Google Analytics 4 represents a fundamental shift from Universal Analytics, moving to event-based tracking, privacy-first architecture, and predictive metrics. Understanding its data model, interface navigation, and core reporting differences is essential before you configure your first property or interpret any traffic patterns.
Universal Analytics organized data around sessions and pageviews, treating each website visit as a container with a beginning, middle, and end. GA4 abandons this structure entirely. Every interaction is an event with parameters. A page view is an event. A scroll is an event. A form submission is an event. This shift solves a longstanding problem: users move fluidly between devices and channels, so session-based measurement artificially fragments their journey. GA4 attempts to stitch these fragments together using Google signals, User-ID, and device-based identifiers when cookies aren't available. For beginners, this means reports won't look like what you expect from tutorials written for the old platform. The interface prioritizes user-centric metrics over session-centric ones. You'll see engaged sessions instead of bounce rate by default. You'll configure events instead of goals. You'll define audiences based on event sequences rather than simple page-rule conditions. Learning GA4 basics requires unlearning assumptions from Universal Analytics if you've used it before.
A GA4 property can collect data from websites, iOS apps, and Android apps simultaneously through data streams. Each stream has its own measurement ID that you embed in your site or app code. For a standard WordPress or Shopify site, you add the measurement ID to your tag manager or directly into header code. The property itself sits above the streams and unifies their data into one reporting view. When you create a new property in your Google Analytics account, you'll choose your reporting time zone and currency. Canadian businesses should select the appropriate time zone for their primary market and CAD as currency if e-commerce tracking matters. After creating the property, you add a web data stream by entering your primary domain and enabling enhanced measurement, which auto-tracks scrolls, outbound clicks, site search, video engagement, and file downloads without custom code. This automatic collection is a major change from Universal Analytics where most of these required manual event setup. Verify data is flowing by checking the real-time report within a few minutes of installation.
GA4 organizes standard reports into four Life Cycle categories. Acquisition shows how users first discover your site, broken into traffic source, medium, and campaign dimensions. Engagement covers what users do once they arrive: pages and screens they view, events they trigger, and conversions they complete. Monetization tracks e-commerce transactions, in-app purchases, and publisher ad revenue. Retention measures how many users return over time and their lifetime value. This structure reflects a customer-journey mindset rather than siloed channel analysis. Within each category, you drill down into more specific views. The Engagement overview shows active users, average engagement time, and your top events. Event detail reports let you see parameters attached to each event type. Conversions are simply events you've marked as important by toggling them to key events in the events list. The interface defaults tocard-based layouts with line graphs and summary metrics. Many beginners find the lack of table-first views disorienting at first, but tables appear when you click into detail reports or use the Explorations workspace for deeper analysis.
Every action in GA4 is an event with a name and optional parameters. Automatically collected events include page_view, scroll, click, file_download, and video_start when enhanced measurement is on. Recommended events follow Google's naming conventions for e-commerce and other common actions: purchase, add_to_cart, sign_up, login. You can also create custom events for business-specific interactions. Each event can carry up to 25 parameters. For a form_submit event, you might pass parameters like form_name, form_location, and user_type. Parameters become custom dimensions when you register them in the property settings, which makes them available as breakdown dimensions in reports. GA4 limits you to 50 custom dimensions and 50 custom metrics per property, so plan what you actually need before registering everything. Unlike Universal Analytics where you configured dimensions at property level and then sent data, GA4 lets you collect parameters first and register them as dimensions later without losing historical data, though you can only query them in Explorations after registration.
Standard reports in GA4 provide overviews and light segmentation. For genuine analysis, you use Explorations, which replace the custom report builder from Universal Analytics with analysis templates. Free-form exploration gives you a drag-and-drop table interface where you add dimensions as rows, metrics as values, and segments or filters to narrow scope. Funnel exploration visualizes step-by-step conversion flows, showing drop-off rates between stages. Path exploration reveals the actual routes users take through your site, either forward from a starting point or backward from a conversion. Segment overlap shows how many users belong to multiple audience segments simultaneously. Cohort analysis groups users by acquisition date and tracks their behavior over subsequent weeks. Each exploration lives as a separate analysis, saved in the workspace accessible from the left sidebar. You can share explorations with other users who have access to the property. For beginners, start with free-form to understand how dimensions and metrics combine, then move to funnel exploration once you've identified key conversion paths worth optimizing.
GA4 does not have goals in the Universal Analytics sense. Instead, you mark specific events as conversions or key events, terms Google uses interchangeably now. Any event can become a key event by toggling a switch in the events list under Admin settings. Common examples include purchase, generate_lead, sign_up, and custom events like demo_request or pdf_download. Once marked, these events appear in conversion reports and can be used as optimization targets in Google Ads campaigns. You can have up to 30 key events per property. Unlike goals, which had a fixed structure of destination, duration, pages-per-session, or event goals, GA4's approach is more flexible because every interaction is already an event. If you want to track micro-conversions like video completions or scroll depth, you ensure those events fire correctly through enhanced measurement or custom tags, then toggle them as key events. The attribution models for these conversions default to data-driven attribution when enough conversion volume exists, otherwise falling back to last-click. You configure attribution settings at the property level, not per conversion.
GA4 was built for a privacy-restricted future where third-party cookies decline and regulations like PIPEDA in Canada or GDPR in Europe require explicit user consent. The platform uses modeling to fill gaps when user-level data isn't available, estimating behavior based on observable patterns from users who did consent. For Canadian sites, you should implement a consent management platform that integrates with Google tag manager, allowing GA4 tags to fire only after users accept analytics cookies. GA4's default data retention for user-level and event-level data is two months, though you can extend it to 14 months in property settings. After retention expires, aggregated reporting data remains, but you lose the ability to create new audiences or run Explorations on that historical user data. Standard reports show aggregated data indefinitely. Beginners often overlook retention settings and then wonder why their Exploration queries return limited historical results. Set retention to 14 months unless you have specific compliance reasons to keep it shorter, and understand that even 14 months is not indefinite storage like Universal Analytics provided.
Yes, and this was recommended during the transition period before Universal Analytics stopped processing data in July 2023. You would have both tracking codes on your site simultaneously, sending data to separate properties. Now that Universal Analytics is retired, new implementations should use only GA4. If you still have historical Universal Analytics data, it remains accessible in read-only mode, but you cannot collect new data there.
GA4 and Universal Analytics are completely separate properties, so your old goals do not transfer automatically. You need to recreate the equivalent tracking by ensuring the relevant events fire in GA4, then marking those events as key events. For example, a destination goal for a thank-you page becomes a page_view event with a page_location parameter, which you can then toggle as a conversion.
Bounce rate exists in GA4 but is calculated differently and not shown by default in most reports. It represents the inverse of engagement rate: sessions that lasted less than 10 seconds, had no conversions, and included no subsequent page views. You can add bounce rate as a metric in Explorations or custom report cards. GA4 emphasizes engaged sessions instead, which measures sessions lasting longer than 10 seconds, having a conversion, or including multiple page views.
No, you can add the GA4 measurement ID directly to your site's header code and rely on enhanced measurement for most standard tracking. However, Google Tag Manager makes custom event tracking, parameter passing, and consent management significantly easier to implement and maintain without editing site code each time. For beginners learning GA4 basics, starting with direct installation is fine, but as you grow into custom events and advanced configurations, Tag Manager becomes nearly essential.
GA4 and Universal Analytics measure differently at a fundamental level. Sessions are defined differently, users are counted differently due to cross-device tracking, and metrics like engaged sessions replace bounce rate. Sampled data thresholds differ. Attribution models default to data-driven rather than last-click. Expect numbers to diverge by 10-30 percent in many cases, not because one is wrong but because the measurement methodology changed. Focus on trends within GA4 rather than comparing absolute values across platforms.
Install the tracking code and verify data appears in real-time reports. Enable enhanced measurement if not already on. Identify your three to five most important business actions and ensure those events are firing correctly, then mark them as key events. Create a basic audience for returning users. Link your Google Ads account if you run paid search. Set data retention to 14 months. Build one simple funnel exploration for your primary conversion path. These steps give you a functional baseline to learn from actual data rather than hypothetical configurations.