Understanding Google Analytics 4

1. Anomaly detection

Security features are positioned at the forefront of the GA4 platform. GA4 uses machine learning anomaly detection to detect “unusual behaviour” within the data being processed on the platform. Anomaly detection aims to drive decisions to audit, investigate or block anomalies.

As online transactions become more common, criminal activity and malicious attempts can arise. Anomaly detection is an excellent security feature on GA4, significantly reducing the risk of false transitions and fraud. As a result, your time can be spent more effectively, fraud can be detected more easily, and losses can be reduced.

Anomaly detection offers value in a second way; it can be used to forecast future occurrences. To avail of the machine learning capabilities, you will need historical data in your GA4 account to “train” Google to predict unusual data behaviour.

To detect anomalies, GA4 will need a training period of:

• Up to 2 weeks (to predict hourly anomalies)
• Up to 90 days (to predict daily anomalies)
• Up to 32 weeks to learn to predict (to predict weekly anomalies)

Once you adjust the anomaly detection settings, hover over the data in the exploration report; anomaly data is highlighted by an empty circle. The anomaly information will show if you hover your cursor over the empty circle.

2. BigQuery connection

GA4 provides a free connection to BigQuery, a cloud data storage and management system, where you can export all your data from GA4 property to BigQuery. Unfortunately, you cannot directly transfer data from UA to GA4. (If you have GA360, you can transfer directly over to GA4.) Therefore, you will need to manually upload to BigQuery before importing the data into GA4. Once exported from GA4, an analytic table is generated to include important data such as the event date, timestamp, event name, and event parameters. GA4 BigQuery’s free tier usage limit is up to 10 GB of active storage and 1 TB of processed data each month. If you need to store or process more data, you must upgrade your tier.

3. Combined app & web view

GA4 no longer requires separate platforms to collect app and web property data. GA4 simplifies your data collection efforts by allowing you to view data from your website and app on one platform. With this feature, you can understand your user journey and touchpoints across devices, giving you the information needed to build better user journeys and experiences. Combined app and web view can help you recognise your customer’s intent, preference, and rationale for their behaviour.

When you add an app data stream, GA4 splits the stream for iOS and Android, helping you get more specific data per device type.

4. DebugView

The DebugView in GA4 provides a more advanced look at the platform. You can see incoming data as GA4 collects it. In DebugView, you can view events at different time intervals. You can also see the minutes stream, top events, current user properties, and a device selector.

You can view events at different time intervals

• Events are logged from 60 seconds.

• Top events are logged from the last 30 mins.

When you add an app data stream, GA4 splits the stream for iOS and Android, helping you get more specific data per device type.

5. Event Based Model

GA4 has shifted from a session-based model that tracked ‘hits’ to an event model that tracks ‘events’. The difference is that hits such as page views, events, social, and screen views will now be tracked as events. The event-based model focuses heavily on how people interact with your site, so keep this in mind when setting up events.

6. Google Tag Manager

GTM allows you to create custom tags specific to your goals on GA4, aiding you in tracking any additional activity not prebuilt on GA4. GTM is more advanced and may require working with a developer to install tags (types of code) onto your website and GA4.

For example:

• Click ID
• Form ID

Once you set up the Google Tag Manager code snippet on your website, GA4 will begin to collect these tags, which you can review in the DebugView report.

7. Google signals

In GA4, Google signals are session data that measure users across devices. Google signals are for users who have turned on Ads personalisation within their Google account. This feature allows you to create highly relevant ads based on user preferences.

If you enable Google signals and don’t change the reported identity to the device, GA4 will start applying thresholds to dimensions will low user count.

8. Predictive machine learning

GA4 employs today’s most advanced analytics, running largely on machine learning technologies. Just like anomaly detection, GA4 uses predictive analytics for building suggested audiences and predicting future buying behaviour. With predictive analytics, you can efficiently invest in retaining customers as the machine learning algorithm is capable of forecast patterns.

Predictive machine learning analytics can be located when building in the suggested audience or the exploration tab. Predictive analytics delivers value in a second way, by helping you focus your resources and efforts on the most valuable and promising set of users.

GA4 can forecast future predictions such as;

9. Improved privacy controls

As the Internet becomes more privacy-centric, Google wants to offer users control over sharing their activity with Google Analytics. In addition, the introduction of data protection laws such as GDPR and CCPA has further impacted how websites handle personal information.

GA4 sets out several improvements in its privacy controls:

10. Insights & recommendations

GA4 offers tailored insights and recommendations to help you get the most out of analytics. You can choose whether to accept or dismiss recommendations as you wish. Recommendations come and go depending on whether it has been applied already, dismissed, or if the recommendation is no longer significant. An additional benefit is that you can create and manage your insights, set conditions, and get notified by email when a custom insight is triggered.

To view the insight in more detail, click on the card.

11. Reporting data

GA4 reports are more customisable, improved, and actionable than previous analytic versions. In addition, you may see different tabs available depending on your GA4 account type. For example, eCommerce or gaming accounts may see other tabs available.

Reporting snapshot

Contains a general overview of your most important data. You can modify what appears on the snapshot page to your liking.

12. Explore

Explore allows you to create your own custom report and interpret the data in your own way. In comparison to GA4 predecessors, exploration reports offer more flexibility. You have the choice of pre-made template techniques to help generate the report. Once you choose the technique, your custom report is split into variables, tab settings, and the custom report. Once you input the data you want to appear in the report, you can change the report’s visualisation type from the standard table to scatter plot, geo map, and many more.

You can select the type of exploration technique you want: