Get comfortable with GA4
Natalie says: “Get comfortable with GA4. Universal Analytics is going away on the 1st of July 2023, when it will stop recording data.
The best version of this scenario would be that businesses will have migrated to GA4 already, so that they have a good amount of data to compare performance year-on-year by the time the data stops recording in Universal Analytics.
The second-best time to do it is now.
Universal Analytics will stop recording data and that impacts all kinds of things from a marketing perspective, and certainly not just SEO. You can’t manage what you can’t measure. Obviously, other platforms are available, but the transition from Universal Analytics to GA4 doesn’t have to be as painful as some people have thought. I was very sceptical of GA4, to begin with, and I’m a new convert to the cause.”
What is different about measuring things with GA4?
“The main difference is the unit of measurement. Also, the way the data is collected is different too. All other versions of Google Analytics have always been based on hits and some event tracking. However, anything that happens within Google Analytics 4 is an event.
That does mean that there are discrepancies between them. If you run GA4 alongside Universal, you’re going to see some differences in the session data. If it’s drastically different then that’s something to look into, but there will always be some discrepancies. My advice would be to install GA4 alongside Universal for as long as you possibly can, to get acquainted with those variances in the data and figure out what’s going on. For me as a practitioner, that’s been invaluable in terms of learning how all of this works. Learn by doing.
It’s a fairly straightforward process. If you want a basic implementation of GA4 - to just get the code on there and see what it tracks out of the box - then it’s usually a five-minute job. The bigger part is when you’ve got a huge website with lots of manual events and goals that you’re tracking. That can become quite fiddly, but take the first step and get the code up there.”
How will the data change when it comes to viewing everything as events?
“GA4 does still use sessions and pageviews; those are metrics that are still in GA4. It’s the step before that where the difference is; in how the data is gathered. The unit of measurement in Universal Analytics is a hit, which is then translated into sessions and pageviews data. Now the unit of measurement is an event, and all those events are untainted in GA4. There’s no change in what the metric is.
I don’t know a huge amount about how that works exactly, but the top line is that the unit of measurement is different, which leads to discrepancies between the two datasets. That needs to be borne in mind when making those comparisons - that some variances are to be expected.”
When moving from Universal Analytics to GA4, how do SEOs take the data, make sense of the changes, and compare what they’re seeing at the moment with what has happened in the past?
“The most straightforward way of doing that would be within Google Data Studio and combining those two sources of data. There are plenty of templates that people have built specifically for this task.
Personally, I still run all my reports in Excel. I like to be very hands-on with the data because, that way, I don’t miss anything. The manual approach is often good.
In terms of making notes within GA itself, you can still annotate with GA4. There are all kinds of custom reports using the Explore function that you can build, but I would say that Data Studio is your best bet for that direct comparison.”
How far should an SEO go back to make sense of that data?
“Personally, I would go back more than a year. The only reason I’m saying that is because of how consumers have behaved so differently over the previous two years. For many campaigns I’m running, the best year-on-year comparison still tends to be against 2019, because 2020 and 2021 were so unusual.
It depends on the sector. You might be running a site where everything’s just been the same all the way through, although I doubt it. For that reason, I would say benchmarking from the start of 2019 at least can be quite helpful, as a minimum for any dataset.”
Why has Google introduced GA4, and what are the benefits?
From what I can tell, it’s a data protection situation. I don’t spend too much time looking at Google’s reasoning behind this, I mainly want to just get on with it.
It seems that the benefits are more for Google than the user. I think that’s why, as an industry, people have been quite frustrated. There hasn’t been as much hand-holding through this process, or as much notice as people might have liked. There are definitely flaws - or things about GA4 that are not as well put together as Universal Analytics - and that’s where some of the frustration comes from, but I think they’ll get there.
There have been recent changes that have made it a lot easier to do eCommerce tracking within GA4, for example, as it was quite fiddly up until a couple of months a month ago. They are ever-evolving and catching up with themselves.
It is tricky to think of the benefits because we don’t like changes within Google Analytics. It takes a lot of time to get used to something. I think Universal Analytics is great and GA4 was so radically different in its measurement and its interface that it took a lot of getting used to. You do get used to it, though, and I can attest to that.”
Some SEOs will be thinking of this as an opportunity to consider other tracking tools. Have you considered learning something entirely different instead?
“I’m fairly brand loyal to Google. Certainly, when there were things that GA4 couldn’t do for quite some time, I understand the motivation to try out things like Matomo.
Personally, I am not doing much with any other analytics platforms, so I can’t really speak to that. I totally understand the desire to, though, and it’s going to make the industry a bit more exciting if there’s a bit more competition.
With that being said, I hold out for GA4, and I think it’s going to do what it needs to do eventually.”
Would you recommend using Data Studio, or another third-party aggregator, to analyse your data and make those comparisons?
“Yes. I think you can do it without Data Studio, but the year-on-year comparison is going to be quite fiddly. Once you’ve got a year’s worth of GA4 data, it will become much easier to do that within GA4 itself and not necessarily need to rely on something like Data Studio.
However, that is what Data Studio is for: pulling together all of the data from all of Google’s platforms and all other kinds of third parties. That’s what’s cool about it. I don’t think Data Studio is compulsory, but I think it’s going to be necessary for this specific comparison exercise.”
What shouldn’t SEOs be doing in 2023? What’s seductive in terms of time, but ultimately counterproductive?
“I would say that, especially when you’re short on time, we can all be guilty of taking the view of just one expert as if it’s the gospel, or definitely best practice. That’s quite unwise. In an area like SEO, where so much of our job is theory, it’s good to get your information about a subject matter from a range of sources.
If they back each other up, then that’s great and you know you’re onto a winner. Blindly following the instruction of one industry expert, however, is possibly unwise. I’m not talking about any one individual. We have an industry full of hugely talented people, and I think we can all learn a lot from collating that information and thinking about it critically, rather than just take taking one thing as read and running with it.
Someone’s experience, no matter how vast, is based upon what has personally worked for them, and their situation doesn’t necessarily apply to your situation.”
Natalie Slater is a Freelance SEO Consultant, and you can find her by searching Natalie Slater (formerly Natalie Mott) on LinkedIn.