UA 4 Property Migration – Begin the journey

Google Analytics 4 will be taking over from Universal Analytics. As of July 1, 2023 (tomorrow), regular Universal Analytics properties will no longer handle data. (360 Universal Analytics properties that have been ordered will receive a special extension for data processing, which will end on July 1, 2024.) If you don’t migrate using the migration tools and you didn’t opt out of migrating certain properties, it will migrate everything for you. If you are not doing a lot of properties and advanced events it should move everything over (the basics) for you. If you are reading this article there is a great chance you didn’t migrate and you need some help updating the automatically migrated properties.

Property Comparisons

Lets talk a bit about the differences in properties and their related events and facets between both versions of Analytics.

Property Setup

In Universal Analytics, the data hierarchy is organized into Account > Property > View. For each website or app, you would generally set up a separate property. For example, if you have a business with a website and a mobile app, you would set up one property for the website and another for the mobile app. Each property has its unique tracking ID (formatted as UA-XXXXXX-Y) that is used in your tracking code.

In Universal Analytics, each property would then have one or more ‘views’ that allow you to segment and filter the raw data from your property in various ways. For instance, you could have an unfiltered view that includes all data, a view that only includes traffic from a specific country, and a view that excludes internal traffic.

In contrast, Google Analytics 4 simplifies this hierarchy by adopting a dual model, Account > Property > Data Stream. In GA4, you can set up one property to encompass both your website and mobile app. You can link multiple ‘data streams’ (websites, iOS apps, Android apps) to a single property, and all the data will be collected together. This makes it much easier to see user interactions across different platforms and perform a cross-platform analysis.

So, if you have a website and a mobile app, you can now have one GA4 property with two data streams – one for the website and one for the app. Each data stream in GA4 has its unique Measurement ID, similar to the tracking ID in Universal Analytics, that you include in your tracking setup.

This unified approach in GA4 is designed to provide a more holistic, customer-centric view of how users interact with your business across different platforms. This setup can potentially offer richer insights, especially for businesses with a significant presence on multiple platforms.

User Identification

Google Analytics 4 introduces the concept of the Google Analytics 4 User ID. This user-level identifier, similarly to the User ID in Universal Analytics, has to be generated by you and should be unique for each user. The GA4 User ID is particularly useful for businesses that have users logging into their websites or apps as it allows you to unify a user’s interactions across different devices and sessions.

In GA4, the Client ID still exists, but it’s named ‘Device ID’. This is automatically assigned by Google and used to identify unique devices. It is stored in browser cookies or App Instance ID in mobile apps.

One significant update in GA4 is that it gives a more straightforward approach to stitching together user behavior from logged-in and anonymous sessions. If a user was previously anonymous but later logs in, GA4 can link the previous anonymous activity with the same User ID. This results in a more complete user journey, assuming that the same User ID is assigned to the user across devices.

Keep in mind, though, that the handling of User IDs, especially where they might link to specific individuals, must be compliant with all relevant privacy regulations and Google’s own policies.

Data Retention

In Universal Analytics, user-level and event-level data is stored for 26 months by default. This means that raw, hit-level data that allows you to drill down to the actions of individual users is kept for 26 months. After that period, the data is deleted.

However, Universal Analytics allows you to adjust this setting to a custom period or even to not automatically expire at all. Also, the data retention period is reset with each new event from the user, so it’s not purely a case of data being 26 months old and then disappearing. It’s 26 months after the user’s last activity.

Aggregate data (for example, the total number of users per day) is not affected by this setting and is retained indefinitely. Mean while GA4 takes a similar but slightly different approach to data retention. It stores user-level data, along with event data that includes conversion events, for 14 months by default. But you can modify this period, from as short as 2 months up to a maximum of 30 months.

The important difference here is that the retention period in GA4 is not reset with new user activity. This means that after the selected retention period, the event data is deleted regardless of whether the user had any new activity or not.

It’s worth noting that GA4’s retention settings apply to event data tied to a user identifier (like User ID or Device ID). Non-identified event data is stored for 2 months and cannot be changed.

In both versions, it’s crucial to select a data retention period that aligns with your data analysis needs while also considering privacy regulations, as longer data retention periods may have implications for user privacy.

Cross-Domain Tracking

Cross-domain tracking in Universal Analytics is used when you want to track user behavior across multiple related domains. For example, if you have an e-commerce site where the shopping cart is on a different domain than the main store, you’d use cross-domain tracking to understand the user’s complete journey.

Setting up cross-domain tracking in Universal Analytics can be quite complex. It involves modifying the Google Analytics tracking code on each page of your website where users might move between domains. You’d have to use the ‘autoLink’ plugin in the analytics.js library and list all the domains you want to track.

Once set up, Universal Analytics can track users seamlessly as they move between your domains. It does this by appending a unique ID parameter to the URL when a user moves from one domain to another, allowing it to understand that it’s the same user on a different domain.

In GA4, the process of setting up cross-domain tracking is simplified compared to Universal Analytics. It doesn’t require the complex code modifications that UA does. Instead, in the GA4 property’s settings, you can list the domains you want to track.

GA4 can then automatically keep track of users as they navigate from one domain to another, treating it as a single session. Similar to UA, it appends a unique identifier parameter to the URL as the user navigates between domains.

The simplified setup process in GA4 is a significant advantage if you regularly need to track user behavior across multiple domains.

Remember, cross-domain tracking requires handling user data across various domains, and appropriate care should be taken to ensure that data privacy and protection measures are in place, complying with relevant regulations and Google’s policies.

Event Tracking

In Universal Analytics, events are user interactions with content that can be tracked independently from a web page or a screen load. These include downloads, mobile ad clicks, gadgets, Flash elements, AJAX embedded elements, and video plays.

When you set up Event Tracking in UA, you typically structure it in terms of Categories, Actions, and Labels. This hierarchy allows you to organize events in a way that makes sense for your particular website or app.

For instance, if you’re tracking a video play, you could use “Videos” as the category, “Play” as the action, and the name of the video as the label.

Event Tracking in GA4 is more flexible and less rigidly structured. Every user interaction that you want to track is an event, including page views. GA4 comes with a set of recommended events that you can use, but you’re also free to create your own custom events.

GA4’s events are basically flat, as opposed to the hierarchical structure of UA. There are no categories, actions, or labels, but instead, each event has parameters. These parameters can hold additional information about the event.

For example, if you’re tracking a video play in GA4, “video_start” could be the event, and then you could have parameters like “video_name” and “video_duration”.

In GA4, you can also mark certain events as “conversions”. This means that you consider those events to be important and want to focus on them. GA4 will then use these conversions in its analysis and machine learning models.

While both UA and GA4 offer Event Tracking, GA4’s approach is more flexible. In UA, you have to fit your events into the category-action-label structure, while in GA4, you can create flat events with parameters that make sense for your specific needs.

Moreover, GA4’s capability to designate certain events as conversions brings an additional layer of focus to important user interactions, contributing to a more comprehensive analysis of your data.

Custom Dimensions / Metrics

In Universal Analytics, a custom dimension is a way to collect and analyze data that Google Analytics doesn’t automatically track. For example, you might use a custom dimension to track the performance of a specific author or category on a blog.

A custom metric, on the other hand, is a way to measure the number of times a specific event happens. For instance, you could create a custom metric to count the number of times a particular button is clicked.

Both custom dimensions and metrics are set up at the property level in Universal Analytics, and you can create up to 20 custom dimensions and 20 custom metrics for each property (or even more if you have Google Analytics 360).

In Google Analytics 4, the concepts of custom dimensions and metrics are replaced by ‘custom definitions’.

Custom definitions in GA4 are essentially the same as custom dimensions and metrics in Universal Analytics, but there is a fundamental difference in the setup. In GA4, you don’t predefine custom dimensions and metrics at the property level. Instead, you collect data through event parameters, and then you can register these parameters as custom definitions in GA4.

Once a parameter has been registered as a custom definition, it becomes available in your reports like a built-in parameter. You can then use these custom definitions in the same way you’d use predefined dimensions and metrics.

An important note here is that GA4 allows up to 50 custom definitions, both for dimensions and metrics combined, per property.

It’s also worth mentioning that GA4 introduces ‘user properties’, which function similarly to user-level custom dimensions in Universal Analytics. They can be used to describe attributes of users or their interactions, such as ‘loyalty status’ or ‘last purchase category’.

User Properties

In Universal Analytics, there is no concept exactly equivalent to GA4’s ‘user properties’. However, Universal Analytics did allow for the creation of ‘custom dimensions’ which could be used to collect and analyze data that Google Analytics doesn’t automatically track.

For instance, you could use custom dimensions to track user-level data, like membership status (member or non-member), user preferences, or any other data that could be tied to a user for the duration of multiple sessions.

GA4 introduces ‘user properties’, which are attributes that you can associate with users. These properties provide additional information about your users, and you can use them to further segment your audience. For example, you can use user properties to track data such as a user’s profession, their level in a game, or their subscription tier.

User properties can hold string or number values, and you can set up to 25 unique user properties per project. They offer a more streamlined way of analyzing user behavior as you can associate specific attributes with users and track these over time.

Once a user property is set, it is attached to the user for all subsequent events and can be used in audience definitions and report filters. For instance, if you have an e-commerce app, you could create a user property for ‘frequent_buyer’. Once you set that property for a user, all events that the user triggers will be recorded with that user property, enabling you to filter reports and create audiences of ‘frequent buyers’.

User properties in GA4 are a flexible tool that can provide deeper insights into user behavior, but they must be used responsibly to respect privacy laws and regulations.

Views

In Universal Analytics, the data hierarchy consists of the Account, Property, and View levels. Each property can have one or more views. Views are where you can access reports, and they provide a way to segment your data within each property.

For example, you could have a raw, unfiltered view that includes all of your data, another view that filters out internal traffic, and another that only includes traffic from a specific country or source. Views give you the flexibility to customize what data you see and how it’s structured in your reports.

You can set up filters at the view level in Universal Analytics, allowing you to include or exclude specific subsets of your data. This is useful for segmenting your data and can help you gain specific insights about different aspects of your traffic. Once a view is created, it only includes data from that point forward, and historical data is not available in the new view.

In GA4, Google made a significant change: the concept of views has been removed entirely. Instead of creating different views within a property, you filter data within the reports themselves in GA4. This is done through a feature known as ‘Comparisons’ in GA4, which allows you to segment your data in your reports. This new approach offers more flexibility in analyzing data, as you don’t need to set up and manage multiple views.

The GA4 model provides a more flexible and comprehensive way to understand user behavior. Instead of segmenting data into different views at the property level, you can analyze and compare different segments directly within your reports, making it easier to derive insights about your users’ behavior.

However, it’s important to note that filters in GA4 are not the same as views in UA. In GA4, filters affect the data that’s collected in the property, not just what’s visible in the reports. Therefore, when you set a filter in GA4, it permanently affects the data that is stored and can’t be undone. This is a significant difference from UA, where filters at the view level only affected what data was visible in that view, and other views could still access the full, unfiltered data.