Managing the GA4 dataset and reading reports: event, report, and Explore guide

After GA4 is set up, the most common question is, “The data is coming, so how do I read these numbers?” This article is for that stage. Based on Google's official Analytics Academy material, GA4 explains its interface, event management, reporting types, and Explore tool in Azerbaijani.

Source: Google Analytics Academy — official course list

GA4 interface: five main sections

There are five sections in the left navigation menu and each has its own task. Not mixing these things is half the analytical work.

• Home — general cards that apply to you and your property. The more you use GA4, the more personalized the cards are.
• (Reports) — ready-made, pre-built reports. Quick responses to frequently asked questions.
• Explore (Discover) — in-depth analysis techniques that go beyond standard reports.
• Advertising (Advertising) — analysis and attribution of online advertising activity.
• Admin — account, property, event, and user settings.

If you're working with multiple accounts or properties, you can switch between the property selector at the top of the page and any moment. In addition, you can also write a question directly in the search box above, such as “How many people visited my site last month?” -as GA4 will suggest a matching report.

Three types of events: automatic, recommended, custom

In GA4, the entire measurement is based on the event. But not all events are created the same way — there are three categories, and each requires a different effort.

Automatically collected events

As soon as you connect your site or app to GA4, the most common interactions automatically start accumulating: clicks, page views, app installation and opening. If enhanced measurement is enabled in the data stream settings, the property collects more events without additional code — scroll, outbound click, site search, file download, video engagement.

Recommended events

These are events that have names and settings predefined by Google, but need to be done by you. The advantage is that GA4 “recognizes” these names and builds richer reports for them. Recommended events are grouped by areas — e-commerce, gaming, travel, etc.

Custom events

These are all events that you have assigned their name and settings to. GA4 neither automatically collects nor recommends, but allows you to measure actions that are critical to your business.

This is explained in two examples in Google's course. An e-commerce site like Google Merchandise Store can add these events: product review, add to cart, complete purchase, refund. And the game app, like Flood-It!, is: starting at a new level in the game, completing the level, opening the achievement, sharing the points.

Custom dimension and custom metric: a forgotten step

This is the most common recurring error in GA4. You create a new custom event and start sending a setting-but nothing shows up in the report. The reason is simple: to make new settings appear in reports, you'll need to create a custom dimension or custom metric that works for them.

The course explains this with an example of Arya, a fashion blog. Arya publishes clothing and accessories reviews. He wants readers to know which brand they're most interested in — so they can create content that's more relevant to this information.

Arya creates a new event: whenever a user visits a page with a blog post, this event runs and sends the brand named in the post as a setting. Changes site tags, creates event in GA4 — but must create custom dimension on that setting to use the “brand” value in the report. This is done in Admin → Custom definitions.

Scope: event, user, item

When creating custom dimension, you need to select scope and this option directly changes the result:

• Event-scoped dimension — value is related to a specific event. Arya's “brand” setting belongs to this group because the setting is sent to the blog post view event.
• User-scoped dimension — value binds to the user and applies to all his sessions (e.g. subscription status).
• Item-scoped dimension — values related to a specific product in e-commerce (size, color, brand).

Important limitation: custom metric can only be event-scoped. The scope of auto-gathering dimension and metrics is already set, and you don't change it.

👉 GA4 setup: account, property, and data stream

Overview report and detail report difference

When navigating between reports in GA4, you'll encounter two different types, and their purpose is different.

• Overview report — summarizes information about the topic on several cards. Each card focuses on a separate question and is usually a preview of the detail report.
• Detail report — allows you to dive deep into a dimension or two, detailing the dataset in table format.

Here's how the practice flows: you see a card that catches your attention in the overview, you go to the detail report with the View report on the card and investigate why.

In the example in the course, Arya's subscription grew last week and she wants to know where new users are coming from. Steps: Select the Reports → Acquisition → Acquisition overview, then select “Last week” from the date picker at the top right and apply. After that, all the cards on the page show last week's dataset and show the trends that are causing the increase.

You can also set which reports will appear in the navigation bar. To do this, go to Library under Reports and add the finished collections or create a new collection. Each property can have up to 7 collections, and an Editor role is required for this action.

Report filter and comparison: different responses from the same report

Reports are built from dimension and metrics. Using those dimension values, you can get answers to completely different questions from the same report.

Report filter

The report filter controls which data will be displayed in the report. For example, you can only research users in specific regions and hide the remaining data. Added with the Add filter + button next to the header of the detail report, rules are based on dimensions.

Comparison

Comparison allows you to compare datasets side by side. With the Add Comparison tool at the top of each page, for example, you can compare desktop and mobile device categories. As a result, data for each comparison and for all users appear on the same screen. It's also possible to combine multiple dimensions in a single comparison.

Remember the difference: filter dataset shrink, comparison dataset sets side by side. You can create a link from the share icon on the top right to share your personalized view with your teammate.

Explore: it starts where standard reports end

The Reports section provides answers to the most frequently asked questions about your data. When you need to go deeper, you go to the Explore tool. It is possible to quickly build a new exploration by scrolling through dimension and metric.

One important thing: exploration is hidden because of suspense. Only you see and edit the exploration you create — unless you choose to share it.

Free form

It presents the dataset in cross-tab format; you arrange the rows and columns yourself, and add the metrics you are interested in. Visualization options: bar chart, donut chart, line chart, scatter plot and map. Note: free form does not support pie chart and donut chart is available.

Funnel exploration

Visualizes the steps users go through to complete a basic task or conversion. Indicates at what stage they enter and leave the funnel. It analyzes a single path that you predetermined.

Path exploration

Unlike Funnel, it's free-flowing: it shows how users are moving from one stage of the customer path to another, including paths you may not have thought of. For example, a cycle (loop) may reveal behavior that signals that users are “stuck.”

Clearing the collected dataset: cross-domain, unwanted referrals, and data filters

The data itself must be clean in order for the reports to be correct. The course explains this over Can's e-commerce site — Can sells sneakers, but processes payments on a separate site.

Cross-domain measurement

When the user makes a purchase decision, they are redirected to the payment site, and once the order is successfully placed, they will return to the main site. If there is no cross-domain measurement, GA4 will see this as two separate users and there will be a cutoff in the report. Can resolves this through Admin → Data Streams → existing web stream → “Configure tag settings” → “Configure your domains” and establishes a rule covering the payment domain.

Unwanted referrals

Referrals are other sites that redirect users to your site. When you look at can traffic sources, along with social media and fitness blogs, you see a third-party site where customer accounts are managed. A user who forgets their password is redirected to that site and comes back-this isn't actually referral traffic. Can Admin → Data Streams → “Configure tag settings” excludes this domain by way of → “List unwanted referrals”.

Data filters

The data filter completely removes the incoming event dataset from the processing process of GA4. There are two main types: developer traffic and internal traffic (by IP range).

Here is a very serious warning: the data filter is not backward, and permanent. Data that is removed while the filter is active never reaches GA4 and cannot be recovered later. If you turn off the filter, only data is processed after the moment of shutdown. Therefore, be sure to check in test mode before applying the data filter.

Conversion to key event: change in naming

Conversion is any user action that is valuable to your business: buying a product, subscribing to a newsletter, etc. To measure it, you must first make sure that the relevant interaction is collected as an event, then mark that event as a conversion.

Important update: The main actions in GA4, marked as conversion, are now key event it's called. The reason was practical — the conversions in GA4 were measured differently than the conversions in Google Ads, which created a discrepancy between the two platforms. Now, the logic is this: important actions for your business success are called key events in GA4, and conversion is when those key events are shared with Google Ads.

In practice, this is also evident in reports. In the Traffic acquisition report, you can scroll to the right and go to the “Key events” column, select only a specific key event (for example, purchase) from the drop-down menu. The same table view is also available in the Pages and screens report — the dataset, for example, is restricted to the newsletter unsubscribe key event, which pages bring the most subscriptions.

Frequently asked questions (FAQs)

I created an Event, but it doesn't appear in the report. Why?

There are two main reasons. First, the event hasn't happened yet — GA4 shows only events collected as fact in the table. Second, and more common: you haven't created a custom dimension for the event's setting. Please check the Admin → Custom definitions section.

Is the report filter the same thing as the data filter?

No, and that distinction is critical. The report filter only affects the view of the report-the data is in place, not just displayed. The data filter does not release the data to GA4 at all, and this cannot be undone.

How many reports can be collected on a property?

Maximum 7 collections. The role of Editor is required in the property to edit collections.

What's the difference between funnel exploration and path exploration?

Funnel exploration analyzes a single path you have predetermined. But Path exploration is free-flowing and shows all the ways users realistically pass, including the unexpected.

Can I choose user scope for custom metric?

No. Custom metric can only be event-scoped. User and item scope are only available for custom dimension.

Results

Reading a report in GA4 isn't about looking at numbers-it's about how a dataset was collected, which event was converted to dimension, and which filter cropped. The Reports section is for daily questions and Explore is for in-depth research. And the data filter and cross-domain settings set the reliability of your background reports.

Once you've mastered this database, the next step is to set up key events, customize reports, and integrate Google Ads.

👉 Setting up a key event, custom report, and audience in GA4

Source: Analytics Academy — Manage GA data and learn to read notifications course