What is Google Analytics 4 (GA4) and how does it work? Complete guide

500 people visit your site per day. How many of these actually turn into customers? Which advertising channel is returning your money, and which one is just spending it? If you can't answer these questions accurately, your marketing budget is working on estimates.

Google Analytics 4 — GA4 for short — is a free tool that answers these questions. This guide explains in Azerbaijani what GA4 is, how it works, and in what order it should be studied, based on the content of Google's official Analytics Academy courses.

Source: Google Analytics Academy — official courses (Skillshop)

What problem does digital analytics solve?

Every business with a site or app has a few top questions they want answers to:

• How do users interact with your site or app?
• How effective are your marketing campaigns?
• What content helps you meet your business goals?
• Where do your users come from and which part of the site do they leave?

Digital analytics answers these questions by aggregating, analyzing, and reporting data. But for effective measurement, you must first ask yourself three questions: what business goals are you trying to reach, what metric or KPI should you improve for these goals, and what value does that metric need to reach to be a success?

The world of measurement has changed

Previously, it was possible to get almost fully observed data thanks to cookies and device identifiers. Now the situation is different: users' privacy expectations have increased, worldwide regulations have tightened, browsers' cookie support has declined. This has created gaps in digital measurement.

GA4 is built for this new reality, with modeling, consent modes, and privacy-based data collection mechanisms.

Basis of GA4: event data model

This is the main difference that separates GA4 from its predecessor: everything here eventis. Page viewing is also an event, a button click, a file download, a purchase.

Practical advantage of this model: the site and the mobile app dataset speak the same “language” and are merged into reports. Previously, it was necessary to set up separate analytics systems for web and application, but now both are merged into one property.

With every event setting sending additional information called, for example, page_view along with the event page_location the parameter — that is, the URL of the page — is also collected. Then these datasets are merged into two categories in the reports: dimension (“who, what, where”) and metric (“how much, how many”).

Learning GA4: four steps

Google's official course structure divides GA4 into four phases, and this sequence is random — each stage builds on the previous one.

Phase 1 — Instruction and data collection

You're setting up the account, property, and data stream structure. You connect the site with Google tag or Google Tag Manager. Receive the initial event set with the Enhanced measurement. With the realtime report and Tag Assistant, you confirm that the dataset is coming. You define user roles.

The most critical decision at this stage is the property structure: the same business sites and applications one must be in property so you can see the entire user path.

👉 GA4 setup: account, property, and data stream

Phase 2 — Data management and reporting

You distinguish between the three types of events: automatic, recommended, custom. You create custom dimension and metric — and master the concept of scope. You learn the difference between an overview and a detail report. With the report filter and comparison, you get different responses from the same report. In the Explore tool, you set up funnel and path exploration.

In this step, you also clean the dataset: with cross-domain measurement, unwanted referrals, and data filter parameters.

👉 GA4 reports, events, and Explore guide

Phase 3 — Deep analysis and activation

You mark the most valuable actions as a key event. You're setting up regular tracking with custom reports. You understand the behavior with segmenters, and turn that concept into an advertising campaign with audiences. With the Google Ads connection, you combine spending and revenue in one landscape.

👉 Key event, custom report, and audience in GA4

Phase 4 — Integration and scaling

With CRM dataset data import, you bring offline events to GA4 with Measurement Protocol. You export the raw dataset to BigQuery and analyze it with SQL. You set privacy settings — data retention, consent mode.

👉 GA4 and BigQuery, data import, Measurement Protocol

GA4 interface: five sections, five different goals

• Home — common property cards; personalised as they are used.
• Reports — ready reports; fast response to daily questions.
• Explore — advanced analysis techniques; in-depth research.
• Advertising — paid marketing, attribution and roas.
• Admin — account, property, event and user settings.

You can also type a question directly into the search box above: “How many people visited my site last month?” GA4 will offer an appropriate report.

GA4 setup by business target

In Google's course, the GA4 setup is explained by four different examples of entrepreneurs. These examples show why GA4 target selection is important — because, depending on your preference, GA4 generates different reporting sets:

• Merve (dog training business) — target: attract new customers. Option: Generate leads. Reports: User acquisition, Traffic acquisition, Landing page.
• Jan (selling shoes online) — target: focusing on the products that earn the most. Selection: “Drive online sales”. Reports: Ecommerce purchases, Promotions, User purchase journey.
• Gamze (photo editing app) — target: understanding user behavior and increasing pro subscriptions. Option: “Examine user behavior”. Reports: Events, Conversions, Pages and screens.
• Refic (online magazine) — target: brand recognition. Selection: “Raise brand awareness”. Reports: Google Ads campaigns, Demographic details, Pages and screens.

This option is not permanent-you can change the report collections at any time. If you're not sure, you can receive all major life cycle reports with the “Get baseline reports” option.

Key event: The most important understanding of GA4

GA4 collects everything as an event, but not all events are of equal value. The actions that matter most to your business key event you mark as: newsletter subscription, add to cart, purchase, create account, quote request.

There was an important change in terminology: these were formerly called “conversion” in GA4. The problem was that GA4 conversions were measured using a different methodology than Google Ads conversions, and there was a discrepancy between the two platforms. The current logic is: in GA4 key event, when shared with Google Ads conversion.

Segment and audience: understanding vs. targeting

These two concepts are the most often confused:

• Segment — The data cluster created in Explore. Objective: behavior understand.
• Audience — User group shared with Google Ads and Google Marketing Platform set in GA4. Purpose: aim.

GA4 uses the last 30 days of data when filling the audience, so if you have qualified users, the audience immediately fills up.

Predictive audience but it goes a step further: GA4 models the data you collect to build up predictions like probability of purchase, probability of abandonment, and expected revenue. This allows you to build audiences like “users who are likely to make a purchase in the next 7 days”.

GA4 and Google Ads: closing the circle

GA4's marketing value is fully opened when combined with Google Ads. After the connection, you see a complete picture of the customer life cycle — from the ad click to the key event on the site.

Key metrics you measure in the advertising section: key events, advertising cost, cost per key event, ad clicks, CPC, total revenue, and roas. Suspension as Attribution Model data-driven attribution it works, but you can compare the impact of other models on the value of your channels.

Key differences between Universal Analytics and GA4

When marketers with Universal Analytics (UA) switch to GA4, they often look for the same logic and don't find it. Knowing the main differences makes this transition easier.

• Measurement Unit — In UA, session and pageview were centered; in GA4, everything is event. This isn't just a name change, it changes the entire reporting logic.
• The Web and the App — Separate properties were needed for the site and mobile app in UA; in GA4, both are merged as data streams within a property.
• Engagement instead of bounce rate — GA4 brought up the concept of “engaged sessions”: a session that lasts more than 10 seconds, has a key event, or has at least two page views.
• Raw dataset access — BigQuery Export was only available in paid 360 version at UA; it is also open for free properties in GA4.
• Attribution — Data-driven attribution for suspension in GA4 works, while in UA the last-click model was standard.

The most important practical conclusion is that the UA dataset is not transferred to GA4. Although the metrics of the two systems bear the same name, they are calculated differently — so it is not a good idea to do a direct comparison.

Datasheet cleanliness: an overlooked but critical phase

The coolest reports lead to the wrong decision, even if they're based on dirty data. GA4 has three parameters that ensure the reliability of the dataset and should be adjusted in the first week of installation.

• Internal traffic filter — Excludes the IP range of your office and team. Otherwise, site visits from your own employees are counted as “customer behavior.”
• Unwanted referrals — prevents domains like payment system or third party account management site from appearing as referrals.
• Cross-domain measurement — when the user moves from one domain to another (for example, to the payment page), it ensures that he/she is recognized as the same person.

Warning: the data filter is not backward. Data that is removed while the filter is active never reaches GA4 and cannot be recovered later.

Is it worth getting a GA4?

Google Analytics Certification is free and has a real shot at a digital marketing resume — especially in agency and freelance work. The certification course consists of about 4 hours of 33-minute content and includes four modules (101, 102, 201, 301).

👉 Google Analytics certification: exam preparation guide

Frequently asked questions (FAQs)

Is GA4 free?

Yes, standard GA4 is free and sufficient for most businesses. There is a paid Analytics 360 version for corporate opportunities such as big data volume, subproperty, and roll-up property.

Is it mandatory to switch from Universal Analytics to GA4?

Yes, Universal Analytics no longer collects data. GA4 is based on a completely different data model and historical data is not directly transferred — so the sooner you build GA4, the more history you have to compare.

How long does it take to learn GA4?

Google's official certification course is about 4.5 hours of content. To work comfortably on a practical level, you'll need 2–4 weeks of experience on a real property.

Need to know programming for GA4?

Not for primary use. Thanks to Google Tag Manager, you can manage tags without writing code. During the Measurement Protocol and BigQuery phase, developer and SQL knowledge are required, respectively.

Is GA4 GDPR compliant?

GA4 contains tools that ensure compliance: IP addresses are not stored individually, you can set the data retention period, consent mode changes tag behavior according to user permission, data deletion request and User Deletion API can delete data. And the full extent of compliance depends on your setup.

Results

GA4 isn't just a visitor tool-it's a data source for business decisions. But to get this value, consistency is important: first proper setup and clean data, then ability to read reports, then activation with key event and audience, and lastly integration with other systems.

A separate detailed record has been prepared for each step in this guide — you can start with the steps you need from the links above.

Source: Google Analytics Academy — all courses