Why Google Analytics 4 might just be a good thing

Organic Search
-

What is Google Analytics 4?

In a nutshell, GA4 is the next generation of Analytics, moving away from the session-based model in Universal Analytics to an events-based model (more on that later) and bringing sites and apps together within a single property.  It marks a shift in both measurement and reporting unlike any seen before and, from 1st July 2023, will be the only Google Analytics package available.  While Google themselves are at pains to represent this transition as a positive, many marketers and industry commentators have expressed a more negative outlook.

What’s everyone getting so worked up about?

For such a progressive bunch, we digital marketers are still human and change can be daunting.  I remember back in 2013 when some marketers expressed concern around the transition to Universal Analytics itself which, by comparison, was a minor adjustment, so some resistance ought to be expected.

However, the industry has raised a number of legitimate concerns, with four recurring main complaints:

  1. “It’s too complex to get everything set up correctly”

At the outset, it’s clear that migrating from Universal Analytics to GA4 is not a seamless process.  This is a whole new approach to measurement and reporting so you’re effectively faced with a blank canvas.  Even Google’s own support pages recognise that most users will need to input medium to high effort to make the switch, and that’s before you’ve even thought about tracking any conversions.

In an ideal world, it would be a simple case of flicking a switch and your data, tags and custom dimensions would appear in your new GA4 property.  Whilst that isn’t the case, thankfully there are already helpful guides for migration and setting up eCommerce available.  Nevertheless, getting your data tracked and recorded correctly is likely to be a lengthy and painful project.

  1. “The UI is less intuitive”

The new user interface layout is much cleaner in terms of appearance, but this brings its own challenges.  For users who are familiar with the UA layout, it can be trickier to unearth some of the reports and insights you regularly need; and there are some aspects of the report builder which appear to be less flexible.  Drilling down into the detail is not as straightforward as it once was, either.

Some may also lament the loss of Views, denying them an area to test and clean up your data.  It should be noted that this is still a ‘work in progress’, and some of the criticism is possibly the result of users not wanting to be taken outside of their comfort zone, but there will undoubtedly be a ‘bedding-in’ period before you can benefit from the full power of GA4.

  1. “Some standard reports and metrics seem to be missing”

As it is based on a different data model, a number of familiar reports and metrics are no more.  On the metrics side, Bounce Rate is no longer provided which arguably limits insight into which landing pages best serve your users’ needs (and subsequently which pages require more attention).  Google instead considers ‘Engaged Sessions’, which can potentially be used as a proxy, but the change in methodology and calculation will lead to some continuity issues if this has been a core aspect of your business to date.

Likewise, there seems to have been a shift in focus when it comes to reporting.  Universal Analytics was fundamentally an exhaustive collection of off-the-shelf reports which enabled in-platform analysis.  GA4, on the other hand, features fewer ready-made reports, instead allowing for greater personalisation.  Whilst this is potentially empowering for marketers and businesses with specific needs, it does pose another headache for users who relied on some of those standard reports for their day-to-day analysis. Whilst it is possible to replicate some of the “missing” reports by customising the reports which are there, this adds yet another layer of difficulty in setup.

  1. “We’re going to lose valuable historical data”

Google has confirmed that your Universal Analytics property will stop processing new data on 1st July 2023, and that you will lose access to all previously processed data six months after that.  For many businesses, this represents a significant loss of valuable commercial information and could cause difficulties in identifying longer-term trends.  Sure, you can export your data (and we would recommend that you start that process sooner rather than later) but, again, this brings its own challenges, not least in the storing and management of all that data for yourself.

So why has Google done it then?

First and foremost, Google aims to future-proof its analytics offering.  Ongoing privacy concerns and conflicts with EU regulations, as an example, mean that Google needs “to adapt to a future with or without cookies or identifiers” (Vidhya Srinivasan, Google).  There can be no doubt that GA4 is more privacy-friendly, supplementing the existing privacy features in UA with some more specific protections, e.g. GA4 does not log or store IP addresses.  Whilst there are still some question marks around whether this will actually satisfy EU regulations, there can be little doubt that Google has designed its new offering with privacy concerns in mind.

On a slightly more cynical note, Google may also be aiming to remove some legacy ‘technical debt’.  There are a plethora of sites with UA installed and gathering data where the corresponding property is not in use.  By sunsetting the old platform, Google benefits from a relatively clean slate, freeing themselves of some of this accumulated debt.

Is there any good news?

In a word: LOADS.  Aside from the improvements in GA4 from a privacy perspective, when set up and used to its full potential this next generation is an altogether more powerful tool than UA ever was.  Below, I’ve explored three main benefits in more detail:

  1. It puts the user at the front and centre of ongoing decision-making

One drawback of Universal Analytics lay in the central role of Sessions in measurement.  By definition, this meant that user journeys could not reliably be followed across devices or platforms.  Measurement now is based on users and their interactions, tracked as ‘Events’, and incorporates interactions with both apps and sites (and even simplifies cross-domain tracking).  GA4 also neatly joins up these events to give you a more accurate overview of how your users actually interact with your brand.

The out-of-the-box reports, too, are designed to align with the user journey, prioritising engagement over site performance.  This is reflected in the streamlined interface, with ‘Life Cycle’ split into Acquisition, Engagement and Monetisation replacing the old reports focused primarily on site performance.  Even the core metrics and pre-defined events within these reports arguably give greater insight into how you effectively are serving your users: events such as scrolls, outbound clicks, downloads and video views give a strong indication of engagement, whilst I would say that the Engaged Sessions metric provides a more user-centric picture of engagement than Bounce Rate ever did.  All of this can then feed into the development of more engaging, useful content – something which is only going to increase in importance following the recent Helpful Content Update.

  1. Improved data quality leads to deeper insights

Any analysis is only as good as the information upon which it is built but data quality is under constant assault from ad blockers, privacy legislation and the like.  GA4 helps to combat these limitations on your data through the application of machine learning to knit together different user journeys and to fill in the gaps between different data points.  The end result is the use of data-driven attribution (previously only available to premium customers of Google Analytics 360) as default, assigning more accurate credit to marketing campaigns and channels.

I confess to some scepticism when it comes to Google’s dual role – how can they take an unbiased approach to attribution when they profit from certain channels? – but there’s even positive news here: whereas previously you could only compare models within a single report (the Model Comparison Tool), you are now able to change the attribution model within individual reports.  This can provide reassurance that the data-driven model is accurate or inform different testing approaches you may wish to take to drive your bottom line.

Another massive improvement, in my opinion, is in the changes to data structure and measurement.  Gone are specific hit types and event taxonomy, which could be limiting and confusing, to be replaced by events and parameters.  Not only is this simpler, it gives you greater flexibility to pass back the key details you need.  As with many of the new features, the onus is on the user to design their approach at the outset, but a well thought out plan will remove excess noise to enable better decision-making.  For those who may not have the time or the resources available to do this, GA4 still gives you more through enhanced measurement, which gives you an additional range of automatically tracked events at the click of a button.  Regardless of which method you use, you will still have access to more insight within the new platform.

  1. Greater personalisation can identify incremental opportunities

I, for one, am also delighted that the new solution offers greater personalisation possibilities.  Having greater control over what data you need to analyse, which customised parameters matter to you, and having more direct integrations with media platforms can only enable you to hone in on what really matters.  The area where this personalisation is most apparent is in the new Explore tab.  This is an immensely powerful tool which allows you to slice and dice your data in a greater number of ways to unearth a wealth of insights which your competitors may have missed.  It is also here that you can create segments and audiences, drilling down into your data to focus on your specific needs on an ad hoc or recurring basis.

As with enhanced measurement, users with less time or budget can still benefit from Explorations by using the template gallery (helpfully broken down into Techniques, Use Cases and Industries).  Here you can see a number of familiar reports and visualisations, albeit in slightly unfamiliar surroundings.  However, the deepest insights can be found through some of the more customisable features now available.

There are two notable developments within Explorations which give far more insight into identifying barriers to conversion and calculating the value of a user.  The first of these is the changes to the Goal Flow (now Path Exploration) report.  Previously, there was limited scope for you to adapt this report and everything was one-directional.  However, you are now able to begin with the end, working backwards from a desired event to identify what users did before this point.  GA4 also uses a larger sample size to provide this insight, so the data quality is better.

Another brilliant addition – albeit one which was available as a beta in UA – is the ability to conduct Cohort and Lifetime analysis at a granular level.  This can inform your strategy and investment at a product- or service-level in much more detail than it was possible in the legacy system.  Ultimately, the ability to create bespoke reports could spark a range of hypotheses and the means to test them robustly.

 

Some final thoughts – The view from Spike

Putting to one side the fact that there is no choice in the matter (unless you intend to move to a third-party analytics tool), adaptability is at the core of digital marketing and we should embrace this development.  GA4 is already a powerful tool and it’s likely that there will be further developments and improvements to GA4 prior to the UA sunset, with fixes and additions based on user feedback almost inevitable.

Google (and everyone else, for that matter) advises that you should get your GA4 property set-up and collecting data now so that you will have some year-on-year comparisons when Universal Analytics shuts down for good in 2023.   Bearing in mind the complexities in setup discussed previously, the sooner you get started on this, the better.