attribution, which you can compare with other models. Lookback windows can be changed (say, from 30 to 90 days), and GA4 might adjust which key events it gives credit to for a few days after they happen, which is another reason why comparing to UA almost in real-time isn’t always the same.

• Channel groups

  The default GA4 channel categories (like organic search, direct, etc.) aren’t the same as UA’s, but you can change them in GA4 if you need to get the same reports as before.

• BigQuery accessibility

  In UA, getting all your raw data into BigQuery required a paid upgrade. In GA4, any property can send all of its raw event data to BigQuery and then use SQL to do much deeper analysis. This lets you work with completely unsampled data, combine it with customer data from your CRM or advertising platforms, ensure all transactions are counted only once, build custom attribution models, and run your own retention models.

• Sampling and thresholds

  Most of GA4’s standard reports don’t use sampling (meaning they use all the data). But Explorations can use sampling and privacy thresholds (particularly if you are using Google signals). If this happens, you can double-check the results against your data in BigQuery.

• Annotations

  GA4 doesn’t have a way to add notes directly to reports. We recommend that your team share a list of changes, or add notes to dashboards in Looker Studio, to remember when changes were made to your website or campaigns. Some teams even create a simple “annotation” event and send it to BigQuery to keep a record.

• Debugging

  UA didn’t have DebugView, but GA4 does. It lets you check that your events and parameters are working correctly in real time as you test them (particularly after you’ve changed your tagging). You can click an event to view its parameters and confirm that important events are marked correctly before reviewing the reports.

8) Integration with Google Tag Manager (GTM)

In Universal Analytics, because events weren’t central to how it worked, GTM wasn’t quite as essential. That’s changed with GA4. While GA4 can be used without GTM, because GA4 treats everything as an event, GTM is now a much more important part of the overall system.

Google is even actively encouraging you to use GTM for collecting data and GA4 for combining, looking at, and reporting on that data, because GTM can gather so much detail.

And remember, in UA you used a Tracking ID (UA-XXXXXXXX-X) with GTM, but in GA4 you use a Measurement ID (G-YYYYYYYYYY).

9) E-commerce workflow

For e-commerce businesses, both UA and GA4 focus on a purchase event, but the additional information you send with it is different. GA4 wants a full items array with all e-commerce events (add_to_cart, begin_checkout, purchase, and so on), and it requires parameters such as transaction_id, value, and currency with the purchase event.

For refunds, send a refund event with the original transaction_id to get the revenue figures correct. Inconsistent or missing transaction IDs are a very common reason for the number of purchases being different between UA and GA4, so be sure to always include a unique transaction_id with every purchase and refund.

Summary: what you need to remember

Google Analytics 4 and Universal Analytics are very different in many ways, so it’s easy to get confused. This summary is to help you keep things clear:

So, having the right mindset for GA4 means…

…. comparing trends and behaviors, not just getting the same numbers as before. And to do that:

• Focus on events and parameters

  Replace UA goals with specific key events and parameters that demonstrate real results. For example, use “generate_lead” with the form’s name, or “purchase” with a full items array.

• Choose the right counting method

  If you were counting a goal once for each time someone visited in UA, in GA4 you would count that key event once for each session to keep business logic consistent.

• Lean into engagement

  Look at engaged sessions, engagement rate, and engagement time instead of UA’s bounce rate.

• Combine data where it helps

  If you have both an app and a website, add the app’s data to your web data and create audiences that work for both. This will make your reports more consistent.

• Tighten admin settings early

  Filter out internal traffic and unwanted referrals, ensure cross-domain and consent settings, decide on your user identification method, and so on. These settings determine how good your data will be from the very beginning, so do them as soon as you can to avoid having to redo them later. We have a “Google Analytics 4 setup checklist for new adopters” to help you with this.)

• Plan for analysis

  Use the standard Reports section to quickly check things are working, Explorations to investigate in detail, BigQuery for all your raw, unsampled analysis, and for connecting your analytics to your CRM.

Final thoughts

Universal Analytics is completely over. Google Analytics 4 is the present and the future. It’s more flexible, organized around events, respects people’s privacy, and is designed for people who use multiple platforms. If you try to get GA4 to show the exact same numbers as UA, you’ll be frustrated, but if you understand the main concepts of GA4, you will have reports that reflect how your business really operates.

And if you want to go even further, use GA4 with Google Tag Manager for precise event tracking. If you need even cleaner data and are worried about privacy and ad blockers, consider server-side Google Tag Manager. We go into detail about this in “How to set up GA4 via GTM server-side tagging”.