Gain a competitive edge by following 7 key steps, beginning with Account Structure Review
Google put a hard deadline on the shift to it’s next-generation Google Analytics 4 (GA4) with the announcement that it will sunset Universal Analytics in July 2023. There remains some debate around when Google will truly hit the “off switch” on UA and UA 360, but whether it’s 15, 21 or 24 months, it’s clear that every business using UA needs to begin planning for the migration to GA4.
Moreover, businesses can gain a competitive edge by accelerating their move to GA4 ahead of the sunset deadlines — mastering GA4’s next-generation data-driven marketing capabilities
A 7-step guide to GA4 migration
This is the first in a seven-part blog series that will walk through key steps to ensuring successful migration to GA4:
Step 1: Review account and property structure
Step 2: Set up and configure a GA4 property
Step 3: Build a measurement plan and implement
Step 4: Configure linking to GMP products
Step 5: Migrate other Universal Analytics features
Step 6: Export historical data
Step 7: Replicate current reporting
Step 1: Account Structure Review
Creating a new GA4 instance requires planning the Google Analytics architecture and revisiting account structure if needed. The goal here is to keep everything in one account — under ownership by a single legal entity.
GA4 Properties
In most cases, you’ll want to keep your new GA4 property in the same account with your existing UA property. A more advanced scenario for enterprise GA4 customers would be to keep all GA4 properties in the same account if they need to be aggregated in a single roll-up property (for UA 360 only).
You’ll most likely want to mirror the setup of your UA instance, keeping the same number of new GA4 properties. Generally speaking, data that spans across the same user base or a single user journey — or any data that should be analyzed together — should stream into one GA4 property. For website tracking, this mostly means a single domain. But if a user journey spans across multiple domains, each domain should send data to the same property.
Google Ads Integration
A property is also a unit for integration with other products (e.g., Google Ads). Once you set up the linking, the data from a Google Ads account will stream into the property, with no ability to keep just a subset of the data and export it partially to a different GA4 property. So, having 1-to-1 correspondence between Google Ads and GA4 will generally be best practice.
Firebase Integration
A GA4 property is also a basis for Firebase integration, as only one Firebase project can be linked to a single GA4 profile. Therefore, if you have multiple Firebase projects, you will need to link them to different GA4 properties or otherwise restructure them for integration with GA4.
Having an existing Firebase project linked to GA4 may impact your decision on keeping the existing GA4 property. For example, Firebase could have been previously connected to Google Analytics for Firebase, which then migrated to Google Analytics App+Web, and after that rebranded to GA4. Therefore, if you have a Firebase project for your app, you may want to first check if there is an existing GA4 property containing the app data.
If your app(s) and website(s) have the same userbase or are connected to the same CRM system, you should consider connecting them to the same property. In this case, best practice is to use the existing GA4 property with the app data — and then start streaming website data to this existing GA4 property.
Unfortunately, even if you’d like to keep two properties — one for the app only and one for app + web — there is not an easy way of doing so, since a single Firebase project can only be linked to one GA4 property.
The decision-making process with Firebase is visualized in the image below:
Data streams
The next level of the GA4 hierarchy is the data stream. Normally, you have one web data stream, one iOS and one Android stream in a property. If you have several apps linked to your Firebase, they will appear as different data streams in GA4.
Using one data stream for several domains is required to set up cross-domain measurement between them. Having multiple streams is possible if you want to additionally segment your data using measurement IDs or stream IDs as distinguishing criteria.
Analyzing data from different environments separately is a typical scenario, so it’s expected that you have at least one property for test environments and another one for production.
Regarding data segmentation that was previously addressed using the views functionality in UA, there is no equivalent view in GA4. Instead, you will have to rely on different methods to achieve the same results, including:
- The Comparison functionality in standard reports, enabling data segmenetation by different parameters (e.g., by Hostname)
- The Explorations functionality, enabling custom reports for more complex segmentation needs;
- Internal and developer traffic filters to exclude unnecessary data from analysis;
- The Modify Events and Create Events features for miscellaneous modification of the collected data (comparable to the Search and Replace or Advanced filters in UA, but more limited);
- Subproperties (GA4 360 only) for data governance purposes, allowing user access to be limited to a specific subset of data
Once you define the structure of your GA4 property, the Administrator role is responsible for granting users access to this property. Users who previously had account-level access will inherit the same permissions to the new GA4 property; all others must be granted specific additional access privileges.
What’s next: GA4 Property Creation and Data Collection Enablement
Creating the framework or architecture for measurement in GA4 is just the first step. In our next installment in our seven-part series, we’ll dive into how to create properties in GA4, configure settings, and more. We encourage you to check out the full Google Analytics 4 Migration Guide here: