Amplitude returns to the main questions of professionals regarding the announced end of Google Universal Analytics and the migration to Google Analytics 4.
Last March, Google shook up the web world by announcing the end of Google Universal Analytics as of July 2023, encouraging its users to switch to Google Analytics 4. However, with more than 70% market share, Universal Analytics is the the world’s most widely used free web analytics and advertising platform. Its removal causes an earthquake and raises many questions among professionals. What are the characteristics of this new version? What can it bring more or less to marketers? How to consider migration? Overview of the challenges of this adoption.
Why is Google switching to Google Analytics 4?
The analytics industry has undergone many changes in recent years, marked in particular by a phenomenon of convergence of the web environment and the field of applications. In previous decades, companies relied on separate tools for marketing analytics – more often referred to as web analytics, experience analytics, and product analytics. With the rise of single-page web applications (SPAs) and cross-platform journeys, the traditional web analytics model, which relies primarily on tracking sessions and page views, is no longer suited to the needs of the business. While these metrics are still useful, most digital analytics vendors have gradually migrated to an event-based data model.
For its part, Google acquired the Firebase web and mobile application development platform in 2014, in order to offer Universal Analytics users analytics capabilities based on event data through the Apps + Web property. For marketers and developers, the transition to GA4 is therefore not a total surprise, but the logical continuation of Apps + Web, marking the evolution of digital analysis towards user logic centered on the behavior of clients.
GA4, a response to data privacy issues?
Users demand more and more transparency and control over their personal data, and in particular over their use for advertising personalization purposes. However, Google Analytics has often been singled out for its non-compliance with personal data protection rules and in particular the GDPR (General Data Protection Regulation). GA4 solves the problem by anonymizing IP addresses and limiting data retention to a maximum of 14 months. Thus, it is no longer possible to make the link between the data collected and the identity of the users, provided that the data sharing option is deactivated in the settings. However, after the formal notice of several organizations by the CNIL last February, due to the transfer of data to the United States – prohibited by the RGPD – legitimate concerns remain as to the use of certain specific functionalities of Google Analytics. .
This is particularly the case of Google Signals, which is a function offered by Google for Analytics, in order to identify anonymous visitors and enrich the data. To do this, Google Signals leverages other Google products (Chrome and Gmail, for example) with the ads personalization feature. Thus, the solution makes it possible to take advantage of the Google advertising network to perform cross-device user tracking, add demographic information such as age, gender and advertising points of interest, and share it anonymously with Google Analytics. Now, Signals can be disabled by admins in Google Analytics, but most organizations that use the platform have it enabled, and very few Google users know how to disable Ads Personalization in their account. Although the use of Signals is currently permitted in the GDPR, if all consent requirements are met, the scenario in which the European Union would force Google to remove this functionality or make it “opt-in”, seems increasingly more conceivable. If necessary, Google Analytics would lose added value, particularly in terms of data enrichment and therefore knowledge of the user.
Should we say goodbye to our historical data?
Although Google Analytics remains available until July 2023, companies should plan their migration strategy to GA4 now. An approach that can be complex, insofar as GA4 is based on a data architecture that is totally different from that of Google Analytics. So migrating can take as much time and resources as moving to a brand new analytics solution. Furthermore, the level of difficulty depends on the use cases, some of which involve specific steps, such as e-commerce tracking, and still have some compatibility issues. For companies with an advanced use of Google Analytics, the change logically involves more stakeholders and requires a structured plan. This notably involves an audit of existing reports and an assessment of the needs of each team, which will make it possible to choose the most appropriate technological solutions and to define a precise budget and schedule. Like any major technological change, this deadline also provides an opportunity to take stock of its digital analysis strategy in order to determine the optimal configuration to strengthen competitiveness and growth.
But one of the main concerns of users is the risk of losing historical data. They will be able to continue to collect and use new data in Google Universal Analytics for less than a year, and retain it for six months after the solution is discontinued. Thus, many organizations fear that they will no longer be able to use their historical data. But above all, they must ask themselves about their real usefulness and to what extent their business depends on them. If there is a real fascination for historical data, in reality, few companies really put it to use. Those for which this is the case generally already have in-house solutions for storing them or have already set up external means for storing them. However, it is essential to configure key properties in GA4 as early as possible in order to have a sufficient data backlog and limit loss at failover.
What about data quality and governance?
Most current Google Analytics users manage its implementation from Google Sheets which lists their data taxonomy. But in a context where event-based analytics becomes the gold standard, this can pose data governance issues. Businesses today need complete confidence in the data they use and activate for downstream applications. Events must be planned, instrumented, validated, organized, transformed and observed over time to generate high-quality information leading to smarter and faster decision-making.
Event-based analytics platforms require significant investments in data management, but are much more powerful and efficient when it comes to tracking and analyzing customer behavior. Without great governance tools, lower adoption rates due to unreliable data and re-instrumentation efforts can quickly become high. This is one of the main reasons why analytics strategies fail. Thus, GA4 and GA360 users are eager to benefit from features such as integrated follow-up planning, observability for event validation, a “developer-first” experience with for example Jira and SDK integration and stronger data transformation capabilities.