UK Search Awards 2026 REBEL is shortlisted four times, across three client campaigns. Read the story →
REBEL Agency logoREBEL Agency logo
Insight / Data & Analytics

Three years on: are you OK, GA4?

July 2023: Universal Analytics was switched off and almost nobody was happy about what replaced it. Three years of Google listening to feedback later, Benni checks back in, and finds five things GA4 now genuinely does well for ecommerce businesses.

By Benni, Client Director20 October 2026Long readShare on LinkedIn ↗
TL;DR
  • GA4 arrived as the only option in July 2023 to near-universal grumbling. Three years of feedback later, it has quietly become a much more capable platform.
  • Full-funnel ecommerce data (product views, baskets, checkouts) turns “is it bad traffic?” into a question you can actually answer.
  • Explorations, multi-touch attribution and behaviour-based audiences move GA4 from reporting what happened to shaping what campaigns do next.
  • Predictive audiences (likely purchasers, likely churners) point at where ecommerce measurement is heading.
  • It still turns simple questions into six clicks and an Exploration. But the platform has grown up.
A laptop showing an analytics dashboard of charts and KPIs
Three years in, the dashboards finally feel like home. Mostly.
01 / 1 July 2023

A breakup nobody
asked for.

July 1st 2023... not exactly a date that jumps out in the memory. But in the world of digital marketing, it was a pretty big one for the vast majority of us.

On July 1st 2023, Google finally made the leap to GA4 becoming the required platform for ecommerce measurement, and Universal Analytics stopped processing data. We had to say goodbye to our dear friend. We'd grown up together. We knew where everything was. We understood each other. And suddenly, we had to embrace the change.

And, at the time, pretty much everyone I spoke to hated it.

“IT’S A STEP BACKWARDS.”

“WHY HAVE THEY GOT RID OF SO MANY FEATURES?”

“WHERE THE HELL HAS THAT REPORT GONE?”

And I have to say, I agreed. After spending years learning, growing and becoming comfortable with Universal Analytics, being handed a product that was arguably worse than the one it replaced was less than ideal.

But that was then. It's been over three years. Google have listened to feedback, added features, made changes and gradually turned GA4 into a much more capable platform than the one we were all reluctantly logging into back in 2023. So, three years on, it feels like a good time to check in.

Let's get one thing out of the way: GA4 still isn't perfect. There are still moments where I find myself wondering why something that should be incredibly simple requires six clicks, an Exploration and a minor emotional breakdown. But there's also a lot it now does really well, particularly if you're running an ecommerce business. So, in the interests of fairness, here are five things we actually like about GA4.

02 / The ecommerce journey

See the whole journey,
not the final number.

One of GA4's biggest strengths is the level of detail you can get around what happens before somebody buys. Product views. Add to baskets. Checkout starts. Purchases. It gives us the ability to look beyond the final revenue number and understand how people are actually moving through the site.

And that matters.

If traffic is up 30% but revenue hasn't moved, the immediate reaction can be: “We're driving bad traffic.” Maybe. But what if add-to-basket rates are up too and customers are falling away during checkout? Suddenly, that isn't just a marketing conversation. It's an ecommerce conversation, and one we've written about before.

GA4 gives both teams the data to find out where the problem actually sits.

03 / Explorations

When the standard reports
aren’t enough.

This is probably one of the areas where GA4 has genuinely moved things forward. Explorations allow you to get much deeper into the data and answer the questions that inevitably come up five minutes into a trade meeting.

How do new customers behave compared with returning customers? What did people do before they purchased? Where are users dropping out of the funnel? Do customers arriving through Paid Social behave differently to those coming through Paid Search?

The standard reports give you the headlines. Explorations let you start asking why. And for me, that's where analytics becomes much more useful.

04 / Attribution

Credit where credit’s
actually due.

Customers very rarely see an ad, click it and immediately buy. They might discover you through Instagram, come back through Google a week later, click an email two days after that and eventually search for your brand and purchase.

Give all the credit to that final brand search and you've got a pretty misleading picture of what actually drove the sale. GA4's attribution reporting gives us a much better opportunity to understand the role different channels play across that journey.

That's particularly important when you're making budget decisions. If we only invest in the channels that appear at the bottom of the funnel, eventually there's nobody left to arrive there. It's the same thinking that sits behind our data analytics work: measure the journey, not just the last click.

05 / Audiences

Audiences you can
actually activate.

This is an area where GA4 gets particularly interesting when you connect it with Google Ads. Rather than simply looking at everyone who visited the website, we can start building audiences around actual behaviour.

People who viewed a particular category but didn’t purchase. Customers who added something to their basket but disappeared. Previous purchasers. High-value customers. Users who have engaged multiple times without converting.

That gives us a much better understanding of the different people visiting the site, but crucially, those audiences can also be activated. Analytics stops being somewhere we just go to look at what happened last month and starts becoming something that can influence what our campaigns do next.

06 / Predictive audiences

Looking forward,
not backwards.

And this is probably where you start to see what Google ultimately wants GA4 to become. Historically, analytics has been retrospective. Someone bought something. Someone abandoned their basket. Someone spent £300. We look at the data and decide what to do next.

GA4's predictive capabilities start to flip that around. Where you have enough data, Google can use machine learning to identify things like users who are likely to purchase or customers who may be likely to churn.

That opens up some really interesting opportunities. Rather than spending money treating every visitor in exactly the same way, we can start thinking about where the biggest opportunity actually sits and build audiences and campaigns around it.

It's still an area that will develop significantly, but it feels much closer to where ecommerce measurement is heading: less reporting on what happened, and more helping us decide what to do next.

07 / The verdict

So, are you OK, GA4?
Actually... yeah.

You're alright. Would I have said that in July 2023? Absolutely not.

There are still things I miss about Universal Analytics. There are still reports that feel harder to get to than they should be. And there are definitely still occasions where GA4 manages to turn a very simple question into a surprisingly complicated answer.

But the platform has grown up. And perhaps we have too.

The biggest change for me is that GA4 increasingly encourages us to stop looking at channels, sessions and conversions in isolation and start looking at the customer journey as a whole. For ecommerce businesses, that's ultimately the conversation we should have been having all along.

Want a second pair of eyes on your GA4 setup? Here at REBEL, we offer full GA4 audits to make sure your setup is working as hard as it can for your business. Get in touch.

Related
insights.

Want thinking like this on your account?

This is what giving a damn looks like in practice. Tell us what's not working.