In 2020 Google finally announced the follow-up to Google Analytics Universal, Google Analytics 4. Their next-generation analytics solution is built on an event-based data model, which differs vastly from Universal Analytics. A question that arose quickly after the announcement was:

‘Can I migrate from Universal Analytics to Google Analytics 4?’

Technically, Google does offer some capabilities to do this, but since the underlying data models don’t match, it would be like trying to fit a square peg in a round hole. Ships have sunken due to such types of quick fixes. To keep your company’s data analysis ship afloat, I will always recommend a new implementation of Google Analytics 4. Regardless if you implement Google Analytics 4 through Google Tag Manager, or place code directly in your source, we have some tips to help you convince anyone of the urgency to start using it today.


ℹ Did you know that Google Analytics 4 works perfectly with Google Tag Manager Server Side? Read more.


No time like the present

The most common advice given today is to implement Google Analytics 4 as soon as you can, even if it is just in its basic form. The best practices seen today are to implement Google Analytics 4 in parallel to Universal Analytics. I tend to agree. Running both Google Analytics versions in parallel is great for comparison purposes in more ways than one.

If you find yourself stuck in a meeting with managers and peers, trying to convince them why you should implement Google Analytics 4, let me give you a few arguments:

1. New data model and new interface

It won’t be easy to compare Universal Analytics data to Google Analytics 4 due to the underlying data models. It will be futile if I can be brutally honest. When collecting data from your visitors, you need to keep in mind that everything will be event-based. Instead of the well-worn (and proven) User > Session > Pageview > Event model, all data points will now be events. Sessions will be derived from built-in session count metrics, and even pageviews are events. The more positive-sounding engaged session metric has replaced even the infamous bounce rate metric.

New metrics, new conversion points, almost everything about Google Analytics 4 is new. The user interface has also gotten a makeover not only in design but also in functionalities. New out-of-the-box reports create the perfect entry point into visualizing the new data model, but it will be the Analysis Hub that will truly be an eye-opener. Krista Seiden and Stefano Menti have a great video on this topic for you to watch.

For privacy reasons YouTube needs your permission to be loaded.
I Accept

2. Practice with your own data

Here in the Netherlands, when you learn to drive a car, you learn in the car of the driving school that you are attending. Not your own car. From experience, we can all relate to the fact that every car, even though the mechanics are the same, drives a little differently. If we translate this metaphor to Google Analytics, I can confidently say that switching from Google’s Universal Analytics to version 4 is a little more than switching cars. I would more easily compare it to switching from a combustion engine car to an electric vehicle. Not only are the driving characteristics different, so is the engine.

We have gotten very used to thinking in terms of Users > Sessions > Pageviews > Events, not to mention Event Category, Action, and Label 😉 It will take time to rethink our view of the data model when starting to work with Google Analytics 4. A big benefit to adopting the new event-based data model, in my opinion, would be to practice using your own data.

It will help reduce the time to understand the event-based data model when you can explore your data. Recognizing your own pages, events and products support familiarization in general.

3. Start replicating existing reports in Google Data Studio

The logical step from working with your own data is to replicate existing reports in Google Data Studio using Google Analytics 4. As stated before, when working with a different underlying engine that processes your data, don’t expect to exactly mimic all Universal Analytics reports using Google Analytics 4 data. It will not be a case of comparing apples and oranges, but more like Granny Smith apples 🍏 and Golden Delicious apples 🍎.

As soon as you start replicating your existing reports, you will quickly learn where the differences lie. These discoveries will help you and your team focus on how to explain the differences to stakeholders. In general, the differences in the data model will be the root cause of most discrepancies. New metrics can substitute these discrepancies. However, what I have run into personally, during several Google Analytics 4 implementations that I have done, is that switching from Universal Analytics is a great moment to reevaluate existing data requirements (based on Universal Analytics implementation) by:

  1. challenge stakeholders on existing/new data requirements
  2. match existing data requirements to implementation
  3. identify data gaps
  4. design new implementation plan
  5. implement the new plan with updated requirements

4. Start working with BigQuery

For some of you analysts out there, Google BigQuery is the holy grail for analysis. Access to raw data collected through the Google Analytics API from your website and/or mobile device, how good can it get? Well, for one thing, sending data from Universal Analytics to BigQuery was never straightforward. It required a middle layer that was either very dependable (expensive) or inconsistent (free). Google Analytics 4 has a pre-built Google BigQuery connector built into its administrator section. The cherry on top, it’s free.

Linking Google Analytics 4 to BigQuery

You have to remember that although populating BigQuery with Google Analytics 4 data is free, querying the data for your reports is not, so there is a ‘gotcha’. Querying is not expensive as long as you test on small dataset samples before querying all your data.

What makes Google BigQuery important for Google Analytics 4? So far, I have found that the API access to Google Analytics 4 reporting is limited. For instance, the creation of acquisition reports leaves a lot to be desired in terms of available dimensions. This is not the case with Google BigQuery, which for analysts should be a key motivational point for migrating to Google Analytics 4. In Google BigQuery, you will have access to all Google Analytics 4 dimensions and metrics. Integrate your query with Google Data Studio, and you will have unlimited possibilities in the visualization of your data.

Final Thoughts

Although implementing Google Analytics 4 is not too much of a challenge when using solutions like Google Tag Manager, there are some considerations to make. A big unknown factor is, for instance, when Universal Analytics will be decommissioned. This makes it difficult to determine the urgency with which you need to switch to Google Analytics 4. I think that because Google Analytics remains free, the original driver many companies chose for Google Analytics in the first place, it is wise to start using Google Analytics 4 as soon as you can.

If you already have a solid implementation of Universal Analytics using Google Tag Manager setting up Google Analytics to run in parallel won’t be too difficult. During this parallel period, give yourself time to get used to Google Analytics 4’s interface, but more importantly, its capabilities as a replacement analytics solution for Universal Analytics.

Once you have followed the above-mentioned steps and your company has transferred its data dependency and trust to Google Analytics 4, then you can pat yourself on the back for having completed a full migration. Migration is more than just code and conversions. It’s about your company’s data value creation ability. Let that be the first and foremost reason to migrate to Google Analytics 4 in the end.

If you would like to know more, please schedule a meeting with me using the button below.