Top analytics tips: Why ecommerce retailers need to embrace GA4, the future of online measurement

Top analytics tips: Why ecommerce retailers need to embrace GA4, the future of online measurement

Google has quietly set in motion one of its biggest changes to analytics in its 21 year history. GA4 enables unified measurement of user interactions and events between web and app for the first time. The platform is set to replace Google analytics, so ecommerce practioners need to engage now in order to realise its benefits and get a head start on competitors. Can Ozcer, Head of Analytics & Insights, UK from data company fifty-five London walks through the platform and the key benefits.

With the launch of GA4, Google has set in motion its biggest changes to Google Analytics in its 21 year history. The new system introduces an official framework to facilitate a unified measurement of user interactions and events between web and app for the first time. With the platform set to become the next chapter of Google Analytics, ecommerce practitioners need to engage now in order to realise its benefits and get a head start on competitors. So what is it, how does it work and why can it help you understand your customers better?

What is GA4?

Simply put GA4 is a measurement system set to become the new version of Google Analytics. Google Analytics has been around for 21 years and has worked well. However in today’s mobile first world the insights we need are different. Customers have markedly different ways of interacting with controlled content within apps and with websites. GA4 combines data from the two sources into a single suite. Despite the name, the new system isn’t just about reconciling data from mobile apps and websites. The new product has different and more advanced capabilities than Universal Analytics, enabling practitioners to develop more sophisticated and personalised user journeys. This applies even if they are only interested in websites and not mobile apps. While it is presently in BETA digital marketers should have a plan now for fully implementing it.

How is it different?

While it was possible to view website and app data on the same screen with Universal Analytics, it was far more complicated to get a clear read without expert support. Google’s new system provides a proper framework, so people don’t have to hack their way around the product.

The technical change is that the entire product and data schema it is based on is changing. The old product was based on the idea that people have ‘sessions’ and ‘page views’. However GA4 takes a more bottom up approach, replacing page views and sessions with a building block called an ‘event’. This can be a specific thing like a ‘page view’ to understand how people consume content. It could also be people completing a specific series of actions, like a funnel. So it is a change in the framework of how we measure people doing things.

Why is Google doing this?

It is not merely a technical choice for engineering sake. Adopting this new data model enables the product to perform certain analytical computations a lot faster than the 21-year-old architecture of Universal Analytics.

Essentially, in the world of Analytics, “Product Analytics” and “Web Analytics” used to be very different things. Now that websites are becoming more like products (and not mere content holders), and mobile apps having always been closer to “product” experiences, a new framework is needed to efficiently service their Analytics needs. Of course there are existing product analytics suites that behave in this way, such as Mixpanel and Amplitude. Google is following but doing it in a bigger way that is compatible with their engineering. In some ways it is not dissimilar to how social media platforms like Instagram have adopted features from disrupters like Snapchat.

How would it work in practice for someone in ecommerce?

If you imagine your view of your customer behaviour across web and app is like a ground level view of two mazes. It is time consuming to then create a map to the destination you want to lead your customers. The new product instantly provides a helicopter view of both mazes, providing a faster and fuller picture with the routes clearly visible. In the past you would have had to go through a process to merge the results for a specific SKU across platforms and then analyse them. Now from this new vantage point, you can do that on the same screen and you can see the product broken down on the app and the web. You can also see the conversion rate right away and take the action needed to drive more sales.

For example, you can now chart the journey of a customer who first discovered your brand thanks to your website, completed their first transaction there, downloaded your app and made all their future purchases in app. You can now do this without leaving your one single interface. Of course this is provided they’ve logged-in to their account on both platforms and provided consent for such analysis.

What are its key features?

A) Reporting

Events: The new data model provides insight on what is happening with your website or app, such as user actions, system events, or errors. These are categorised as ‘events’. There are three categories of events that can be sent to GA4. These include automatically collected events, recommended events or custom events that can be set up.

Audience analysis: Because of the fundamental shift in the data model, a lot of new and detailed audience analysis across different platforms is now possible.

Mobile App and Web data integration: The new system creates a joint and approved and well documented way of having analytics across both mobile app and websites.

New reporting capabilities: The new data model facilitates certain types of analysis which are computationally intensive and were previously only reserved to Google Analytics 360 customers in Universal Analytics, as well as adding previously unavailable types of reports even on the paid platform. A drag & drop interface lets you build pivot tables, pathing reports, segment overlaps and more.

There are also a host of other features such as a native BigQuery integration and the introduction of hourly custom alerts.

B) Personalisation

Google Optimize: The platform can be coupled with Google’s personalisation platform – Google Optimize. Currently, Google Analytics based audiences in Optimize take several minutes to be processed and calculated. The new data model is set to enable this to be seconds instead. Previously challenging analyses, such as user journey pathing and segment overlaps, can be built in a matter of minutes thanks to the new Analysis section of the platform

Experience optimisation & personalisation: By introducing this new data model Google started thinking about much faster activation, whether personalised ads through Google Ads and Google Marketing Platform (GMP). You can now do this in seconds thanks to a far leaner data model.
Improved Machine Learning functions: Machine Learning is changing the way we do things in a lot of areas, Analytics is no exception. Universal Analytics paved the way for Machine Learning (ML)-based automated insights Google provides users. Going forward, ML-based enhancements will be exclusive to GA4 as its data model is fundamentally more suited to data science activities.

How do you get started with it?

This is a great step from Google in creating a modern data schema rather than continuing to rely on one first created in the 90s, but it is still in development. Despite the potential for a more intuitive and automated approach, it is critical to work with the right experts to maximise the benefit and have the right parameters set up. This is particularly the case while the system is at an early stage. As with any new system, upgrades are occurring each quarter and being refined based on user feedback. A lot of features and changes are still to happen. Expectations need to be managed as to what is possible and what is not possible. Understanding these differences is crucial.

It is therefore crucial to retain access to both the new analytics and old analytics. You don’t want to be without your year on year reports for example for a holiday season comparison so you will need the historical analytics data to be there when you migrate. By adopting the platform now digital marketers can get a head start on their competitors. With GA4 set to become the mainstream Google Analytics platform in 2021 and beyond, it is crucial to engage now.

By Can Ozcer

Head of Analytics & Insights, UK

fifty-five