In-depth analysis in GA4: key event, custom report, segment, and audience

After mastering the basics of GA4, the next question is, “What do I do about this dataset?” This article is dedicated to this passage — the step of converting the dataset from report to decision. Based on Google Analytics Academy's mid-level course content.

Source: Google Analytics Academy course list

Key event: the most valuable action of your business

GA4 collects everything that happens on your site and app as an event. However, not all events have the same value — a purchase with page views is not essential in the same size. That's why you mark the most important actions in GA4 as a key event.

Typical key event examples: newsletter subscription, add products to cart, complete purchase, download the app, create an account, request a quote.

Why was the Conversion called a key event?

In the past, actions that were flagged as conversions in GA4 were measured using a different methodology than conversions in Google Ads. As a result, the same campaign showed different numbers across two platforms, making it a serious mess for marketers.

The current logic is this: events that measure critical actions for your business success are in GA4 key event called, but when these key events are shared with Google Ads conversion perceived as such, so the performance of advertising campaigns is measured more accurately.

How key event reports work

The true power of Key events is revealed in reports. In the Traffic acquisition report, scroll to the right and go to the “Key events” column — you can only select a specific key event (for example, purchase) from the drop-down menu. This shows which channel is actually bringing in sales, not just traffic.

The same table view is also present in the Pages and screens report. If you limit the dataset to a newsletter subscription key event and sort it by that column, you'll instantly see the pages on your site that bring you the most subscribers.

Key events answer these types of business questions:

• How do I measure the movement that's particularly important to my business?
• Is my site experience effective for boosting key events?
• What are the actions that result in a purchase event? For example, how does sending a form or downloading a PDF affect your purchase at the end?
• Are social media shares effective at boosting key events?
• How do I measure gross revenue from purchases, in-app purchases, subscriptions, and advertising revenue?

E-commerce data requires additional steps

Caution: e-commerce events aren't sent automatically because they require extra context to make sense. To collect e-commerce data, you'll need to add e-commerce events to the site, app, or Google Tag Manager container. You can also see how customers interact with your products by adding settings like product, price, category to these events.

👉 GA4 reports and Explore guide

Custom report: where ready reports don't arrive

GA4 has a lot of ready-to-go reports, but sometimes your question is a little different from the dataset in local reports. In this case, you customize reports. Maximum on a property 150 custom reports could be.

Personalization options: change dimensions, change metrics, change graphics, save filter to report, create summary card, close or remove report from navigation.

Editing an existing report

Here's the simplest way. Example from the course: Traffic acquisition report is based on session default channel group dimension because of suspension, but you want to see it by campaign, source, or vehicle.

1. Select the pencil (edit) icon in the top right corner.
2. Open the Dimensions box on the right — you'll see all the dimension included in the report.
3. Select “Set as default” from the three dots icon next to “Session source /medium” dimension and Apply.
4. You can also adjust the graphic type — hide it with the eye icon next to the bar chart.
5. When saving, there are two options: Save changes to current report (writes on existing report) and save as new report (generates new report).

Recommendation: if you write on an existing report, keep it as a new report and give it a meaningful name. As such, team members continue to see the standard report in its entirety.

Creating a custom report from scratch

Go to Library (at the very bottom of the Reports sub-navigation), select the Create new report button and click the Create detail report option. You can start with a template or start with a blank report.

In the example in the course, Google Merchandise Store's member_level a report is set up with a custom dimension called — indicating which status layer the client belongs to. Sessions, Total users, Purchases and Total revenue metrics are added to this dimension and the report is saved as Member Info.

Custom report or exploration?

These two tools are often confused. Here's the difference:

• Custom report — you adapt familiar metrics to your own needs. Ideal for information you view on a regular basis; once set up you use it continuously.
• Exploration is a more in-depth, one-off survey of customer behavior. You're setting up ad-hoc requests and scouring the data quickly.

Segment and audience: two concepts mixed

The difference between the two is the most frequently mistaken topic on the exam as well as in practice. Simple definition:

• Segment — GA4 is a subfolder of your dataset used in Explorations. It is created in Explore and appears there, and is for analysis.
• Audience — is a user group that you set in GA4 and then share with Google Ads and/or Google Marketing Platform. For activation (targeting).

In a sentence: segment understand for audience aim to

Segment overlap technique

In Explore, you can compare up to three user segments with the segment overlap technique and see how they intersect.

The example in the course is very clear: new users who see a summer promotion make a purchase, but old (returning) users who see a summer promotion make a purchase more purchases. Based on this finding, you can change the campaign's targeting settings to focus on returning users. Then it is possible to narrow the targeting further by looking at the distribution of purchases by device types.

How Audience is Filled

Every time a user visits your site or app and its behavior matches your audience conditions, it is added to that audience. Important technical detail: end when filling GA4 audience 30 days it uses data. That is, if you already have users who match your definition, the audience fills up immediately — you don't have to wait from scratch.

Predictive audience

Predictive audience is audience based on at least one conventional predictive metric. By modeling the data you collect, GA4 predicts how your users will behave in the future if the minimum data volume is collected:

• Which users are most likely to buy (purchase probability)
• Which users are likely to leave the site/app (churn probability)
• How much revenue is expected of users (forecasted revenue)

That's the practical value: figuring out who you're going to spend the marketing budget on. For example, you can create an audience of “users who are only likely to make a purchase in the next 7 days” and focus their remarketing budget on them.

Business questions answered by Segments

• How do I identify customers whose purchase volume is declining as a competitor offers a cheaper price?
• How do I set up a segment for remarketing from users who visit specific pages?
• How do I see users who add products to the cart and don't complete the purchase?
• How do I know which landing page or traffic source users spent more than $100?

Connecting GA4 with Google Ads

When you link your Google Ads account to your GA4 property, you get a full view of the customer life cycle-from clicking an ad to completing a key event on the site.

Key benefits of connection:

• Cross-channel measurement — you measure the impact of each channel on behavior; including both paid and organic efforts. GA4 values cross-channel attribution as a signal for a promotional offer, which can improve ROI.
• Detailed behavior reports — you see the behavior of users who come within a specific promotion on the site and optimize the experience.
• Powerful audiences — You get access to precisely targeted audiences based on the behavioral dataset in GA4.

To set up a connection: Admin → Product links → Google Ads links in GA4 Select the → “Link” button → accounts → Confirm → Submit. It may take up to 24 hours for the Google Ads dataset to appear in GA4. Prerequisite: You must have an Administrator or Editor role in GA4 and an administrative login in Google Ads.

Advertising section and attribution

Once the connection is established, the Advertising section of GA4 opens. This section focuses on paid marketing conversions and attribution data.

In the course scenario, Google Merchandise Store's marketing manager evaluates the performance of campaigns from the beginning of the year to the present. Here's what the flow looks like:

1. Advertising snapshot — quick overview. Switch from Primary channel group card to key events and see which channels bring results.
2. All channels — hurts the dataset in more detail; sees how much credit each dimension value receives with line and column graphs.
3. Promotional key events — the dimension of the report has changed from channel group to Campaign to measure the performance of specific campaigns.
4. Conclusions — decides the budget based on the metrics in the table.

The main metrics in the Advertising section are:

• Key events — the number of users launching the key event
• Advertising cost — total amount paid to advertisements
• Cost per key event — dividing the total cost by the selected number of key events
• Ad clicks — total number of clicks
• Cost per click (CPC) — average amount paid per click
• Total revenue — gross revenue from purchases, subscriptions, and advertising
• Roas (return on ad spend) — Split revenue from selected key events into advertising spend

Attribution modeling is one of the key tools for assessing marketing performance. Due to suspension in Advertising section GA4 data-driven attribution the model is used, but you can compare the effect of different attribution models on the value of your channels.

Frequently asked questions (FAQs)

Can I convert a segment into an audience?

Yes. The segments you create in Explore can be converted into an audience for remarketing or bid optimization. This is the fastest way to switch from analysis to activation.

When you create an audience, is the data collected from scratch?

No. GA4 uses the data from the last 30 days when it was available to fill the audience, so if you have qualified users, the audience immediately fills up.

How many custom reports can be found on a property?

Maximum 150 custom reports.

I closed Google Ads, but the data is not visible.

It may take up to 24 hours for the Google Ads dataset to appear in GA4. Also, check that you have an Administrator or Editor role in GA4 and an administrative login in Google Ads.

What attribution model works in the advertising section?

Suspension data-driven attribution. You can check other models in comparison mode.

Results

At this stage, GA4 is no longer just a reporting tool-it's the source of marketing decisions. Key events show you what's really valuable, custom reports allow you to monitor this value regularly, segments allow you to understand the behavior, and audiences turn that concept into an advertising campaign.

The Google Ads connection and the Advertising section connect this entire chain to revenue: what channel spent and how much revenue it generated.

👉 Combining GA4 with BigQuery and other tools

Source: Analytics Academy — Dive deeper into GA data and reports course