How does Google Analytics 4 work with events?

Understanding how Google Analytics 4 works stopped being a migration project and became a basic job requirement. The platform is Google's measurement standard, and anyone opening a report today is reading data organized by events, not by pageviews.

That structural change is bigger than it looks. It redefines what counts as a visit, what counts as engagement and what counts as a conversion, which is why old numbers do not compare directly with current ones.

For anyone building digital marketing strategies, the effect is immediate. A goal that made sense on the previous platform may be measuring something else today, even when the dashboard label stayed the same.

The purpose here is conceptual and direct. You will see what the tool measures, why it measures that way, and how to read the main indicators without drawing the wrong conclusion from a chart that looks familiar.

What Google Analytics is and how it works in version 4

Google Analytics is Google's digital measurement platform, and Google Analytics 4 is the version in operation since October 2020. It collects interactions from websites and apps inside the same property, organizes everything as events, and returns reports on acquisition, behavior, engagement and conversion.

Caption: How Google Analytics 4 works after the Universal Analytics shutdown: every visitor interaction becomes an event carrying its own parameters.

The question of what Google Analytics is and how it works usually arrives with an old expectation attached: that the tool counts visits and pages. That expectation describes the previous generation.

In the current design, the smallest unit of information is the event. A pageview is an event, a button click is an event, a completed purchase is an event, and each one carries parameters that describe the context.

That uniformity is what allows website and app to be measured with the same logic. The same purchase event can come from the browser or from the app, and the report treats both as the same business action.

For anyone working with web analytics applied to business decisions, the practical consequence is simple. The report stops answering "how many pages were viewed" and starts answering "what did people do".

What changed with Google Analytics 4 compared to Universal Analytics?

What changed with Google Analytics 4 was the data structure, not only the interface. Universal Analytics organized measurement around sessions and pageviews, with events as a complementary resource. Google Analytics 4 reverses that: everything is an event, and the session becomes a derived reading.

Universal Analytics is a closed chapter today, not an alternative in use. According to Google's official documentation on the shutdown, standard properties stopped processing hits on July 1, 2023.

The next cut was access itself. From the week of July 1, 2024, current and historical data stopped being available, and most users lost access to the old interface and to its API.

Google's own page explains how to read what is left: after July 1, 2024, a property that no longer appears in the interface should be considered terminated. There is no way back and no parallel operation.

Universal Analytics therefore remains useful only as a conceptual reference point. It explains why certain indicators changed value, and that comparison is what prevents misreading historical series.

Three differences account for almost all the confusion among teams that moved from one platform to the other. Here is how they line up:

Table: Conceptual comparison between the two generations of Google measurement.

Those three lines explain why an old chart does not overlap with a current one. It is not a data discrepancy, it is a different metric definition.

Teams that still need to organize the current property will find the operational walkthrough in the content about how to set up Google Analytics 4 today, which covers the hands-on part left out of this text.

What are the new Google Analytics 4 features?

What are the new Google Analytics 4 features: unified measurement across website and app, machine learning models applied to prediction and to gap filling, consent controls aligned with privacy regulation, and the Explorations area, where the more open-ended analysis lives.

Machine learning shows up in two places. It feeds predictive audiences, which group users by probability of purchase or churn, and it supports the modeling used when part of the collection is limited by tracking restrictions.

Privacy stopped being an accessory and became part of the design. The platform works with consent mode and with measurement that does not depend exclusively on cookies, which brings it closer to the requirements of rules such as GDPR and CCPA.

The interface separates three reading environments. Standard reports answer recurring questions, the realtime report shows what is happening now, and Explorations let you build custom analysis.

Funnel analysis lives inside Explorations, and it is what shows where the journey breaks. The detail of that build is covered in the content about the conversion funnel in Google Analytics.

For the marketing professional, the consequence is a clearer division of labor. The standard report serves daily monitoring, and the Exploration serves the investigation of a specific problem.

How does event-based collection work in Google Analytics 4?

How Google Analytics works at the collection layer is the question that separates people who read the report from people who trust it. The answer sits in the event model: each recorded interaction becomes a named event with parameters, and Google's official documentation on events sorts that collection into four types.

The first type covers automatically collected events, which arrive by default as soon as installation is complete. They need no extra configuration and cover basic navigation interactions.

The second covers enhanced measurement events, collected when the feature is active on the property. They widen the capture of common interactions without requiring development work.

The third covers recommended events. You implement each one, but the name and the parameters come predefined by Google, as happens with `article_read`, `video_play` and `newsletter_signup`.

The fourth covers custom events, defined by you when none of the others fits the case. Google documentation recommends treating them as a last resort, and the reason is practical.

An unnecessary custom event falls outside the ready-made reports and outside the platform's standard vocabulary. The result is a property full of private names that nobody can compare across periods or across projects.

The decision order is therefore always the same. Look first for a recommended event that describes the action, and only create a new name when the action genuinely does not exist in the list.

Understanding how Google Analytics works at this level of hierarchy prevents rework months later. A badly chosen event name cannot be fixed retroactively, because the history was already written under the old name.

What is engagement rate and how to analyze an engagement rate?

Engagement rate is the percentage of engaged sessions within total sessions. According to Google documentation on engaged sessions, a session is engaged when it meets any one of three criteria, not the combination of them.

The three criteria are objective. The session lasts longer than 10 seconds, or records a key event, or adds up to two or more page or screen views.

One criterion is enough for the session to count as engaged. That is the part that produces most reading errors, because many people read the definition as if all three were mandatory at once.

Current bounce rate is exactly the inverse of that calculation. It measures the percentage of sessions that were not engaged, which is why it is not comparable with the bounce rate of the previous platform, which looked only at single page sessions.

Knowing how to analyze an engagement rate without fooling yourself requires, above all, knowing which criterion is pulling the number up. A site with a long video and a single page can show high engagement by time, with no extra navigation at all.

The way through is to cross the metric with what matters to the business. Compare engagement rate by channel, by device and by landing page, and check whether engaged sessions are reaching the key events that represent value.

It also pays to compare equivalent periods. An engagement swing between weeks with very different volume usually says more about traffic source than about the user experience on the page.

When the reading shows reasonable engagement and weak conversion, the problem is rarely in the measurement. It sits in the page, and that is where work on user experience applied to performance comes in.

How to integrate with Google Ads and measure e-commerce in GA4?

How to integrate with Google Ads is one of the most used capabilities of the platform, because it connects media investment to later behavior. With the link active, conversions measured in Google Analytics 4 can be imported into the ads account, and audiences built in analysis become available for targeting.

The gain is continuity. Instead of looking at clicks on one side and results on the other, the team follows the campaign end to end, from the ad to the event that represents closed business.

That stitching supports marketing strategies that do not depend on an isolated report. The success criterion becomes the same on both sides, and the budget discussion gets more objective.

The budget conversation also changes its basis. Comparing media platforms, the subject of the content about Google Ads or Meta Ads, starts to rest on behavior measured on the site, and not only on metrics declared by each advertiser.

In e-commerce, the event logic fits almost naturally. Every stage of the purchase journey has a matching event, from product view to checkout start and completed transaction.

With that sequence instrumented, you can see revenue, average order value and abandonment at each step of the funnel. The report stops showing only how much was sold and starts showing where the sale failed to happen.

The same design works outside classic e-commerce. Forms, sign-ups and contact requests can also be marked as key events, which allows demand generation to be treated with the same rigor as an online store.

How to analyze the data in Google Analytics 4?

How to analyze the data in Google Analytics 4 starts with a business question, never with a report opened at random. Define the decision that has to be made, choose the metrics that inform that decision, and only then open the screen that answers it.

A simple sequence covers most cases:

• Set the period and the segment that make sense for the question.
• Compare the indicator against an equivalent period, not against last month out of habit.
• Isolate the variable that changed, whether channel, device or landing page.
• Check that the event supporting the metric is actually configured.
• Write down the hypothesis the data suggests before changing anything.

The last item is the one that usually goes missing. Without a written hypothesis, analysis becomes a retroactive justification for what the team already wanted to do, and the data loses its role as arbiter.

Conversion rate is the indicator that ties all of this together. It only makes sense when read by source, because the same site converts very differently depending on the channel bringing the visit, as the content about conversion rate by channel shows.

This is where analysis meets CRO (Conversion Rate Optimization). Optimization work needs a testable hypothesis, and the hypothesis comes from reading engagement, funnel and page-level behavior.

The relationship runs both ways, and it is detailed in the content about CRO and conversion rate optimization. The platform says where it hurts, and the test says what fixes it.

When an operation reaches that point and needs a method for prioritizing tests, structured CRO run by specialists speeds up the move from diagnosis to real conversion gains.

A marketing professional who opens the report with that discipline can defend marketing strategies with evidence instead of impression. That is the difference between reporting a number and leading a decision.

One caveat about expectations is worth stating. Well collected data does not produce digital marketing strategies on its own, it narrows the space for guesswork and makes errors cheaper to find.

Frequently asked questions about how Google Analytics 4 works

How Google Analytics 4 works when it turns into a decision

Knowing how Google Analytics 4 works only changes the outcome when the report starts answering business questions. The platform is already installed in nearly every digital operation, and the competitive edge is not in holding the data, it is in interpreting the data with judgment.

The shortest path has three moves. Confirm which events represent real value, read engagement rate knowing which criterion supports it, and always compare conversion rate by traffic source.

Once that is settled, the next decision stops being "which report do I open" and becomes "which hypothesis do I test". That is the point where analysis turns into revenue, and not just a tidy dashboard at the monthly meeting.

If your team already has the data and still lacks the method, talk to the mkt4edu team to structure reading, test prioritization and conversion optimization.