The Google Analytics Certification rewards understanding the GA4 model, not memorising where buttons live. GA4 rebuilt analytics around a single idea, the event, and almost every concept the certification tests is a consequence of that one decision. This guide is a full self-study course built around the GA4 model: it explains the event-based data structure, how a property is put together, how data is collected and shaped, how the standard reports and the Explore workspace turn that data into answers, and how key events, audiences and attribution measure whether marketing is working. Then it turns the lot into a short, dated study plan. It is original teaching material only. It contains no real or simulated assessment questions, and because GA4 changes often, you should confirm current behaviour against Google’s Analytics Academy and Help and the Skillshop certification path before you sit it.
Chapter 1: Exam overview and how to use this guide
What the certification measures and how it is run
The certification measures what you can do in GA4: set up a property, collect the data a business needs, use the reporting tools, and recognise the measurement features that show how effective marketing is. Google describes those four skill areas but does not publish a percentage weight per area or an official question count, so plan to study all four evenly rather than chasing a heaviest topic. Third-party figures of “around 50 questions / 75 minutes” circulate, but they are unofficial, so confirm the current format on Skillshop when you sit it.
The logistics are simple and forgiving. You take it online on Skillshop, signed in with a Google account, in your browser, with no proctor and no fee. You need 80% or more to pass. If you fall short you wait 24 hours and retake it at no cost, with no limit on attempts. A digital certificate then appears in your Skillshop profile. The certification is valid for one year and then expires, so you retake the assessment to renew. That expiry shapes the whole approach: because the badge signals current GA4 skill and costs nothing, a short focused run beats a drawn-out one, and you should plan to refresh it annually.
The one fact that reorganises everything: GA4 is event-based
The most important thing to fix before any detail is that GA4 is built on an event-based data model. In the retired Universal Analytics, data came in as distinct hit types (pageviews, events, transactions, social hits) and the session was the central unit. In GA4 there are no separate hit types. Everything is an event, and each event can carry parameters that describe it. A page view is an event. A purchase is an event. A scroll is an event. Sessions still exist, but they are derived from events rather than being the core unit. This single shift is the source of most of the differences the certification cares about, and the most common way candidates fail is by revising Universal Analytics concepts that no longer map. Throughout this course, when something looks unfamiliar, the explanation almost always traces back to “because everything is an event”.
How to use this course
Read the chapters in order once. They follow the natural workflow of measurement, which is also how the four skill areas are organised: you set a property up, collect clean data, report on it, then use the measurement features to judge performance. Treat the bold terms as a checklist you can explain in a sentence each. The worked illustrations are teaching examples, not assessment questions. The final chapter turns the content into a paced plan and describes the assessment day.
Chapter 2: Setting up a GA4 property
The first skill area is structural: creating and organising the containers that hold your data and choosing where data comes from.
Accounts, properties and data streams
GA4 has a clear hierarchy. An account is the top-level container, usually one per organisation, and it governs settings and access at the broadest level. A property sits inside an account and is the unit that actually collects and holds data for a business, product or app. Inside a property, a data stream is a source of data: a web stream for a website, or an Android or iOS stream for an app. A single property can hold multiple data streams, which is how GA4 brings web and app data together in one place, something that was awkward in the old model. The teaching point is to hold the nesting clearly: account contains properties, a property contains data streams, and data flows in through the streams. As an illustration, a company with one website and an iOS app might run a single property with a web stream and an iOS stream, so it can analyse users across both surfaces in one property.
What setting up a stream actually switches on
Creating a web data stream does more than register a URL. It gives you the GA4 tag to place on the site (or to deploy through Google Tag Manager), and it turns on a set of automatic collection behaviours you control. Understanding that a stream is the on-ramp for data, and that some collection starts automatically the moment it is live, sets up the next chapter, where the kinds of events are the main subject.
Why structure comes first
Property and stream structure decide what you can later analyse together and what stays separate. Decisions here, such as whether web and app live in one property, ripple through every report and exploration. The certification puts setup first for the same reason this course does: clean structure is the precondition for clean data and meaningful analysis.
Chapter 3: Collecting the data you need
The second skill area is about what GA4 records and how you extend it. Because everything is an event, this chapter is really a taxonomy of events.
The three kinds of events
GA4 events fall into three groups, and telling them apart is one of the most reliably tested ideas. Automatically collected events are recorded with no setup at all once the tag is live, such as first_visit, session_start and page_view. Enhanced measurement events are an optional layer you switch on per data stream that automatically tracks common interactions without code, such as scrolls, outbound clicks, site search, video engagement and file downloads. Custom events are ones you define yourself to capture something GA4 does not record by default, configured either in the GA4 interface or pushed from your site or a tag manager. The way to keep these straight is by who creates them and whether you did anything: automatic events need nothing, enhanced measurement needs a toggle, and custom events need you to define them. As a teaching example, the fact that a user scrolled 90% of a page is an enhanced measurement event you enabled with a switch, whereas “clicked the pricing-plan comparison toggle” would likely be a custom event you defined.
Parameters and user properties
Events carry parameters, which are the details that describe each event, such as the page title, the link URL, or the value and currency of a purchase. Parameters are what make events analysable beyond a simple count, because they let you break an event down by its characteristics. Separately, user properties are attributes of the user rather than of a single event, such as a membership tier or a preferred language, and GA4 can use them to segment data across that user’s activity. The distinction to carry into the assessment is event-versus-user: a parameter describes one thing that happened, while a user property describes the person it happened to. There are also limits and reserved names to respect, and personal data must not be sent, which is the responsible-collection point the certification expects you to recognise.
Testing collection with DebugView
GA4 gives you DebugView, a report that shows events arriving in real time while you test a setup, so you can confirm that an event and its parameters are firing correctly before you rely on them. The teaching lesson is that collection is something you verify, not assume. When data looks wrong later, the discipline is to check whether the right events and parameters were ever collected in the first place.
Chapter 4: Using the reporting tools - standard reports and Explorations
The third skill area is turning collected data into answers, and GA4 splits this into two tools with different jobs: the standard reports for monitoring, and the Explore workspace for investigation.
Standard reports and the lifecycle
The standard reports are the pre-built summaries, organised largely around the customer lifecycle: Acquisition (how users arrive, from which channels and campaigns), Engagement (what they do, including events, pages and screens), Monetisation (purchases and revenue), and Retention (whether they come back). GA4’s engagement metrics are themselves event-derived, which is why terms like engaged session (a session that lasts longer than a set time, has a key event, or has at least two page or screen views) and engagement rate (the share of sessions that were engaged) replace the old bounce-rate framing. The teaching point is to read the standard reports as the lifecycle story: acquisition brings people in, engagement shows what they do, monetisation shows value created, and retention shows whether it lasts.
The Explore workspace and its techniques
For questions the standard reports cannot answer, GA4 provides Explorations in the Explore workspace, where you build custom analyses from dimensions, metrics and segments. The certification expects you to recognise the main techniques and what each is for. A free-form exploration builds flexible tables and visualisations, the general-purpose tool. A funnel exploration shows the steps users take toward a goal and where they drop off, which is how you diagnose a leaky conversion path. A path exploration shows the routes users take through events or screens, which surfaces unexpected journeys. A cohort exploration groups users by a shared starting behaviour and tracks them over time, which is how you study retention. A segment overlap exploration shows how different user segments intersect. The way to remember them is by the question each answers: “where do people drop off” is a funnel, “what path did they take” is a path, “do groups who started together stick around” is a cohort. As a teaching example, to find where shoppers abandon checkout you would build a funnel exploration of the steps from cart to purchase and look at the biggest drop.
Matching the tool to the question
The overarching skill is choosing the right tool: standard reports for routine monitoring and quick answers, Explorations for deeper, custom investigation. A scenario that asks “where exactly are users dropping off in this multi-step process” is pointing at a funnel exploration, not a standard report, and recognising that is the kind of judgement the reporting area tests.
Chapter 5: Recognising the key measurement features
The fourth skill area is the features that turn analytics into a verdict on marketing: key events, audiences, attribution and data retention.
Key events (GA4’s conversions)
A key event is an event you mark as important to the business, such as a purchase or a lead-form submission, and it is GA4’s version of a conversion. The mechanism follows directly from the event-based model: because everything is an event, you create a conversion simply by designating an existing event as a key event, after which GA4 counts it in conversion-related reporting. The teaching point is that key events are not a separate object you build from scratch; they are ordinary events promoted to “this one matters”. Knowing how to mark a key event and find it in reports, and that the term replaced the older “conversion/goal” language, is what the assessment looks for.
Audiences and predictive audiences
An audience is a group of users defined by shared behaviour or attributes, used both for analysis and for remarketing in linked Google Ads. GA4 also offers predictive audiences, built using its predictive metrics such as purchase probability or churn probability, so you can target, for example, “users likely to purchase in the next seven days”. The lesson is to see audiences as reusable definitions of “people like this”, and predictive audiences as GA4 using machine learning on your data to anticipate behaviour rather than only describe the past.
Attribution and data retention
Attribution is how GA4 assigns credit for a conversion across the touchpoints in a user’s journey, and GA4 lets you choose an attribution model that decides how that credit is shared. You do not need to compute models by hand, but you should understand that the chosen model changes which channels appear to deserve credit, which in turn affects how you judge marketing. Data retention is the setting that controls how long GA4 keeps user-level and event-level data for use in Explorations (standard aggregated reports are not limited the same way). The practical teaching point is that retention settings affect how far back your custom analyses can reach, so it is a setting worth getting right early rather than discovering its limits later.
Chapter 6: Universal Analytics versus GA4, and why the difference matters
This short chapter exists because the single biggest source of wrong answers is studying the wrong product. Universal Analytics has been retired, and the certification is entirely GA4-based, so old habits actively mislead.
What does not carry across
In Universal Analytics, the session was the central unit, data arrived as distinct hit types, success was configured as goals, and bounce rate was a headline metric. In GA4, the event is the central unit, there are no hit types, success is a key event (a promoted event), and engagement is framed through engaged sessions and engagement rate rather than bounce rate. Views, a sub-level of UA properties, also work differently in GA4. The teaching instruction is simple: when you read older material, mentally flag UA-only vocabulary (hit types, goals, bounce-rate-as-headline) and replace it with the GA4 equivalent, or skip it. If a concept depends on the session being the atomic unit, it is a UA idea and does not apply.
Why Google rebuilt it this way
The event model exists to handle web and app together, to be more flexible about what you measure, and to fit a privacy landscape with less reliable cross-site tracking. You do not need the history for the assessment, but understanding the motivation helps the pieces cohere: the unified property, the predictive features and the event taxonomy are all consequences of building around events rather than sessions.
Chapter 7: Study plan, hands-on practice, and the assessment day
With the model understood, the work is pacing it and, above all, doing it in a real property.
Choose a timeline that fits your starting point
If you already use GA4 day to day, a few hours (roughly four to eight) reviewing the free Analytics Academy courses is usually enough: skim the familiar areas and concentrate on Explorations, key events and attribution, which are where working users most often have gaps. If you are new to Google Analytics, plan 15 to 25 hours over two to three weeks. A workable shape is days one to three on GA4 foundations and property setup; days four to six on the three event types, parameters and user properties; days seven to nine on standard reports and building Explorations; days ten to twelve on key events, audiences, attribution and data retention; and the final days on review and the concept checks. To turn that into dated days for your own start, use the free study-plan generator.
Learn GA4 by doing, not only reading
GA4 makes sense once you have clicked through it, so set up a free GA4 property on a personal or demo site, or use Google’s demo account, and practise each skill area in the real interface. Send a test event and watch it arrive in DebugView; mark an existing event as a key event and find it in reports; build a free-form and a funnel exploration; create an audience. Doing this once turns abstract terms like “parameter” or “funnel exploration” into things you have actually built. Work through the Analytics Academy courses alongside the hands-on practice and re-do the concept checks, looking up anything you miss in the official Help articles. Avoid third-party “answers” sites: they breach Google’s policy, teach you nothing usable, and are often out of date because GA4 changes so often.
Final review and the day itself
In your last session, confirm you can explain the difference between automatically collected, enhanced-measurement and custom events; tell a key event from an ordinary one; say what each exploration type is for; and name what changed from Universal Analytics. On the day you sit the assessment online on Skillshop with no proctor, needing 80% or more to pass; if you fall short you wait 24 hours and retake it free, with no limit on attempts. Because it is free with unlimited retakes there is no penalty for sitting it before you feel flawless, but aim to be comfortable across all four skill areas first so you pass cleanly and actually learn the platform. Remember it is valid for one year, so plan to refresh and retake annually.