November 29, 2023 (Updated: January 24, 2024)
Explore the best ways to approach the newest version of Google Analytics, avoid massive frustration, “learn to love the bomb,” and survive GA4 and E-E-A-T signal optimization in this exclusive interview with veteran SEO, Ruth Burr Reedy.
She covers the steps you can take to implement actionable, specific changes to your site and strategy so you can capitalize on Google’s emphasis on “E-E-A-T”.
If you haven’t had a chance to tune in yet, be sure to catch the replay.
When it comes to Google Analytics 4, Ruth says it’s important to treat it as a totally new and different analytics platform that needs to be properly configured and “onboarded” before you dive in. So, focus on configuring GA4 as the ultimate data collection tool and leverage the nearly seamless Google Tag Manager to ensure all your “events” are properly captured. Several things to consider here:
Ruth also suggests using Google’s Looker Studio to create charts, tables, and visualizations from GA4 data for better visibility. Connect GA4 to Google Big Query and refer to this GA3 template for GA4 data as an example.
Regarding actionable E-E-A-T, remember:
And when it comes to establishing this authorship, make sure your schema markup matches your “author” call-out on the page. Add “Same As” elements for your author’s other works, profile pages, bios, and citations (think FindMyDoctor and others). Lastly, invest in your writers. Get them published on other sites, in interviews, and on podcasts. This will change link-building into “authority building,” helping your brand build credibility and lasting authority in your industry.
Jeremy Rivera: Hello, I’m Jeremy Rivera with CopyPress, and I’m here with Ruth Burr Reedy, and she is a powerhouse in SEO. She’s been around for a long time, knows a lot of things and works, has worked with a lot of great teams. If you do need an excellent SEO on your team, she is in the market. So hit her up on socials.
Today we’re going to be talking about what’s on our mind. And we had a conversation last week, I was struggling with something and she really. Has helped me on a path on something that’s been really jabbing at me and that’s Google Analytics for implementation. So, let’s jump into that.
I know there’s some other things we were thinking about in the helpful content area, but let’s talk about GA4 first. For me, it’s a bit of a prickly pear and I’ve been putting it off for a long time, but I think you kind of helped therapize me a little bit. So, how do you look at the transition to GA4 and not start crying?
Ruth Burr Reedy: That’s a great question. I mean, part of how I’ve been able to do it is that I didn’t put it off until 2023.
So, I’ve had I’ve had some opportunity to cope and to come to terms with it. You see all of these, like, polls going around and these conversations going around, like, SEOs hate GA4…and if you look at the reasons that we hate it, it’s because it doesn’t have as many features as UA. We hate it because it takes too much time to set up.
And we hate it because we can’t find anything. And I think all of those problems boil down to the fact that for the first time, really, in 15 years with Google Analytics, we are having to treat it as though we’re onboarding a new analytics tool. And for me, that has been the helpful mindset reframing. If you look at how UA was, the analogy I like to use is if it’s Pokémon, UA was Charizard.
It was fully evolved, it had 15 years to grow and grow and evolve into its iterations. It was a fully grown analytics tool. It was very robust. It had a lot of features. It had a lot of support. There was also a ton of information on the web on how to use it.
Unless you had a brand-new baby website, most people were coming into existing instances of UA that had evolved naturally over time from classic analytics and even, you know, from Urchin. And there was still some of that remaining Urchin code from way back in the day that was part of the data collection mechanism.
So, it was one tool that had grown a lot, and GA4 is a Bulbasaur. It’s a different Pokémon, and it’s a baby. It is a brand-new baby tool. And I think that a big part of why people hate it is because we don’t want to have to onboard a new analytics tool. We want it to just work, and it to be set it up. – Ruth Burr Reedy
Jeremy Rivera: Yeah, I think back and, you know, my employer paid for me to go through the GA3 training process. And there was, like, a week and there were sections and tests and, you know, I became very familiar with the difference between “Sessions” and “Page Views,” where this part of it is using page views and this part is powered by sessions. Through my long process I think there’s consultancy, you know, freelancing, looking at different UA profiles.
I became very proficient at looking at the “when and how did that traffic spike occur?” And what channel came very proficient? Okay, well, you do this and, you know, you unlock this segment, and you can drill down to this. So, I think there’s certainly a deep discomfort with the layers of familiarity that I had with GA3. I could make it sit up and beg.
I could separate the channels and I could see the pages that I want. Here’s the channel data that I need. And I knew exactly where that was every single time. I knew what limitations there were, you know, like you had to go through extra steps to get specific click tracking and you had to add extra event tags.
And I think the data collection we kind of just assumed, “Oh, it’s always going to be poor.” So, it’s going to be what it is and just kind of exist with it. It feels like I have seen examples where, like at MozCon, they did some fantastic deeper dives into pathing and the collection of data — that was impressive. But it seems even more frustrating for me because I don’t seem to be able to get there, you know, I don’t seem to be able to get out of the boxes to navigate. I have to move the data around and tell these reports, “Okay, use this data here and go through this specific section to create something that’s usable.”
Yeah, it’s very frustrating because I know already that small business owners struggled out of the box just with Google Analytics 3. And I cannot imagine that any small business owner is really going to be able to use GA4 out of the box.
Ruth Burr Reedy: I agree with that. And that is something that I’ve been thinking a lot about as well.
GA4 is designed to be more customizable. And if you are somebody who knows a lot about web analytics and is willing to take the time to set it up and customize it, you can get GA for doing stuff more or less exactly the way that you want it, in a way that UA couldn’t quite do. Like, you could hack it to do that, and, you know, people like you and me got very good at doing that.
But now we have to learn to do it again, which, you know. It is what it is. We’re SEOs. Sometimes everything we learn how to do stops working and we have to learn to do new things and that’s okay. But yeah, because of that there is less out-of-the-box functionality. GA4 requires some setup and that is something that I think Google has done themselves a little bit of a disservice.
I understand, and I empathize with the reasoning behind it, but a lot of their messaging and marketing around the UA to GA4 switch has been, “Oh, it’s super easy, like, complete this setup wizard, roll over this stuff, it’s all automatic, if you don’t do it, we’ll do it for you,” and I think there are a couple of problems with that.
One is that there is more to it than that. So, there’s sort of a cognitive dissonance between what you’re hearing from Google and what you’re seeing when you log into the tool. And then the other is that, as somebody who has spent many years fixing people’s analytics for a living, a lot of people’s analytics, like UA Data was garbage and there’s this huge opportunity for your data to not be garbage anymore because you’re forced to reset it up.
But if you’re just rolling over what you already had, it’s garbage into garbage. And you know, I think a lot of us small business owners don’t really know that their data is garbage, so they’re probably not concerned about that.
One of the things that you said just now that really stood out to me is you knew how to go in there and find what you wanted. And I think that that’s another thing that is causing so much frustration is that most people, not me, but most people don’t just log into Google Analytics for fun.
Like poke around and see what’s going on. Usually when you open up GA, you are looking for a specific number. You’re looking for a specific report. You have a specific data problem that you need to solve and you’re having to spend that extra time to find it and take those extra steps that you didn’t have to do in UA, because either it was there out of the box, or because you’d already customized it.
So, I think a part of the reality of GA4 is that you have to take some time with it when you don’t have an immediate data need to customize it, so the next time when you do need that data you can just get it. But I think if you’ know if you’re somebody without a lot of time and without a lot of analytics expertise, I think that’s a really steep learning curve.
Jeremy Rivera: Let’s say that you’re given a fresh site and we’re thinking about, [00:09:00] you know, okay we want to think about it as onboarding our new software. What are the first two things that SEO should do to their, their new GA for implementation to help them transition and see the power? And have some hope that it’s not always going to be this painful.
Ruth Burr Reedy:
The first thing that people should do that many, many, many people don’t ever do is make sure that data collection is actually happening on every page. Make sure that the code is actually firing on every page. And that includes landing pages and subdomains. – Ruth Burr Reedy
You know, so taking a minute to make sure that your code is implemented properly is something a lot of people don’t do and didn’t do — whether that’s running a crawl with a tool like Screaming Frog and looking to make sure that your GTM container is firing on every page, or if you’ve got it directly hard-coded in there. Make sure that tag is present on every single page.
I also like to run a crawl and do a custom extraction and find pages that don’t have the code. Often that’ll be landing pages and especially if you have something that’s on a separate backend for whatever reason. And that’s just analytics housekeeping, regardless of what tool you’re using.
The second step I’m going to cheat and make it a two-part step. I mean, what you will want to do is set up enhanced measurement and enable enhanced measurement as part of the setup wizard that GA4 tries to make you do. Not all of the steps in the setup wizard are equally important but setting up enhanced measurement is.
Enhanced measurement is going to do what it can to automatically detect and track various events. Everything in GA4 is an event. It could have been called something else, because “Events” meant something different in UA, and words have meanings, but it isn’t, and here we are. Everything’s an event passing parameters into GA4.
Turning that on, it will do its best to automatically detect things like form submissions and things like clicks on outbound links (you know, some of the basic things that most people want to track) and start collecting those as events. The second part of that, that I think a lot of people aren’t doing, is actually QAing that and not just pushing the button and saying, “Oh, data’s coming in, it must be fine,” but taking a minute to say, “When we have a file download, you know, there’s a specific way,” because there are so many different ways to build a website.
And there are so many different ways to code a lot of very common actions. And I think a file download is a great example. Enhanced measurement file downloads events are basically tracking when you click on a link that’s internal that ends in a file extension like pdf. This is one way many people have file downloads.
But if you have a file download where you’re actually clicking on an HTML URL that takes you to another page — especially if that page is on a subdomain or on another site entirely on a different server — and then prompting them to download from there, that to you is a file download. But GA4 is not going to see it that way. So, you have to take the extra step to hard code it.
But there are also things that can fire what looks like a form submission that aren’t. So, one example is, and I think that Facebook has made some changes to this, but the Facebook ad tracking pixel.
It’s technically a form submission. It’s submitting that third-party data via the pixel to Facebook. When people first implement enhanced measurement, a lot of them are seeing a huge inflation of form submit events because that pixel is firing a set event. It’s an easy thing to fix. And that’s one of the nice things about GA4 is that it’s a lot easier to edit data without having to go into the hard coding of it. You can edit data as it’s being passed into GA4, which is awesome.
But you do have to do that. And if you don’t check on it, then a couple of weeks later, you’re like, “Why does my data look like this?”
Jeremy Rivera: Got it. So, it’s the enhanced measurement. Like, if you inherit a GA4 that somebody set up, is that something you can go into the settings and trigger and go through the wizard after the fact?
Ruth Burr Reedy: Okay, yeah. And you can turn on not just enhanced measurement itself, but the individual events, whether or not you want them to automatically track. You can toggle those on and off within your data stream settings. So, it’s easy to do, but you do have to do it.
Jeremy Rivera: It does seem like GA4 is much more closely married to Tag Manager than GA3.
So perhaps taking some time to take a tutorial and go back over to GA3 because I know setting up, make sure appropriate triggers are configured and you have the right measurement ID to connect this profile to this profile. Because I had a situation where someone had made a GA for instance, for a backend tool or content analysis tool. And it was technically tracking. Even though it’s a single page application, it was using a different container code.
Then, you know, the container code that the marketing site was using. So, it seems like Google Tag Manager is something we would want to recommend brushing up your skills on to get that more thoroughly implemented as a compliment to GA.
Ruth Burr Reedy: Yeah, I think that the built-in GA4 event tag in GTM is pretty good. You just would need to (especially if you’re firing a lot of events into UA via GTM) make sure that you create a GA4 event to do the same thing. And that’s it, that’s another thing that is really grinding my gears about the Google-assisted transition from UA to GA4. They have essentially tried to sort of shoehorn the category action label framework of event tracking, which GA4 purposely moves away from, which I think is great if you automatically roll over your events from UA into GA4.
Then, the event action will become the name of the event and will be firing parameters called “Category” and “Label” that are populating whatever values they are before. But a lot of times, there’s an enhanced measurement event that will track the same thing more effectively. And, you know, category action label like the problem with it historically has been that it’s three parameters that mean anything.
And every single website and every single marketing team was using them differently. Often within the same website you could watch the evolution of different naming conventions over time as various people had built events. Again, like garbage-in-garbage-out, you have an opportunity to not do that.
But it requires setting things up and having some thought about how you are going to fire events and what parameters you are going to collect with those events going forward. Because now the problem with clinging to category action label is that GA4 is going to continue to evolve away from that event tracking model.
So, you’re just going to have to keep reverse-engineering it, and you won’t be able to scale and move forward as the tool evolves.
Jeremy Rivera: Got it. And something that you suggested, I guess, would be suggestion three. We treat GA4 more of a data collection engine. And, look, I hate to say it, Looker instead of Google Data Studio.
Ruth Burr Reedy: And it doesn’t have to be Looker.
One of the great things about GA4 is that BigQuery integration is included, and you don’t have to have 360 like you did in UA. So, you know, you can enable BigQuery, especially if you don’t have hundreds of thousands of sessions a month. – Ruth Burr Reedy
It’s not that expensive. And then you can push BigQuery into whatever data visualization tool you want. It can be Looker, it can be Tableau, it can be, you know, whatever you want to use. But I do think as of right now, this has always kind of been the problem with GA is that GA was built by engineers and often has a sort of engineering-focused layout in terms of the reporting capabilities. I think right now GA4 is really good at collecting data and less good at visualizing it.
So just pushing it into the database tool where you can mix and match things the way that you want, I think it’s going to make you happier.
Jeremy Rivera: Yeah. I mean, it already made me happier. You gave me this advice and I asked around and I found somebody who made a Looker Studio. That is the classic GA3 model. So, it’s got the, you know, behavior and layout approach of GA3.
And this made me so much happier than I was with GA4 out of the box. So just knowing that, hey, you can, make a looker studio that looks like GA3 and can do a lot of the similar things. That makes me a lot happier. And I think you’re right — that the data was bad in a lot of situations with Google Analytics 3.
We just kind of pushed along to get along.
Ruth Burr Reedy: It’s a huge opportunity right now. I think the more you can treat GA4 like you are onboarding a brand-new analytics tool, it’s just like looking at a clean slate. What do we want to track? How do we want to track it? Let’s get some rules in place around things like naming conventions and make sure that we’re using the description field when it’s available. So, when we create a custom report, somebody else could look at it and know what it’s supposed to do.
We’re being forced almost to onboard a new tool and to rethink our data and make it work for us going forward. And I think people are wasting that opportunity right now, which makes me sad. They could be really excited about a new tool, new data, new accuracy, and, you know, making your data exactly the way you want it. But instead, they’re sad because it’s not the way that it was before, when for a lot of people, the way that it was before was also jacked up.
Yeah. So, get excited, yay!
Jeremy Rivera: A new tool! Yeah. Dr. Strangelove learned to relax and love GA4.
Ruth Burr Reedy: That was the title of my talk at Engage this summer. Yeah.
Jeremy Rivera: I love it. So, thinking about transitioning topics, but also, you know, about data and signals. There’s been a lot of people talking about E-E-A-T: expertise, experience, authenticity, and trust — all of these things. But those type of concepts that SEOs have been talking about in sessions, how does that actually translate into usable signals, editable signals, and physical changes?
You can go to your site and say, “Okay I want to change this thing on my site, that’s going to help Google understand this part of it,” and/or not do something specifically that is likely to fall foul of, you know, helpful content parameters.
So, for me, a lot of the discussion around helpful content has kind of been overblown expectations of what Google is going to be targeting with that. And I’ve been very closely watching, you know, what Lily Ray is discovering and trying to spot. An update came out here and we think these guys, you know, crashed right after that kind of analyzing postmortem.
What were the similarities there? What’s your take on the Helpful Content updates and what do you think it breaks down to as far as actionables that we can really step away from? You know, super actionable, like “Oh, I need to be more authoritative.”
Ruth Burr Reedy: It’s the new just-build-good-content. Okay. But what is good content? What does that mean? How do you know?
For me, SEO has always, and especially when it comes to interpreting what Google says, it comes down to understanding the human. Readable quality signals, human readable traits of quality that Google is trying to exhibit to its end users and then figuring out or at least hypothesizing about what the machine-readable equivalence of those human readable quality signals are.
And I think that’s part of the reason why conversations around E-E-A-T can kind of go around in circles or not be super actionable. This is because E-E-A-T is a way of thinking about all of this for people, for person brains, for marketers, and content creators — all of whom are people. – Ruth Burr Reedy
But. The algorithm is a machine and, you know, algorithmically, there’s no reason for Google to try to build a human brain machine-wise. Human brains are not actually very efficient in that way, which is fine.
But the way that Google (and when I say Google, you know, it’s really like 100 of the algorithms — many, many algorithms stacked on top of each other in a giant trench coat) collects, processes, understands, and returns information is looking for, you know, wherever possible. They want to build that in, they don’t want manual processes, they want algorithmically to build it in. And this is something that I think E-E-A-T and the rise of machine learning-generated content are inextricable.
You know, originally Google said no machine learning-generated content, then they walked that way back. I think it is because E-E-A-T is difficult to say with confidence that you can (at an algorithmic level) detect whether or not something was created using AI.
AI, it’s not AI, it’s machine learning, whatever.
I’m a curmudgeon about this, but AI is too shorthand. I don’t think that they can do that with enough confidence to come out and say, don’t do this. Instead — and I think this is smart — they’re emphasizing quality signals that are difficult to do with language learning models.
I think that’s part of where the experience point comes in. So, thinking about machine-readable signals that you actually did this, that you actually created the content. You know, I mean, part of what’s interesting for that to me is that people have been creating content that they know nothing about for years and years.
There are many, many, many content marketing agencies in the world. You know, when I was 22, I wrote a blog about how to find and win contracts with the federal government. Do I know about that? No, I don’t. But I can do some research and create some content about it. And so can a predictive content tool.
So, looking for the things that not everybody is saying you can do that on a machine detectable basis. If you are using tools to do entity extraction at scale, this is looking at what the ranking pages are for a given term or topic set and then extracting the common entities. But it’s also taking a look at, like, the Wiki Data entry around those same topics.
What are the entities that are not commonly discussed that you can add in taking a look at who is creating the content? I think, and this has been true for several years now since E-A-T before the extra E was added, but I think even more now, is treating your content producers like brand terms. You know, treating your people like your product.
So, it’s not just saying here’s this person, here’s their bio. Anybody can say anything about themselves. It’s very easy, but how are you building machine readable citations across the web around this person? I think we’re going to see a rise in sort of a new wave of SEO-based influencer marketing as people who do have demonstrable real-world expertise that can be shown with citations across the web. – Ruth Burr Reedy
And I think links, again, then become important in a different way. They become important almost more as an expertise signal rather than a signal in and of themselves.
Jeremy Rivera: So, slowing that down and kind of breaking it apart to make sure. Because I just talked to somebody, and they had a little box that said, “Michelle Wolf, PhD” and “Learn more about her here.” That was in the text. And then I looked at the code and realized that they had a separate scheme of texts that had a totally different name. And I figured out that three quarters of their 3,000 articles were attributed to one person.
And she was no longer with the company. And because they hadn’t gone into WordPress to change the author internally, they had just added, you know, a special dynamic coding to put the PhD. They intentionally tried to up their E-E-A-T and show expertise and say, “Hey, this is because it’, so y ‘s a Your-Money-Your-Life (YMYL) niche.
They knew what was important, but they put it into the text. But they, they were sending two signals for that same author for that.
So, one, be making sure that you have one signal for your author, and two, that you have some aspect of it that’s connecting to that social graph. Have links to their LinkedIn profile, have links to their personal site, have a schema markup saying, “Hey, this is the same entity.”
Ruth Burr Reedy: Yeah, I think the same way about markup around your authors, especially around their other citations. And I think with YMYL, you know, a lot of medical doctors have pages on Find My Doctor or Physicians Today or whatever it is to build those citations. But the thing with schema is I think that so many people still see semantic markup as a page-by-page or a page-template-by-page-template activity where they’re like, “Okay, this is a blog post, so I’m going to implement blog posting markup.”
But I think the real opportunity with semantic markup now is to interrelate those different pieces and link them to each other via internal linking and IDs. And that can help you create a knowledge graph within your own website that is explicitly linking these pieces of knowledge together to make them more machine discoverable. Then pulling data from other places in the web to demonstrate this is a real person and here’s how you can tell, here’s their digital footprint.
That’s harder and more expensive to fake. But taking the time to connect it all together, I think, is another missed opportunity.
Jeremy Rivera: And I think that’s another actionable piece of the puzzle is to stop thinking about your content as just a written item.
Start thinking about, you know, the aspects of marketing that, traditionally in the seventies, a company would get behind a particular person and do thought leadership and have an industry figure.
And that’s not something where you can just publish things on a site and then you’re an authority. That person needs to be out there in different ways, pushing. You know, instead of these really exhausting conversations about, “Okay, we need to do link building because we need link authority to be sent.”
And then the business owners, “Well, how do we measure that?” Well, there’s this domain ranking metric that kind of, sort of, works. And then the business owners, “I want a DR of 75 and I’ll pay $5,000 for, you know, a DR of 75.”
Whoa, whoa, whoa, wait. We need to tell Google that we’re relevant. You need to get out there in some way.
So, it’s almost like rehashing and changing the link-building concept into authority building and the processes of building those authority signals. And if we remind people to treat authors like a resource that you can build up, not just by attaching pieces of content but by getting their pieces of content on other sites. If the link isn’t necessary, you’re not getting a link back to the site. That person building up their authority transfers it. And that’s a change of mind, of thinking.
Ruth Burr Reedy: I love that you brought up the seventies because, to me, I think one thing that I’ve had success talking with business owners about is that link building at its heart is a set of disciplines that has its roots in PR and brand building, and in practices that are older than the internet.
I think too often we divorce links from that, but if what you’re doing is building a brand, if what you’re doing is not just, you know, spraying out your press release to a million websites, but real public relationships where you’re doing outreach and you’re promoting your business. You’re promoting the people behind your business, especially in any kind of services-based business, including most medical sites and practices, which are [in the] YMYL category.
The more that you are taking the time to build that brand on the web and taking an extra moment to make sure that’s happening, you’re getting those citations in a machine-readable way. A lot of times people are already doing those activities, but they don’t think about them as link building — but they are.
It’s a mindset change more than anything else.
Jeremy Rivera: Absolutely, I love it. Thanks so much for your time today.
Are there any last takeaways on GA4 or on thinking about how you change link building to fit into this new schema of how to approach authority?
Ruth Burr Reedy: Really, I mean, one of those is a way bigger question than the other.
With GA4, I would say, let go of UA. It’s gone. It’s okay to feel sad about that. Let it go. Breathe it out. Namaste. We release it back into the world. It no longer sparks joy, and that’s okay. We honor the service that it did for us, and then we release it into the world.
And really take a minute to think about onboarding GA4 as a new tool, even if you’ve been using it for a while. Set aside four hours. That’s my challenge to anybody who watches this. Set aside one afternoon, half a day, four hours, to look at your data collection, to think about what you want to report on, and how you want to label that information.
Just set it up. Build some custom reports. Build some dashboards. It really doesn’t take that much time, but it takes time, and it takes intention. And the reason that you’re annoyed is because you’re only logging into GA4 when you need something else. If it was not GA, if you were onboarding some other third-party analytics tool, you would have to take the time to onboard it.
You would have to take the time to define what you’re collecting and then make sure it’s collecting and reporting the way that you want to. If you can view it through that lens and do that setup one time, you’ll be so much less irritated with it every other time you do it. I know, I know you don’t want to.
But you can spend four hours once and be less annoyed forever, or you can keep being annoyed. And you know, that’s a personal choice that you have to make. I know which one I would pick.
Jeremy Rivera: Okay. Thank you.
Ruth Burr Reedy: And then yeah, I think with authority building, really take the time to. One of the things that I think is great about machine learning models is that you have an opportunity. And a lot of content tools are incorporating this into, hey, does this piece of content say what I think it says? Can you tell who wrote this?
You know, use those tools to figure out how easy it is to tell who wrote your content, what they do, and who they are. And if you have that expertise, then look at whether or not Google can tell.
And if not, Think of it as brand building. You’re just building a brand, or you’re doing it around a person instead of around a product. And you’re a marketer. You know how to market. You just have to start thinking about your experts as part of your offering.
More from the author: