Google Analytics is dead, long live Google Analytics 4

Portrait of John Hughes
By John Hughes

06 April 2022

Google's Universal Analytics is at end of life. Learn what you need to consider for migrating to Google Analytics 4 now.

Google has announced that from July next year it will stop collecting data into the older version of Google Analytics. It is time to migrate to Google Analytics 4.

Google Analytics 4 has been with us for about 18 months now. Initially when it was first released, it came as a shock to the system for many Google Analytics users with such a different interface, and particularly with many features so keenly used in Universal Analytics absent from it. Over the last couple of years features have grown and it has become more useful and richer. However, in our experience, most end-users remain considerably more comfortable in Universal Analytics. There is comfort in the familiar.

So, it comes as a shock to those users that Google has now announced that from 1st July 2023 (three months later for 360 accounts), you will no longer be able to track data into Universal Analytics. Google is moving to support only Google Analytics 4 instead.

Moreover, Google is only guaranteeing that it will continue to give you access to the data you collected in Universal analytics for six months beyond that date. This means that for most users, there is a reasonable chance that you will not be able to access data collected in Universal Analytics at all beyond the end of 2023.

What does this mean to the end users of Google Analytics?

Some of you may already have set up Google Analytics 4 properties and started tracking, in which case that is great – you have a head start on others. However, now is the time to start planning to switch your entire analytics function to focussing on the data from Google Analytics 4, and to capitalise on its many features.

There are many differences between each platform.

Firstly, the structure of data is very different. Universal Analytics was really built more like a platform for digital marketing data – helping you understand the channels that drove users to your website, how they engaged with it, and how many users converted. Its structure was essentially defined 15 years ago and has not really changed since – it has been expanded, but the original structure is still there. If you built a website 15 years ago and used the original urchin.js tracking code, Universal Analytics would still collect and store the data correctly now, despite two overhauls of the tracking code later!

This changes for Google Analytics 4. The entire data model has been ripped up and rebuilt to give you more flexibility. This meets the needs of the modern internet much better – there is no longer the assumption that what is being tracked is a website, for example, you could track sites, services, apps, and devices such as smart fridges in ways which the data would make sense contextually. Clearly, this is a vast improvement on the restrictive data model that existed in Universal Analytics.

Universal Analytics has a measurement API that enables you to send data to it directly rather than use Google’s own tracking code. We have utilised this at Storm to retain control over tracking libraries in some circumstances (e.g., to track security conscious apps and services). However, the API was completely open meaning anyone could send data to any GA property. We protected against this using filters and custom dimensions, but this openness was the root case for things like referral spam.

Google Analytics 4 also has a measurement API, but it requires an API key. This should prevent issues like referral spam from recurring.

Some features that were commonly used in Universal Analytics, such as advanced segments, are absent in Google Analytics 4, or changed so much from before that they are essentially not the same thing. Many features were not available at launch for Google Analytics 4 but have been added (such as cross domain tracking) so it may only be time before other features follow suit.

Google Analytics 4 has no views. Its account structure is one-level shallower than Universal Analytics. However, there is the concept of sub-properties, only available in Google Analytics 360 – the expensive premium version of Google Analytics.

In short, the new data structure and account structure of Google Analytics 4 enables considerably more flexibility, but the cost is the familiarity with the interface which we have all become accustomed to over the last 15 years.

Furthermore, as Google Analytics 4 uses slightly different cookies to Universal Analytics, you need to update cookie policies to ensure that your digital service remains compliant with PECR and GDPR legislation.

What should you do to move over to using Google Analytics 4?

Before you dive in, do a little planning first. A migration to Google Analytics 4 is an opportunity to improve how you report service use, and to align better with business objectives and important KPIs that you might not have tracked or reported well before. The flexibility of Google Analytics 4 is primed to make it much easier to do this.

Measurement plan process going from identify KPIs to record accurately to report effectively to predict confidently

At Storm, we focus on measurement planning and creating STAG documents. A STAG is a Site Tracking and Audit Guidance document, inspired by Brian Clifton, previously Head of EMEA at Google Analytics. Our version of a STAG lists business objectives and translates these into actionable KPIs that we want to be able to track. It then describes how to track and report these using Google Analytics metrics and dimensions.

Importantly, the STAG is a reference for data analysts to help them understand how data is tracked, and for data engineers as a bug-fixing reference when something goes wrong. Given the flexibility in Google Analytics 4, it becomes more important than ever to maintain a STAG document to enable customised and focused tracking and ensure reporting models remain accurate and well understood.

The cost of a well-designed measurement plan is of course the resource to create it, but the benefit is harvested through accurate, trustworthy and actionable data that helps you continuously improve a service informed by data. A commercial service can optimise marketing channels with confidence. A public service can fulfil its duty for transparency to stakeholders such as government and citizens.

Once you move to Google Analytics 4, prepare for a shock

The interface is not so intuitive

The reporting interface for Google Analytics 4 is much less intuitive than that for Universal Analytics. This is in part because the data flexibility means that the reporting interface cannot predict as easily how you will structure your data. This will be a major hurdle for casual self-service Analytics users in marketing and digital departments, especially those people used to handling requests for data and information coming from other stakeholders.

Screenshot of the GA4 explorer dashboard showing traffic to specific URLs

The navigation is different

Gone are the Audience, Acquisition, Behaviour and Conversion menus of old, replaced by Acquisition, Engagement, Monetisation and Retention. It is notable that the new navigation menu is similarly themed, if differently named to the REAN model we commonly use in Storm to understand marketing user journeys (Reach, Engagement, Activation, Nurture).

In fact, the navigation is customisable – you can remove reports that are irrelevant. You can even create and add your own reports after a fashion, albeit the custom reporting function is a little limited for now (this may of course improve with time). It also has ‘User’ reports, replacing ‘Audience’ in Universal Analytics.

There is an “Explore” option that enables you to deep dive into your data and understand complex analyses. Here you can create different kinds of reports, such as funnel explorations, path analysis and segment overlaps.

The advertising reports are more focused on marketing but require conversions to be correctly configured. Goals are no longer a thing in Google Analytics 4, instead you have an interface to set up conversions.

Screenshot of GA4 showing a reports snapshot with a graph of changing user numbers over a month.

Some of your tools may no longer work

You may previously have linked Google Analytics to other tools for richer data analysis. For example, at Storm we quite often used Google’s own tools to pull report data into a Google Sheet for data transformations and to join with other data sets. In many cases, these tools will not work with Google Analytics 4.

There are some relatively expensive API tools available to solve some of these problems. There are also new opportunities that you may not have previously exploited. For example, Google Analytics 4 natively links to BigQuery, and low-to-mid volume websites may well be able to continue to use that functionality for free.

Your stakeholders will ask questions you may not understand how to answer

The issue with new data models, new reporting interfaces, and new terminology is that you must become accustomed to them before you can confidently answer questions others have of you. If you are a gatekeeper for analytics data at your organisation, you will encounter this problem.

In fact, if other users in your organisation previously self-served simple data, you may find that the volume of questions about the data also increases quite dramatically, making this problem exponentially harder for you.

Some metrics will appear differently

In Google Analytics 4, Bounce Rate is dead (and frankly I dance on its grave, but I’ll explain why another time). In its place is an event called engaged_session. This event is triggered after a user has been on your website for ten seconds, or if the user has a second pageview. This makes the count of engaged_sessions as a fraction of the count of session_starts a much truer measurement of user engagement.

Session timeouts are counted differently also and are easily adjustable in the admin interface.

Users clicking alternate source/medium (tracking emails etc.) do not break an existing session like they do in Universal Analytics, likewise sessions are not broken in half at midnight like in Universal Analytics.

With Google Analytics 4’s new data model, custom dimensions and metrics cannot be session scoped.. Since session counts are derived from the session_start event, older session scoped custom dimensions and metrics must now shift to fit the new structure.

All of this means that the metrics recorded in Google Analytics 4 are not quite the same as the metrics recorded in Universal Analytics. They are typically better and more accurate, but you may have slightly fewer sessions in Google Analytics 4.

How can you tell when it is broken?

If you are not familiar with the reporting interface or data structure, how can you tell it is working correctly? Google Analytics 4 has a debug mode than can be used for testing. On top of this if you are familiar with the tracking model, you can look to see that all the events you expect to be tracking are – another good reason to keep and maintain a STAG document. However, it takes experience to know where to look in the interface for this information and to test it quickly. Each new deployment to the site or service might introduce a bug.

Screenshot of GA4 showing the debug view with a timeline of events the user is tracking

How do you find the time to learn to use Google Analytics 4?

As the interface is so different, how do you find the time to upskill in Google Analytics 4? What other work priorities must take a back seat to enable you to make the best use of Google Analytics 4? This is one of the biggest challenges facing digital teams using Google Analytics 4 over the next two or three years.

Many digital marketers had a good working knowledge of some parts of Universal Analytics, but to shift that knowledge to Google Analytics 4, knowing what is still relevant and what to throw away, coupled with learning entirely new interfaces, reports and functions quickly introduces a significant barrier to using the platform to its full potential.

How do you update and manage your reporting service to stakeholders?

If the platform will take some time to adjust to in terms of knowledge, in terms of user capabilities, in terms of understanding and in terms of tools readiness, how are you going to manage stakeholder reporting?

How we overcome these issues at Storm

At Storm, we have carefully planned our Google Analytics 4 migration services.

We see there being three different ways we may need to use to help a digital service to move to Google Analytics 4, depending on what Google Analytics setup they already have.

Basic installs (with or without ecommerce tracking)

These installations use the basic tracking functionality of Google Analytics 4, but with no significant customisation. There is no STAG, and no customised tracking code is created.

Customised installs (with or without ecommerce tracking)

If your existing Universal Analytics tracking code has some customisations in it, we may be able to recreate these as closely as we can. With no STAG, these customisations are possible, but tricky to manage in the long term.

Measurement plan (with or without ecommerce)

This version of the service gives the best long-term opportunity for success. We will go through a full measurement plan process, including working with you to identify business objectives and KPIs, and to build a customised tracking and reporting plan that aligns closely with these.

How we can help you get the best long-term value from Google Analytics 4

Installation of Google Analytics 4 is only the beginning. As we described above, it takes knowledge and experience to be able to use Google Analytics 4 effectively. We can help you overcome these barriers with some of our ancillary services.


We can help you overcome the knowledge gap between Universal Analytics and Google Analytics 4. This helps you to become self-sufficient in the platform itself so you can get the answers you need from the reporting and data exploration features.

  • Walkthrough your own Google Analytics 4 setup
  • Learn how to find the data you need
  • Discover features that help you make faster decisions

Customised reporting setup

We can help you customise the reporting interface on Google Analytics 4 to include your custom dimensions and metrics derived from your own flexible data model. Find the data you collect quickly and easily in the reporting interface.

  • Mix default and custom dimensions and metrics
  • Organise the reporting section exactly how you need it
  • Replace or enhance Google Analytics 4’s own reporting structure

Google Data Studio dashboards

We can create helpful and meaningful reports customised to your specific Google Analytics 4 setup, in a manner that is understandable to you and your other data stakeholders. Overcome the barrier of the limited in-platform reporting and stay on top of your digital performance.

  • Understandable and easy to use report formats
  • Control who has access to report data
  • Extensible to join data from other sources, e.g. CRM, social media, etc.

Monthly reporting and insights

We can deliver monthly human curated analytics reports to help you tunnel through your data into real insight and recommendations. Make decisions quickly and with confidence based on actionable insight and expertise.

  • Report across the entire customer journey
  • Actionable insights backed-up with data and expertise
  • Includes telephone or video call support

Health monitoring

We can use a combination of automated and periodic human checks to ensure that your tracking code and data collection continue to track data accurately. This gives you the confidence in your data to know that the decisions you make are based on true and accurate information.

  • Automated checks intercept problems quickly
  • Periodic human checks validate the full data collection journey
  • Have confidence in the validity of your Google Analytics 4 data

Universal Analytics data exporting

We can help you export reporting data from Universal Analytics so that you retain access to it beyond the date that Google is expected to delete it. This ensures you can still make useful long term data comparisons and understand long-term trends in engagement and conversion data.

  • Export any existing Universal Analytics report data to MS Excel or Google Sheets
  • Keep your historical digital performance data for YoY and larger comparisons
  • Recycle your data into Google Data Studio dashboards etc.

What next?

If you are not an existing Storm client but would like help and advice on migrating to Google Analytics 4, please get in touch.