How to analyze data and create reports with Google Analytics 4
Joining us today on the Marketing Analytics Show is Charles Farina, an expert in Google Analytics 4.
Tune in to learn:
- What Google Analytics 4 is
- What the differences between GA4 and Universal Analytics are
- What kind of reports you can create with Google Analytics 4 data
- How the data model has changed
- How to set up event-based tracking if you want to start using GA4
Links mentioned on the show
- Follow Charles on Twitter.
- Start your 14-day free trial of Supermetrics.
- Follow Supermetrics on Twitter.
- Follow Anna on Twitter.
Subscribe to the Marketing Analytics Show newsletter
Hello, Charles, and welcome to the show.
Hi, so excited to be here.
Awesome. Yay, we’re really happy to have you here. And I’m actually super excited about this episode because I heard a lot about the new GA4, and I’m sure our audience would love to learn more about it.
So let’s start with the first question. So first of all, could you please tell us: What is Google Analytics for, and how is it different from the previous versions of Google Analytics and from Universal Analytics?
Yeah, so Google Analytics 4 is essentially a completely new version of Google Analytics. The best way I like to describe it is Google Analytics has been around for about 15 years now, and this is the first time that Google’s actually updated the entire platform, both the front end and the backend.
So because of that, for most people, it’s not backward compatible, it requires new tags and brand new implementation, and it differs from Universal Analytics in almost every way.
It’s a new data model, it’s new features, it’s a new UI, and what I’m most excited about it is many of us have been waiting for features like pathing, and funnels, and big query integrations in Universal, and now for the first time ever in GA4, we have some of these new capabilities that have simply not been available before to us.
Wow, that definitely sounds awesome, and actually, I didn’t know they also updated the back-end, I only thought that they updated the UI and made some improvements to their reporting, but yeah, that totally sounds like a whole new world.
So with that in mind, could you please tell us a bit more, why should marketers start using the new GA4 if they haven’t yet started it?
Yeah, so one of the biggest reasons you’re going to want to start using GA4 today is because, since it has this new data model and needs this new implementation, you have to start using it and start implementing it in order to get access to all these new features. You can’t carry your existing data from your old implementations over to this new platform. And to your point about the new backend, really what’s changed here is Google’s adopting an event-based data model. This data model is very common in other platforms, like Snowplow and many others, and it’s going to make it easier to work with.
So at the end of the day, marketers are going to want to start using GA4 because, in a few months, it’s just going to be the best place to be. It’s where these new features are, where you can create funnels. We can actually do pathing analysis and understand content performance in ways that were really hard to do in Universal, and simply in order to get access to all these features you need to use GA4. And what we’re going to talk through in our session today, Anna is it’s very easy to actually use both at the same time. And for right now, that’s actually the best recommendation, is to continue to use Universal, but just start using GA4 in addition to the GA properties you already have in place.
All right, yeah, it definitely does sound like a wise idea to start using both maybe in parallel, and then switch to GA4. So could you please, actually, maybe talk more about this? So what should marketers consider when they’re starting to implement the GA4, so any prep work that needs to be done, and maybe then you can touch on how they could run both in parallel, and what would be the best outcome of using both of these in parallel?
Yeah, I think the best place to start is to think about what vertical, and what your use cases are with Google Analytics. So some things are, like for Firebase for mobile apps, so if you have an iOS app or an Android app, Google’s already deprecated some of the functionality from Universal Analytics for mobile apps and has been pushing users over to Google Analytics for Firebase for quite some time. So for mobile apps, actually not that much has changed, because in fact, GA4 is basically the same exact data model as Firebase, and Google actually allows you to upgrade your Firebase properties to GA4, and it carries all of your data forward. So for mobile apps, you don’t really have to think about doing anything that different.
Now for the web, there could be a variety of things you want to think about here. One of the biggest challenges I think we have is GA4 is not yet complete. It doesn’t have all of the use cases that we have and we need from Universal Analytics. And I’m sure Google’s going to address this throughout the year, there are just all these new features, like attribution, we need more view level components, so how do we manage user permissions? How do we do country-level views, all this sort of stuff is going to be coming out, I think, throughout this year.
So really the biggest thing right now is to just consider where you are at. If you already have Universal Analytics in place, then my recommendation is just to start with the foundation. Just try and get the basic page view tag implemented so you can start seeing the new event-based data model and the new features, and then my best advice is, you can basically take the implementation and break it out. So just plan over the next 12 months of starting to move your events over from Universal into GA4, and you can do it in phases. And I always like to think about, start with the events that are most important, which are your conversions, then move on to the second level of what events lead to conversions. And then lastly, you can just move over all of the nice to have events. So those random things where you’re tracking buttons or scrolls, or all sorts of other engagements, you can do that last. So just break it down into phases, make it easy, and take your time would be my advice there.
Yeah, it does definitely sound like great advice. And also another question. So you’ve mentioned these event-based data models, and that GA4 technically uses events as their main concept. So could you please tell us more about this data model? So as it can combine data from different devices, how has the data model changed? And maybe then you could dive deeper into what events are here in QA4, and then you’re talking about different kinds of events, so how should marketers set the events correctly? Are there any specific things they should take into consideration?
Yeah, great question, Anna. So I think to start, let’s think about Universal Analytics. So the version of Google Analytics we’ve been using forever, that model was based on a hit-based data model. So when we track things, we haven’t determined what sort of hit it was. Was it a page view hit? Was it a transaction hit? Was it an event? And each of those different hit types had very different schemas and different capabilities, and because of that, it was one of the reasons that somethings in GA didn’t always work so well.
So as an example, if you wanted to figure out something around content performance, and you wanted to understand, how does a particular piece of content lead to conversions? Or how do particular pieces of content lead to e-commerce transactions? If you go into the content reports in Google Analytics today, there are no conversion columns, and there are no revenue columns for transactions because hits don’t, or at least these page view hits, don’t really work well with other event hits. The data model just doesn’t really support it.
So what’s changing with GA4 is it has this new event-based data model where simply everything is an event. Even a page view is an event. So simply you have an event name, and then you have these event parameters, and the parameters you just make them relevant to whatever your use case is. So that’s a big difference here is that it’s kind of adopting this new event-based data model with these parameters that just allow you to make it very flexible depending on your use case.
Yeah, that definitely sounds very interesting, and I’m sure that many marketers would love to use it for a variety of different use cases now that the data model is so flexible. And also you mentioned the schema has changed because the data model has changed, so basically now we should have a whole new set of metrics and dimensions, which leads me to my next question, so what kind of reports can marketers create using the GA4 data? And maybe another question here would be, what reports would you recommend starting with?
Yeah, so once you get started with GA4, there are two key areas in the platform, and this is a big change from Universal. So GA4 is going to give you a set of canned, out-of-the-box reports. So in the old version of Google Analytics, we had the ABCs that we all talked about, audience, acquisition, behavior, and conversions. In GA4 they’ve kind of rebuilt it, and now we have the AEMR. I haven’t figured out a catchy acronym for that, but they’re organized into acquisition, engagement, monetization, and retention. And you can even see right there, it’s giving a newer feel to how most of us talk about these key areas today, which I like.
Now those reports are kind of predefined. I’m hoping they get a lot more flexible in the future. My recommendation would be, if you’re starting out with GA4, one of the coolest things is Google actually announced, just very recently, there’s a new demo account that you can try without even implementing and get familiar with it. The best place to go, in my opinion, is you should go straight into the analysis reports. So analysis is really exciting because previously this was only available in the paid version of Google Analytics, Google Analytics 360. Google’s actually taken this report component and now in GA4, everyone has access to it for free. And what I’m excited about is in here, we have some really cool features that I think you’re all really going to like.
So for the first time, you can actually build a meaningful funnel in GA. So you can build funnels that are retroactive, they’re user scoped, you can define your own steps, they’re very flexible. There’s also pathing in there, so for the first time, we can do meaningful pathing. You can do forwards pathing, you can do backward pathing, all sorts of rich components. And one of my favorite features that shows the new capabilities in GA4, to your point about new metrics, is, as an example, when you’re building audiences, or you’re building funnels, for the first time we can actually do meaningful analysis with time.
So in Universal Analytics, we really only could do time-based analysis on session time or time per page. Those were the only kind of time metrics. GA4, because of the new data model and how these events work, time is actually way easier to work with. So here’s an example. If you wanted to do analysis and figure out how long does it take your users to do any kind of two things if it was filling out a form, and when did they start it and when did they finish it, or when did they get to your homepage, and how long did it take them to get to your blog? You couldn’t really do that analysis in Universal. With GA4 you can just add a metric called elapsed time, and you can say, okay, let’s build a funnel of five steps, and just show me how much time it took in days, minutes, seconds for each user to complete those steps.
Or, when you build audiences, you can say, hey, I have these two random interactions that I want to analyze. Show me all of my users who did it within two minutes, or five minutes, or six days, whatever you want. So because of this, a lot of that functionality that I’m mentioning that’s just new, most of that is going to be inside this analysis module. So my recommendation is if you’re getting started with GA4, and you’re either in the demo account, or you have some data to play with, head over to analysis because really that’s where a lot of these exciting new features and components will be. And I think in the future, there’ll be even more in here as Google releases new features throughout the rest of the year.
That does sound like a whole new set of reports, and yeah, really interesting to hear about the funnels and pathing, and especially about the time dimensions, because I’m sure many more users would appreciate that, and that would really help improve the UI/UX experiences as well. And yeah, so a bit more about the reports. You mentioned there are a couple of reports you can create with GA data and do you have any thoughts about what cross channel reports can marketers create with data from different platforms combining it with GA data? So for example, we can have a beta report like Facebook plus GA, any kind of opportunities that you see there, or maybe you could throw in a couple of ideas for marketers to experiment with, now that you also mentioned that they can export this GA4 data?
Yeah, so cross-channel is really interesting, Anna. So right now we have a lot of excitement, and then a little bit of a challenge. I’ll start with the challenge. The challenge here is this is one of the use cases that Google still has a variety of features they need to release. So if you’re familiar with Universal Analytics, then you probably use multi-channel funnels, and you use the attribution components a fair bit. That entire attribution module is not yet available in GA4. I’m hoping it comes soon, and once we have that it opens up a variety of easy-to-consume cross-channel reports. So anyway, I’m sure Google is going to be releasing new features to help us with that.
Now what we have today I think is pretty exciting. So when we talked about Google analytics in the past, once you become a more sophisticated user, you realize that all the default reports all use a last non-direct attribution model. And in fact, you can’t even change that attribution module, and the only place you can really do cross channel is if you’re doing advanced segments, or perhaps you’re using the big query feed with Google Analytics 360, or some other niche components in the existing platform.
Now with GA4, if you go in there today, one of the things you’ll discover is you can hit customized reports in any of the standard reports, and one of the options it shows you is that you can actually change the default attribution module for all of the reports. So that default attribution, right now it’s still really only last click or last engagement, but the fact you even have some options to change just gives you a preview of the future. So I’m hoping that sometime soon Google will actually release other abilities, maybe a data-driven option, maybe a linear option, just a bunch of options for us to actually, for the first time ever, change what the default attribution module is in GA4. So that just gives you some hopefully previews of what we might have in the future.
And then to the last point, Anna, you mentioned the big query feed. So this is a huge change, and in my opinion, the biggest change with GA4. All of us who use existing GA all have the same challenge at some point in time, which is once you become a more intermediate or advanced user, you just can’t do everything in the UI. All of us at some point need access to the raw data to do our own data visualization, to do our own data joins, integrate it with our CRM, or our e-commerce platform, or to even just do this cross channel and build our own attribution reports and components. That just wasn’t really possible in the free version of Google Analytics because you are just very limited in how you can get data out of the API. You had to deal with sampling and all sorts of other things.
Now GA4 has the free big query integration, which used to only be available in GA 360, the enterprise version. And that’s really exciting because there you actually have all of your raw data, you have all of the user and client IDs, you have all the traffic source information. So when you’re thinking of cross-channel, you just have the ability to pretty much do whatever you want. Predictive, you can build your own models, all of these advanced use cases which just weren’t possible before. So with that cross-channel, there are some really exciting components for GA4 in there.
Yeah, sounds awesome. So to add to that, I actually was really curious to know about more practical applications of GA4, so could you actually please share a couple of examples of how GA4 changed the way analytics and reporting are done at your company, Adswerve.
Yeah, so the biggest change right now that’s happening is in a few different areas, I would say. The product’s still really new, so most of us are still primarily using universal for most of our workstreams, but there are a few use cases where we’ve moved over to GA4. Some of the use cases are first around BigQuery. So we use big query very heavily. One of the challenges that we had with BigQuery on GA was in GA 360, the paid version, which we have. Our BigQuery feeds, actually the fastest we could get them updated was every 10 to 15 minutes, which was the streaming inserts. With GA4 the BigQuery feeds actually have real-time streams that are within seconds.
So if you ever get into CDPs, your kind of customer data platform, and you want to activate data on real-time based on what someone’s doing in GA, you can open up a whole world around CDP. So you can do email activation, you can trigger things in your CRM. So our team is starting to use these new big query feeds to basically build their own CDP. So that’s architecture and innovation.
Another big change and this is another one of my favorite features in GA4, Google rebuilt conversion tracking. So for the first time, conversion tracking is way more flexible. You can reuse goals, you actually have 30 conversions instead of 20, and simply my favorite feature is that you can actually take any audience that you build and make it a conversion. So previously in Universal, you could only really have a conversion based on a page URL, or an event: what was that category action label? Or what was the page URL?
In GA4, you can build audiences with whatever conditions you want. So I could say, I want an audience of users who are in the US, who viewed more than four pages, and who did not convert, but also spent 15 minutes on a certain device. And using this audience trigger functionality, you can create an event from an audience when someone enters it, which you then use as a conversion. So simply to your question, Anna, one of the other changes there is it’s opening up a whole new layer of sophisticated conversion tracking, which then you not only have for reports, but you can also layer into Google Ads for audience activation, and other components as well. So right there, that event-based data model is bringing some pretty sophisticated changes there. So those are two examples for you on that.
Yeah, wow. I think this has moved GA to a totally new level. It does sound like it has so much more granularity to the whole analytics and reporting suite. Yeah, Charles, thank you so much for sharing all these useful insights, and where can the audience learn more about you?
Oh, so that’s pretty simple. I post a lot about GA4 on my own Twitter handle, so just Charles Farina, and also on LinkedIn, and occasionally I also blog under the same URL. So just follow me anywhere on social media, and if you want to hear more about GA4, I hope to be a really great resource for you.
Yeah, awesome. And do you actually want to mention a couple of words about your CXL course? Because I think our audience would find it super useful.
Yeah, so I had an opportunity to launch the first conversion XL course on Google Analytics 4, so what I really liked about the course that I built is I tried not to just use slideware, but I tried to make it really interactive and engaging, and get you experienced by being hands-on with the product. So in my course, I basically share everything I know about GA4 today, and we actually implement it together. So we go to a Weebly, we create a CMS together, we implement GTM. I show you how to implement GA4, and then with this implementation that we all do together, we learn all the new features, like event modifications, audience triggers, how to set up your big query integration, all the new privacy components. So there’s about, I think, four to five hours of the content thereon GA4 to try and get you everything you know on how to get started with this new platform. So I really enjoyed that and would love to get any feedback on anyone that goes through that.
Yeah, amazing. And thank you so much for coming to the show today.
It was really great to be here. I really enjoyed the questions, and thanks so much, Anna.
Try Supermetrics for free
Get full access to Supermetrics with a 14-day free trial.
No credit card required.