GA4 vs Universal Analytics: What Actually Changed
If you used Universal Analytics for years and then opened GA4, you probably felt like you walked into the wrong building. The reports are different. The metrics are renamed. Your old bounce rate number is gone, and the new one does not match what you remember. None of this is because you missed something — Google redesigned the entire measurement model from scratch when they launched GA4 in October 2020 and forced the migration by sunsetting Universal Analytics in July 2023.
This post maps the old system to the new one so you can stop comparing numbers that were never meant to be compared.
Try it on your own data
Brass-SEO reads your GA4 and GSC data and explains it in plain English. Start your 3-day free trial — $25/month after that.
Quick Navigation
- What Universal Analytics Was (and Why People Liked It)
- What GA4 Changed
- The Side-by-Side Metric Map
- Why the Numbers Do Not Match
- What to Actually Do About It
- How Brass-SEO Handles GA4 Data
- Frequently Asked Questions
What Universal Analytics Was (and Why People Liked It)
Universal Analytics tracked website activity using a session-based model where every visit was a container of pageviews, and metrics like bounce rate had simple, intuitive definitions that business owners could understand without training. Google launched Universal Analytics in 2012, and it became the default analytics tool for over 28 million websites worldwide.
People liked it for specific reasons:
Bounce rate made sense. A bounce was any visit where someone viewed only one page and left. That was the whole definition. You could explain it to anyone in ten seconds.
Sessions were straightforward. A session started when someone arrived and ended after 30 minutes of inactivity. Every pageview, event, and transaction lived inside a session.
Reports had clear names. Audience, Acquisition, Behavior, Conversions. Four categories. You knew where to look for what.
Goals were simple. You defined a URL (like a thank-you page) as a goal, and Universal Analytics counted how many sessions reached it. No configuration beyond "which page counts as a conversion."
The numbers were stable. You could check your bounce rate every month for five years and know you were measuring the same thing each time. The definitions never changed.
Universal Analytics had limitations — it struggled with cross-device tracking, its session model did not capture complex user journeys well, and it could not handle app analytics alongside web analytics. But for a small business owner who wanted to know "how many people visited my site and what did they do," it worked.
For a broader look at the GA4 experience from a business owner's perspective, see I Don't Get Google Analytics.
What GA4 Changed
GA4 replaced the session-based measurement model with an event-based model where every user interaction — pageview, click, scroll, form submission — is recorded as a discrete event, fundamentally changing how sessions, bounce rate, and conversions are calculated. Google built GA4 to handle both web and app analytics in a single property, to work with privacy-first tracking (limited cookies, consent mode), and to support machine learning for predictive metrics.
Here is what changed:
Everything became an event. In Universal Analytics, pageviews, events, and transactions were different hit types. In GA4, everything is an event. A pageview is an event. A scroll is an event. A purchase is an event. This is more flexible, but it means the interface and reports are organized differently.
Sessions are calculated, not measured. Universal Analytics started a session when a user arrived and ended it after inactivity. GA4 generates a session_start event and groups subsequent events into that session, but session counting rules differ. GA4 does not restart a session at midnight or when campaign parameters change — both of which Universal Analytics did. This means GA4 typically reports fewer sessions than Universal Analytics for the same traffic.
Bounce rate was redefined. In Universal Analytics, a bounce was a single-page session. In GA4, a bounce is a session that was not "engaged" — meaning the visitor stayed less than 10 seconds, viewed only one page, and completed no key events. A visitor who reads your blog post for three minutes and leaves was a bounce in Universal Analytics but is an engaged session in GA4. For a deep dive, see Bounce Rate Is Dead. Here's What Replaced It.
"Conversions" became "key events." In 2024, Google renamed conversions to "key events" in GA4 to avoid confusion with Google Ads conversions. Same functionality, different name.
Reports were restructured. The four familiar categories (Audience, Acquisition, Behavior, Conversions) were replaced with Acquisition, Engagement, Monetization, and Retention. The Behavior section — where most business owners spent their time — was effectively split across Engagement reports.
Attribution changed. Universal Analytics used last-click attribution by default. GA4 uses data-driven attribution, which distributes conversion credit across multiple touchpoints using Google's machine learning model. Your traffic source reports may look different even if the underlying traffic is identical.
The Side-by-Side Metric Map
Of the 12 core Universal Analytics metrics, 8 changed their definition or calculation method in GA4 — including sessions, bounce rate, users, and conversions.
| Universal Analytics Metric | GA4 Equivalent | Did the Definition Change? |
|---|---|---|
| Sessions | Sessions | Yes — GA4 does not restart sessions at midnight or on new campaign parameters |
| Users | Active Users | Yes — GA4 defaults to "active users" (had engaged session in period) rather than total users |
| Pageviews | Views | Minor — GA4 "Views" counts both page and screen views |
| Bounce Rate | Bounce Rate | Yes — UA: single-page sessions; GA4: sessions under 10 seconds with 1 page and no key events |
| (no equivalent) | Engagement Rate | New — inverse of GA4 bounce rate (engaged sessions / total sessions) |
| Avg. Session Duration | Average Engagement Time per Session | Yes — GA4 measures active time, not elapsed time from first to last hit |
| Pages/Session | Views per Session | Minor rename |
| Goal Completions | Key Events (formerly Conversions) | Yes — renamed in 2024; setup method changed from URL-based goals to event marking |
| Conversion Rate | Session Key Event Rate or User Key Event Rate | Yes — GA4 offers both session-based and user-based rates |
| Source/Medium | Session Source/Medium | Similar, but default attribution model changed from last-click to data-driven |
| New vs Returning Users | New vs Returning Users | Similar — GA4 uses the same concept but underlying user identification differs |
| Landing Pages | Landing Page report | Similar — available under Engagement reports |
For an overview of which GA4 numbers deserve your attention, see The Only 5 GA4 Numbers That Matter for Your Business.
Why the Numbers Do Not Match
GA4 and Universal Analytics will show different numbers for the same time period on the same website because they use different counting rules, not because either system is wrong. Business owners who compare their historical Universal Analytics data to current GA4 data will see discrepancies in nearly every metric.
Sessions are lower in GA4. Universal Analytics started a new session at midnight and whenever UTM parameters changed (like clicking a second Google Ad in the same visit). GA4 does neither. The same browsing behavior that Universal Analytics counted as two or three sessions may count as one session in GA4.
Bounce rate looks better in GA4. A visitor who read your page for two minutes and left was a "bounce" in Universal Analytics (single-page visit) but is an "engaged session" in GA4 (stayed over 10 seconds). Most sites see their GA4 bounce rate 10-30 percentage points lower than their old Universal Analytics bounce rate.
Users are counted differently. Universal Analytics showed "Total Users" by default. GA4 shows "Active Users" by default — users who had an engaged session. This typically produces a lower number, which can look like you lost visitors when you actually just changed what counts as a user.
Conversion counts may differ. Universal Analytics counted one conversion per session per goal (if a user completed the same goal three times in one session, it counted once). GA4 counts each key event occurrence separately by default, though you can configure it to count once per session.
Traffic source attribution differs. Universal Analytics used last non-direct click by default. GA4 uses data-driven attribution, which may credit different channels for the same conversion. A conversion Universal Analytics attributed entirely to organic search might be split between organic and email in GA4.
The correct response to all of these discrepancies: stop comparing. Universal Analytics data is a closed book. GA4 data is your new baseline. Compare GA4 to GA4 going forward. For help reading your GA4 traffic data, see How to Read Your GA4 Traffic Report.
What to Actually Do About It
The transition from Universal Analytics to GA4 is not reversible — Universal Analytics stopped processing data in July 2023, and Google shut down access to historical Universal Analytics data in July 2024. Accepting the new system and learning its conventions is the only productive path.
Stop comparing old numbers to new numbers. Any comparison between Universal Analytics and GA4 metrics for the same site will produce misleading conclusions. The definitions changed. Comparing them is like converting miles to kilometers and wondering why the number went up.
Establish a GA4 baseline. Whatever your GA4 metrics show right now is your starting point. Track month-over-month changes within GA4 from this point forward.
Learn engagement rate instead of bounce rate. Engagement rate is the metric GA4 was designed around, and it gives you a clearer signal about content performance than the old bounce rate ever did. Higher engagement rate means more visitors are spending meaningful time on your pages.
Set up key events. If you had goals in Universal Analytics that you have not recreated as key events in GA4, you are missing conversion data. The most common key events for small business sites are form submissions, phone number clicks, and specific thank-you page visits. GA4 Admin > Events > Mark as Key Event.
Use the GA4 reports that matter. You do not need to learn every GA4 report. Traffic Acquisition, Landing Pages, and Engagement are the three reports that tell most small business owners what they need to know. See I Don't Get Google Analytics for a practical walkthrough.
Check your data regularly. Monthly is enough. A consistent cadence prevents data from becoming an overwhelming backlog. If you want to know whether your SEO efforts are producing results, see How to Tell If Your SEO Is Working for the specific signals to check.
How Brass-SEO Handles GA4 Data
Brass-SEO connects to GA4 using read-only OAuth access (the analytics.readonly scope) and translates your engagement metrics, traffic sources, and landing page performance into plain-English answers. Instead of navigating GA4 reports and interpreting engagement rate vs. bounce rate, you ask a question and get a direct answer.
GA4 is one of two required data sources — Brass-SEO also connects to Google Search Console so it can cross-reference what people search for (GSC data) with what they do when they arrive (GA4 data). This cross-referencing is what separates Brass-SEO from reading each dashboard independently: a page with high impressions in GSC but low engagement in GA4 tells a different story than either data source reveals alone.
You do not need to know whether a metric uses the Universal Analytics definition or the GA4 definition. Brass-SEO always uses GA4's current definitions, and it explains results in terms you can act on.
Setup takes about 2 minutes: sign up, connect GSC, connect GA4, and start your 3-day free trial at $25/month. See the setup guide for a step-by-step walkthrough, or go straight to the dashboard to get started.
Frequently Asked Questions
Can I still access my old Universal Analytics data?
No. Google shut down access to Universal Analytics properties in July 2024. All historical data from Universal Analytics is permanently gone. If you exported your data before the shutdown, you have those exports, but you cannot log into Universal Analytics to pull new reports. Going forward, GA4 is the only Google Analytics option.
Is GA4 harder to use than Universal Analytics?
GA4 has a steeper initial learning curve because the interface is different and core metrics were redefined. However, the underlying questions it answers are the same: who is visiting your site, where do they come from, and what do they do. Most business owners find it manageable after spending a few weeks with the three core reports (Traffic Acquisition, Landing Pages, Engagement). For a guided introduction, see our GA4 guide.
Why did Google replace Universal Analytics?
Google built GA4 to address three limitations of Universal Analytics: it could not track web and app data in a single property, it relied heavily on cookies (which are being phased out by browsers for privacy reasons), and its session-based model struggled to represent complex, cross-device user journeys. GA4's event-based model and machine learning capabilities were designed to work in a cookie-restricted, multi-platform environment.
Should I worry that my GA4 bounce rate is lower than my old bounce rate?
No. A lower bounce rate in GA4 does not mean your site improved — it means GA4 uses a stricter definition of "bounce." In Universal Analytics, any single-page visit was a bounce. In GA4, a visit only counts as a bounce if the visitor also stayed under 10 seconds and completed no key events. Your GA4 bounce rate is your new baseline. Track whether it improves or declines from here. See Bounce Rate Is Dead for the full breakdown.
Do I need to keep checking both GSC and GA4 separately?
You get the most insight when both data sources are used together, since GSC shows search performance (impressions, clicks, positions) and GA4 shows on-site behavior (engagement, traffic sources, conversions). Brass-SEO connects to both and cross-references them automatically, so you can ask questions that span both data sources without switching between dashboards. Both GSC and GA4 are required to use Brass-SEO — see the GA4 guide for why both matter.