This week, Google made a series of announcements about Google Analytics, including that App + Web properties would fold into a new version of GA called “Google Analytics 4”.
Google does not often release new versions of Google Analytics: “Universal Analytics” – the current version, now implicitly referred to as “GA 3” or ”GA classic” – was released eight years ago, in October 2012, so it’s worth taking a small break to get familiar with what this announcement really means.
Google Analytics needed a refresh to keep up with marketer needs
Over the years, the need for Google to redefine their Analytics solution has become more pressing as both the market and the user have evolved. The times when we focused our analysis on sessions and “last click” attribution are over, and we have moved to a user-centric approach with a focus on behavior.
Websites nowadays are more like applications within a web browser. A single web page can convey a large variety of user interactions, not to mention Single Page View sites (SPV) where everything happens in the same “page”. In this context, the analytics model had to shift from Hits, Pages, and Sessions to Events to stay relevant and allow for finer understanding and analysis of online user behaviors.
Two years ago, Google introduced a GA360-only feature called “Analysis”. This was obviously a first attempt to offer new tools to gain more insights into the customer journey. But in order to really move to a user-centric approach, Google has had to fully redesign the way it processes navigation data and the kind of metrics it creates.
In October 2014, Google also acquired Firebase, a backend service for iOS and Android app developers that helps to collect and sync data instantly. Within the Firebase platform, Google was especially interested in Firebase Analytics, a solution that has greatly inspired the design of App + Web Beta that we have become more familiar with since it became a public beta last year.
A new data model based on Events
If there is one take-away from this new paradigm, it would be that the new Analytics is now all about events, coupled with user audiences. Inspired by Firebase and App + Web Beta, the new Analytics has rendered Universal Analytics pages or sessions obsolete.
Of course, in this new model, there are standard events, custom events, and more possibilities like client-side rules for “editable events”.
This new unified data model, built around events and users, allows for easy integration of websites and mobile apps data streams. However, this comes with a price: you won’t be able to compare your current Google Analytics property numbers with the ones from your new Google Analytics 4 properties.
Thus, it is crucial that you consider implementing a dual tracking setup on your UA and GA360 properties today, in order to start collecting historical data in your new Google Analytics 4 properties.
What benefits should you expect from Google Analytics 4?
From a business standpoint, the three main advantages highlighted by Google are the following:
- Built with the power of Google’s machine learning at its core, the new Analytics will offer capabilities to automatically reveal business insights
- Bridging the gap between desktop and mobile, it will allow for a deeper understanding of the customer journey across devices
- Designed to be future-proof, it will work regardless of browser cookies or identifiers
The most obvious change in Google Analytics 4 is the integration of website, mobile web and app data, as well as IoT devices in a single data model. But this is far from being the only interesting aspect offered by the new Analytics. Among all newly-introduced exciting features, we would like to highlight three.
Data collection flexibility
Enabling to declare and fine tune events directly from the new Analytics UI, without having to write custom tracking code and modifying the implementation. Another exciting feature introduced with Google Analytics 4 is the Audience Triggers, which offers a high degree of flexibility by giving the possibility to fine tune complex custom metrics. In general, GA4 will provide solutions towards higher data quality.
Machine learning-based insights & activation
The new Analytics will introduce predictive metrics such as churn probability, conversion probability and also predictive revenue. Combined with audiences, predictive revenue will enable, for instance, the creation of dynamic segments to place adjusted bids on mobile app install campaigns. It will also help reach untapped high value-potential customers through look-alike modeling.
New lifecycle reporting
Because of the major change in Google Analytics around consumer and behavior, the new lifecycle reporting derived from Analysis is a reorganized and simplified collection of predefined reports, including advanced ecommerce reports. But beyond predefined reports, the new Analytics makes it easier for customers to conveniently create custom reports, save and share reports.
As a matter of fact, Analysis, which was already available in GA360 classic is now out of beta and offers 4 additional types of reports: ‘ Cohort’ and ‘Path analysis’, ‘User lifetime’, ‘User explorer’.
With the new Analytics becoming the default experience for new analytics properties moving forward, Google Analytics 4 is where Google will be focusing all their efforts. However, Google has made no explicit announcement yet about immediate plans to deprecate GA3 (Universal Analytics).
Rather than this, Google highlights a staged approach, as it is going to take some time for clients with complex implementations to adapt their tracking to this new model. Google’s advice is to create new Google Analytics 4 alongside your existing GA3 properties, to benefit from the latest innovations as they become available, while also keeping your existing implementation intact. On the web, it is possible to do so without additional developer integrations if a TMS has been set up.
At fifty-five, we are obviously excited by this announcement: we are convinced that Google Analytics 4 will bring an unprecedented flexibility in measurement and analysis, and allow for a future-proof 1st party data collection.