Where are we at with AI Overviews?
Where do we stand with AI overviews on the SERP? What does the data show and what doesn't it show? How are these overviews reshaping the search results by including more sources, introducing subtopics beyond the main SERP content, and why do certain sites consistently appear in these overviews while others don’t?
And let’s not forget the all-important question: How stable are those URLs in AI overviews, and what does it mean for your SEO strategy?33
Wix Studio’s dynamic duo, Mordy Oberstein and Crystal Carter, team up with Mark Traphagen from seoClarity to break down the intricacies of Google’s AI Overviews.
Plus, Rich Singer’s eye-opening study on the overlap between organic search results and AI-Overview comes under the microscope, along with juicy details and practical tips for staying ahead.
Plus, learn how to sift the SERP’s most recent feature and track AI Overviews!
So join us as we summarize the AI summaries on the SERP and sum up what it all means for your SEO practices on the SERP’s Up SEO Podcast.
Episode 106
|
October 16, 2024 | 39 MIN
This week’s guests
Mark Traphagen
Mark Traphagen is Vice President of Product Marketing and Training for seoClarity. His previous careers as a salesman and then a teacher were excellent preparations for what he does now. Mark's deep dive into online marketing began about 14 years ago when he was tasked with turning around a failing independent bookstore. He took the store online and developed practices of content marketing, influencer relationships, exceptional customer service, and community building that brought the store to profitability in two years. Since that time he has dedicated himself to learning what really works for online business success. He is a sought after speaker at conferences such as PubCon, State of Search, and SMX, and a featured writer for major online publications such as the Moz Blog, Search Engine Land, Marketing Land, and a VIP Columnist for Search Engine Journal.
Transcript
Mordy Oberstein:
It's the new wave of SEO podcasting. Welcome to SERP's Up. Aloha, mahalo for joining the SERP's Up Podcast. We're pushing out some groovy new insights around what's happening in SEO. I'm Mordy Oberstein, the head SEO brand here at Wix and I'm joined by the amazing, the fabulous, the incredibly uncompromisingly unparalleled head of SEO communications here at Wix, Crystal Carter.
Crystal Carter:
That was very kind of you to say, thank you, and yeah, you too.
Mordy Oberstein:
You're welcome.
Crystal Carter:
Back at you. Back at you. How about that?
Mordy Oberstein:
Go for it. Repeat it right back at me.
Crystal Carter:
I can't say it back. I don't remember. People have different ways of processing things. I find if people talk at me for too long, my brain goes a little fuzzy. I'm not going to lie. So I have to-
Mordy Oberstein:
I have that with my kids. Sometimes they just go on, I'm like, "I lost you like five minutes ago, dude."
Crystal Carter:
I need it very, very structured, so I find these are the things. Everybody learns in different ways. It makes the world a many splendorous thing, that's my opinion.
Mordy Oberstein:
And thus the SERP's Up podcast is brought to you by Wix Studio where you can not only subscribe to our SEO newsletter, Searchlight, over at wix.com/seo/learn/newsletter, but where you can also track search performance with our deep integration with Google Search Console. Get immediate insights and dive deep into large quantities of Google Search Console data with our Google Search Console Analytics Report. This as we pick the data on AI Overviews apart to help you make actionable sense of it all. Where do we stand with AI Overviews on the SERP? What does the data show and what doesn't it show? And what do you actually do with all of this data? To help us sift through the data itself, Mark Traphagen of seoClarity will clue us in on where things stand on the SERP as AI Overviews, I'm going to just call them AIOs at this point, AIOs are concerned.
We'll also sort through some tools that help you track the AIOs, and of course, we have your snappiest of SEO news and who you should be following on social media for more SEO awesomeness. So generating, generating, generating. If I'm not done generating in five seconds, wait a little longer. And here's our generative sum up on the surmising of the SEO tool providers and what they've done on Google AI SERP Summaries on this the 106th episode of the SERP's Up Podcast. This is a deep pull, by the way, the if I'm not done generating in five seconds, wait a little longer. That's an Ace Ventura, if I'm not back in five minutes-
Crystal Carter:
Oh, yes. Yes.
Mordy Oberstein:
... wait a little longer.
Crystal Carter:
There we go. Yes, yes. That's a good one.
Mordy Oberstein:
Woo, do not go in there. We were just talking about Ace Ventura, weren't we?
Crystal Carter:
I know, but I can't quote it chapter and verse, I'll be honest.
Mordy Oberstein:
It's like gospel to me.
Crystal Carter:
I remember all righty then. That I got. That's there.
Mordy Oberstein:
I don't know, I was just, I don't know, 10 when it came out. When you're a 10-year-old boy, that thing is like gold. Gold. Anywho. You know what's gold? AI Overviews. They're gold for tracking.
Crystal Carter:
Oh, yeah.
Mordy Oberstein:
Or not, who knows? No one really knows.
Crystal Carter:
Oh my gosh.
Mordy Oberstein:
The AI Overviews, which produce a summary of whatever you're searching for with theoretically a bunch of URLs, pulling the data on this stuff is really, really tricky. I know firsthand, I'm not going to get into it, it's super tricky. And it's really hard to get a deep look and there's a whole lot of limitations and it's hard to know. When you're looking at a piece of data or a study about the AI Overviews, it's hard to really know what to do with it, what's accurate, there's so many numbers all over the place, what's what, and whatever.
And then of course, what do you do, actionably speaking, with all of this data, which is why we're going to go with a little bit of a different format. Usually, Crystal and I wax poetic on the topic a little while. We talk, we schmooze, we have insights, then we go to a guest. This time though, we are going to get right to our guest, seoClarity's own Mark Traphagen because he's had a deep dive on the data around AI Overviews as he tells us where we stand with AI Overviews and the data itself. Here's Mark.
Mark Traphagen:
Here at seoClarity, as an enterprise-level SEO platform, we've been keeping a close eye on Google's AI Overviews since their infancy as Search Generative Experience in Google Labs. So we were completely ready in our platform when Google went full public with AIOs in the U.S. in August to track them and report on them in greater detail than any other enterprise platform. That also meant we had access to unprecedented levels of data on AIOs. And our data science team has been hard at work on covering them. Let me cover just a few of the most interesting insights. First and most importantly, in a LinkedIn post a few weeks ago, we were the first to break to the world the news that Google had shifted sometime recently to including more sources in AIOs that matched those on the first page of Google. We saw that almost always now, 99.5% of the time, at least one top 10 organic source appears in an AIO for the same query.
Now that alone is a dramatic shift as all agreed prior to this there was almost no correlation between top rankings in traditional Google and appearing in an AIO. But it gets better. Over half of all AIOs contain the number one organic ranked source, and wait for it, over three-fourths of the time, all of the AIO sources are now from the top 10 organic results. And what that means should be obvious: AIO SEO is now just SEO. Improving rankings for your content where an AIO is shown is your best shot at getting mentioned in that AIO. We also recently revealed that 96% of all AI Overviews have informational search intent. That's relevant if you're using top-of-funnel content as part of your marketing strategy but means that e-comm and local business sites can breathe a sigh of relief for now. Now we'll have plenty more insights about AIOs in the weeks ahead.
Mordy Oberstein:
Thanks so much, Mark. Check out seoClarity and all the work that Mark's done there around AIO Overviews. We'll link to a bunch of this stuff in the show notes. Okay, so it's really interesting. Okay, so first off, when he put out that data that 99% of AIO Overviews contain at least one result from the organic results, people misunderstood that. I saw around the web people... They were saying, "Oh, the AIO Overviews, the links overlap 99% of the time with the organic results." It's one result is in there in 99% of the AIO Overviews. So that nuance in all of this stuff is so tricky. Just nice piece of data, by the way, and then 96% of all AIOs being informational, yada, yada, yada, yada.
Crystal Carter:
Yeah, no, I think that's really important to think about as well, in my own, and I think that that's really, really useful. And I think that what we see is that sometimes the AI Overviews aren't generating in real-time, that's the other thing. They're pulling from something already, so you go to the page and it's not generating in real-time, it's just has an AI Overview on the thing already. Which tells you that they're pulling from some stored data, and all of these machine learning systems, they have a training set. And the training set for an AIO Overview of the SERP, it would stand to reason that the training set would be the SERP, would be Google's repository of search results. So they have search data for years on content and it makes sense that they would pull at least some of the information from it.
And I think that there's a lot of different data around there, and I think it's really fascinating as a topic for search researchers right now because it's rapidly changing. Since they've added AI Overviews into the EMEA space, we're going to see a lot more people chiming in on that data, which is really great to see. My first initial hunch is that it's building on the featured snippet data set. Because what I very often see is that you have an AI Overview for something that has predictive text. Like the predictive search will have an AI Overview because they know that search and they can make an AI Overview for it. And as I said, they're not always generating it.
But the way that the featured snippets work, that system is slightly different from the core plain blue links system. That system has always included an AI element and has always included a remixing element. So you'd have places where on the page it would show bullet points, but on the list on the SERP it would show numbers, or it would show numbers on the page and then it would show bullet points on the SERP for instance. So to me, it would make sense, and certainly coming from a SaaS angle, from working with a SaaS team, it would make sense for Google to not completely reinvent the wheel. For Google to use some of their existing frameworks in order to build out their AI Overviews, especially if they're trying to do it at a pace.
Mordy Oberstein:
Right, and that's what's interesting. So you have so much data on this. You had Rick Sanger who partnered with Authoritas looking at the overlap between the organic results and the AI Overview links, and something around, I think it was 45% of the URLs in the AI Overview are also ranking in the top organic results. I think he looked at the first page, top eight or so. And then when he clicked through to the related search features, people also ask, people also search for, related searches, that number jumps up to 64%. I think that makes a lot of sense because... I can't remember who I was talking to about this. It might've been Garrett Sussman. It doesn't matter, either way.
When you're looking at the topic, let's say you're searching for how to build a log cabin... It was Garrett. You're going to get, let's say on that SERP, you're going to get, I don't know, eight or so results on how to build the best log cabin. Let's say just for argument's sake, they're all about how to build a log cabin. There's no other search intent. All of those URLs about how to build a log cabin. When you now have the AI Overview telling you how to build the log cabin, yes, it's going to include some of those URLs, but how to build a log cabin because what that content's talking about, but the AI Overview breaks into a subtopical breakdown.
So for example, it might say, "Step one, select the right material, use whatever, maple wood or birch wood," and it might link out to what birch wood is. That URL is not going to be on the main SERP because what birch wood is has nothing to do with that main SERP. So when you go subtopic inside of the AI Overview, it's going to link out naturally to things that should not even be on the actual SERP itself. So when I see numbers like 50 or so percent, that to me makes a heck of a lot of sense because there is alignment. But when you go subtopically, there's not alignment. You're pulling in topics from somewhere else.
Crystal Carter:
Well, absolutely, and I think that the other thing that the AI Overviews, I think, certainly, in the long run, maybe not in this first iteration, but certainly, in the long run, are hoping to do, is hoping to solve some of the challenges that Google has been trying to address for years. So Google have said themselves that very often if somebody makes a query, they have to make eight questions before they can get to an answer. And the multidimensional search that you're talking about where they would also have some of the subcategories in that question, and whether it's birch or oak or pine or whatever on the log cabin, what kind of log, for instance. When you're thinking about those questions, these are things that Google have been thinking about for ages.
So this was something they were trying to address with MUM, for instance. This is something that they've been trying to address with lots of different ways with some of the disambiguation they've added to the knowledge graphs and the featured snippets where you have the drop-down where you'd say somebody would search for blogging and they'd say more about this topic, things you should consider, all of those sorts of things. Google have been trying to address this in a lot of ways, and I think that they're trying to see if they can address it with the AI Overviews. And it stands to reason that yeah, you would have a more rich collection of search results.
Mordy Oberstein:
So I want to switch topics really quick because I want to talk about why there are so many data discrepancies among the AI Overview data. You have certain tools saying X, certain tools saying Y. So first off, one of the things to look for is consensus. So for example, generally speaking, most of the data studies that have come out, there are obviously outliers, are basically saying as of the time of this recording, by the way, that AI Overviews appear at around or anywhere between 10 and 20% of all SERPs. I think SEO ranking had it at 12%. I could be misremembering, but it's generally around there.
So look for a trend. I wouldn't rely on the exact data from the exact study the exact way you normally would with an SEO study because I feel like pulling the data here is very, very difficult and the tools aren't really built for you doing multiple click-throughs. And then Google's also changing formats all the time and getting it right is incredibly difficult. So while the tools may be built to scrape rank, they're not built to scrape AI Overviews the same way. I'll prove it to you. If you look at how when AI Overviews came out, at least in Search Labs, and you look at when the SEO tools started, actually being able to track them we're talking months later, which meant they had to do months of development in order to change how they're tracking things. So I think that one of the reasons why you're seeing a little bit of discrepancies is A, because the AI Overviews are changing all the time, and B, pulling this data is just wonky. It's a little bit tricky, so look for trends.
Crystal Carter:
It's also tricky too because there are so many different... I think that we're not seeing how consistent it is. So sometimes I think you've been looking at how often the AI Overviews are switching links, and certainly, for some of the-
Mordy Oberstein:
Yeah. I'll get to that. Yeah, yeah, yeah. I'll get to that.
Crystal Carter:
And some of the AI Overviews that I've been tracking, which includes some of my own content, I've been seeing those, and there's not entirely consistent on some of those. So for instance, one of the things that I've seen is that they'll have the core AI Overview and then they'll have a few topical links, a few recommended links on the right. And sometimes what you see is that the recommended links on the right aren't referenced in the AI Overview. So in the AI Overview, they have links within that, then they have links on the side and they're not necessarily the same links.
So that's interesting, and I think that also what I'm starting to see is that on some of the search results they'll say, or sorry, on some of the search tools on their cache, they're saying, "Oh, this is an AI Overview." And then you go to the SERP and then it isn't. And so I think that it's still early days in terms of the data, which makes it tricky to track and it's still early days in terms of Google figuring out how they want to roll this out. So it's really a live space at the moment.
Mordy Oberstein:
That's totally true, and I want to get the URL thing, okay? That's the next thing I want to cover really quickly is that what do you do with all this data? Okay. Oh, AI Overviews are in 10% of all SERPs, I don't know, 15% of all SERPs, or 95% of all information queries, yada, yada, yada, yada, yada. But there's two fundamental questions. One, I have absolutely no clue about what I'm trying to get a clue about. One is how often are users engaging with the AI Overview, how often are they not? Or subset of that, how often are they engaging with it, bouncing out of it, and going to the organic results? So even if they are engaging with it, are they still going to the organic results anyway?
And if they are engaging with the AI Overviews, what's the stability of URLs? In other words, is it Google swapping out URLs all the time and there's no stability? In other words, what's the real impact of AI Overviews on traffic and then what's the real necessity of trying to optimize, whatever that means, for an AI Overview? Even if they are impacting traffic, but Google's swapping out the URLs on a daily basis, you have no market share. That's a waste of time.
Crystal Carter:
Right. Because if people aren't clicking on them, which we don't know what the click-through rate is on it, if people aren't clicking on them, should you optimize for them if you are getting straight to the SERP and things? So then it's a question of whether or not Google's taking the share of the click or whether or not there's a click at all, which we don't know if there's a click at all, or whether or not somebody else who's in the AI Overview is taking it. And I think that if we think about the featured snippets element, featured snippets are really, it's very much a similar evolution from when we didn't have featured snippets to when we do have featured snippets. Because in some cases, the featured snippet gives you enough information so that you don't need to make a click. That is true. It is true.
Mordy Oberstein:
And that's when the format of featured snippet matters most. The paragraph featured snippet you might have the answer right there. List is usually not the full list, so you usually get a higher click-through rate.
Crystal Carter:
Right. So these things will change, but the visibility is very important. So if people are constantly seeing your content in the featured snippet, then they'll go, "Okay, this is a great source."
Mordy Oberstein:
Totally, 100%.
Crystal Carter:
And I can just go straight there and get the information rather than having to go to Google to find the thing, they'll probably have it.
Mordy Oberstein:
Yeah. I think you always want to be in the AI... If there's one there, you always want to be there, which brings the question: what about that URL stability? For example, I just did a study... I didn't just do a study. I did a study 10 years ago, maybe less, maybe seven or eight. I don't remember what I ate for breakfast, so I don't remember what year it was when I did the study. And it was featured snippet URL market share. If you're in the featured snippet, how long do you tend to last in the featured snippet? And basically, the pattern that I saw years ago, I don't know if this holds true, is that there's generally speaking, two URLs: a dominant URL, and a secondary URL. The dominant URL stands in the featured snippet around 75% of the time, and then the non-dominant URL is in there 25% of the time. So that's pretty good. If you could be one of those two URLs either way, that's at least decent market share, 25%. With the URLs inside of AI Overviews, what's that pattern look like?
Crystal Carter:
Yeah, yeah, this is it. So it'd be interesting to see how that develops and it's so fresh, it's so new, particularly in different markets as well that we'll have to carry out the data and I think it's going to have to be a little bit manual for a while.
Mordy Oberstein:
It's a little wonky because a little peek behind the curtains, I've been working with Semrush to try to do something like this. And we did one version, had to redo it because Google changed the format of the URLs inside the AI Overviews, they don't want that data. And even then, it's still a little bit tricky. When I'm telling you it's tricky, I'm not telling you because I think it's tricky. I'm working with companies and I'm telling you it's tricky. I don't know if I could spill the tea or not. I don't know when this episode is coming out. I don't know when the study's going to come out. I will tell you-
Crystal Carter:
Spill the tea, Mordy. It's your podcast, it's your study. Me too. You do what you want.
Mordy Oberstein:
I already drank all the tea. I'm literally holding my tea mug upside down. Oh, I just actually... There was one drop left and I just dropped it on my keyboard.
Crystal Carter:
You can't tell us about all this juicy stuff and then hold back. That's just-
Mordy Oberstein:
I'll say it's less stable than you think it is. And I'll throw one stat at you. If you're out of there over a 31-day period, and this is only for AI Overviews where the AI Overview showed for at least 20 out of 30, 31 days. So you're talking about AI Overviews that are already consistent. If they're not consistent, then what's the point of URL consistency? The whole thing's not consistent, right? So I guess I'm saying all this data is so layered, so let's now only look at AI Overviews that are consistent. They show up 20 out of 30, 31 days, whatever we looked at. If you're out, you're out. Only 45% or so of the URLs that got kicked out ever came back in over that data period.
Crystal Carter:
Over the course of how long?
Mordy Oberstein:
31 days,
Crystal Carter:
Right. Right.
Mordy Oberstein:
I don't know, maybe they come back after 60 days. I don't know. Secondary data point, I don't remember this exactly off the top of my head, but it's not like Google... Because maybe you'll say, "Okay." When Google's swapping the URL out... That's how my brain works. It's why I'm good at data studies. Maybe it's... And it takes out the Wix SEO Hub and it puts in another URL from the Wix SEO. All right, that's fine, right? I still have the market share. I think 100% of these AI Overviews that were in this category saw domain changes, meaning Google's not just saying, "Oh, one URL for one website." It's swapping, I don't know to the extent, but significantly so, swapping out domains. And when you add it all together that very few AI Overviews have the level of consistency that you're looking for. Very few URLs are returning, and a lot of domains are switching. It doesn't paint a very stable picture as of right now, as of the time of this recording. It could all change tomorrow. Who knows?
Crystal Carter:
See, this is interesting because I think this plays back into what Mark's talking about is that the SERP right now is a little bit unstable. We do a weekly or a daily news thing with Barry, and he's constantly like SERP volatility, SERP volatility, SERP volatility. If it is the case that the AI Overview optimization is essentially SEO, then it would stand to reason that the URLs would switch around in a volatile SERP. That would make sense. I think that this also points to, if you're thinking about AI Overviews, thinking about some of the information around... Or some of the fact that there are still some big, big players in terms of the AI Overviews of who has the most visibility.
So Kevin Indig shared who's got the biggest share of AI Overviews, and he was saying that the biggest one is YouTube. Again, that's really similar to featured snippets. Then followed by Wikipedia, followed by YouTube mobile, followed by LinkedIn, and then followed by a few others, including usual suspects like Healthline, and WebMD, and Mayo Clinic, and things like that. We also see that those same folks see a big share on the SERP, and those same folks will have a fairly stable level of visibility on the SERP. They have fairly stable rankings in terms of where Google shows their domain and how often that Google shows their domain. But yeah, I think it's a really fascinating time in research and looking at this particular part of the SERP.
Mordy Oberstein:
Yeah, exactly. But you were saying what Mark is saying, "Do good SEO."
Crystal Carter:
Yo.
Mordy Oberstein:
Just do good SEO. That will get you there.
Crystal Carter:
Right? Everybody wants some fancy thing, but it's like people are like, "How do I lose weight?" It's like, "Eat better and do some exercise." It's nothing complicated.
Mordy Oberstein:
Just the execution part, that's the hard part.
Crystal Carter:
That's true. That's true.
Mordy Oberstein:
Yeah. Some of it gets a little tricky.
Crystal Carter:
Because I love donuts, so if I could just have a delete donut. Have donut, delete donut, if I could just have that, that would be great.
Mordy Oberstein:
Oh, I love donuts so much. Krispy Kreme or Dunkin' Donuts.
Crystal Carter:
Oh, I love Dunkin'.
Mordy Oberstein:
I love Dunkin' flavors. They have better flavors, but the donut itself, I think Krispy.
Crystal Carter:
Krispy Kreme is great, but I find Krispy Kreme to be very expensive, and also, they don't oblige my dietary requirements, whereas Dunkin' Donuts comes in better for me. So it's not a flavor thing. It's not explicitly flavor, it's for other considerations.
Mordy Oberstein:
Either way, donuts for the win. You know what else for the win? Tracking your AI Overviews. Global AI Overview data is great, super interesting, super informative, super everything, but how do you track your own AI Overviews? Well, find out as we look at the data providers as we go tool time on the SERP's Up Podcast. All right, so there's a whole heap of tools that do different things. I don't know, like Semrush and SE Ranking. Let's start with them maybe. I found them. They both have it in position tracking and then they were like organic research so you can see how stable your AI Overviews are and what keywords you have. And you can also analyze the competitor and see how many overviews they have, which I think is cool for looking at brand power for example. If I look at, I don't know, Nike versus New Balance, which brand is more powerful? Or the one that has a ton of AI Overviews kind of thing.
Crystal Carter:
Right. Right. I think it's interesting to see that and I think that it's great to see some of these different tools showing this because people need to know. So I think that what is their general summary for the overall visibility on AI Overviews?
Mordy Oberstein:
Yes, if you run it through the organic research tool and then you can filter out by SERP feature, you can see all of the AI Overviews that that website or that domain is grabbing.
Crystal Carter:
And on their Semrush sensor, they also have included it under their SERP feature there so you can get an idea of the general trend. So according to-
Mordy Oberstein:
Yeah, but that data set is super normalized. I think that data set is... Because you're looking... I think now they're showing 2 or 3%, something like that. I know that data set because I work with it when I'm doing their algorithm update stuff. It's not reflective of what you do every day. The Moz data set would be. The MozCast is very much high search volume keywords, assuming that they're informational kind of thing. That would align with AI Overviews. I think that the Semrush data is just more normalized across the board, which is what I mean. This data, looking at the numbers, it's so hard. So hard.
Crystal Carter:
So then help me understand this. So according to their Semrush sensor, they're currently saying that for the US, on desktop, it's showing at 1.23% of search results and on 1.63% of search results. So what are you saying when you're giving that information?
Mordy Oberstein:
So what I'm saying is, okay, that's interesting because you compare this to say, I don't know like what Mark was talking about, or SE Ranking, and BrightEdge, and their data. The Semrush data set is lower, but I think it's because they're using their volatility data set to track it. When they track the volatility of SERP features, they have your featured snippets, and local packs, and images, they're not creating a separate data set that targets that feature or aligns with that feature. It's a universal data set to track all volatility. So it's not built... The way I search, I search mostly for informational keywords. That data set is not built to function the way I or most people search. It's meant to try to track the web. But AI Overview, so you're trying to track those. If you wanted to really see that you would have to include more things like how people actually search, in which case you would probably see a higher number. So I would look at the trend, not at the actual number in that one, and the trend's up.
Crystal Carter:
Yeah, and the other thing I think is interesting on the trend is that the Semrush sensor, for instance, filters by categories. And if you look at arts and entertainment, for instance, it's showing AI Overviews for 0.14%. But if you look at health, and this goes, again, back to some of the data that Kevin was showing, that some of the top AI Overviews include Healthline, WebMD, Mayo Clinic. That's three out of 10 for the top 10. And this is saying that health is showing 3.63% of the time. So again, maybe the numbers, maybe the number percentage might not be hard fast-
Mordy Oberstein:
But the trends, yeah.
Crystal Carter:
Exactly, but the trend that, on average, it's showing 1.63%. On average, it's showing a certain amount, but it's showing three times as much in the health sectors is interesting. And that also follows with some of the data that Lily's been sharing around that particular sector.
Mordy Oberstein:
Yeah. And there's so many tools like the seoClarity tool, for example. They show you the actual AI Overview content itself so you can actually see the snapshot of the AI Overview and what's actually in there. So you can see, obviously, where you're being impacted or where your competitors are being impacted, but you can get an actual sense what does that actually mean by looking at the actual content and the snapshot of the AI Overview. Which is great if you are, for example... If you're not in the U.S. and your clients are in the U.S. and you want to see what that actually looks like without using a VPN. You can use a VPN also, but you can use seoClarity and it's right there.
Crystal Carter:
Right. And another one that's got some is ZipTie.
Mordy Oberstein:
Yeah, ZipTie is interesting. I checked them out.
Crystal Carter:
So ZipTie.dev is detecting AI Overviews. They are claiming that they're detecting 17 times more AI Overviews than certain other tools. So they were checking us. According to ziptie.com, wix.com is showing a 27% of AI Overviews, which is interesting. Yeah, so they've got various different tools. I think they've got a freemium setup as well and they've been putting out some reports around that. So shout out to Tomasz Rudzki for some of his data there because that's really interesting.
Mordy Oberstein:
Yeah, and it's cool because they'll tell you with AI Overview, remember there was one where there was automatically triggered, and then one format where it wasn't automatically triggered. They'll tell you which one it was, or they'll tell you if you have the AI Overview and the featured snippet, or just the AI Overview, or just the featured snippet. They'll tell you what AI Overview features are inside of it. Is there a topic breakdown? Are there pros and cons lists? Are there bullet points? Is there a shopping list? They'll actually break it down for you. So it's a really interesting tool. It seems like a really powerful tool for tracking AI Overviews for sure.
Crystal Carter:
Yeah, absolutely, and they've got their... On their blog, they're sharing a lot of data. So for instance, they were saying that, in June, it rose from 8% in June to 11% in August, which is interesting over things. And again, this goes back to your trend point, so just like it's on the up. How much? It's difficult to quantify exactly, but yeah, they are also breaking it down on their blog as to how many... What types of terms people are seeing it on and things like that, so yeah, interesting times in SEO.
Mordy Oberstein:
Oh, yeah. We'll link to all those tools in the show notes. Check them all out, see which one is best for you. We are agnostic here on the SERP's Up Podcast, so pick what is best for you. You know what is a great place to learn about AI Overviews and the latest news happening around AI Overviews? I find that to be such a tongue twister, by the way, AI Overviews.
Crystal Carter:
It just makes me think of Old McDonald. I'm not going to lie.
Mordy Oberstein:
The what?
Crystal Carter:
Old McDonald, like A-I-A-I-O, like E-I-E-I-O, come on.
Mordy Oberstein:
That’s a great connection. That's great.
Crystal Carter:
Yeah, like << On the SERP there was a... And Google had a SERP, A-I-A-I-O. And you get a link, and a link over here, everyone link, link, link, link, link >>
Mordy Oberstein:
By the way, no one actually... We should have a separate conversation about the actual impact CTR-wise. Side note, I see a bunch of cases where a lot of rankings in AI Overviews and not real traffic change.
Crystal Carter:
As a user, I very often skip over it. As a user, I very often skip over it.
Mordy Oberstein:
I wonder if they check it out and come back, maybe then I click... I'm interested if they click on it and just don't even see the links and just go back and click on the links in the SERP, that would be interesting.
Crystal Carter:
Yeah, I think that's interesting. But also, I find that sometimes the AI Overview doesn't make any sense for what I'm searching. I'll search for one thing and then it's-
Mordy Oberstein:
Eat glue!
Crystal Carter:
Well, yeah, I'm searching for one thing and then it shows up with something else, and I'm like, it's not what I'm talking about.
Mordy Oberstein:
Yeah. It's very broad.
Crystal Carter:
Yeah. And I think the other thing with the AI Overviews, they're trying to make a cohesive statement in the AI Overview, and they're trying to make a cohesive statement about whatever it is you're looking at. And sometimes if you were to look up Bluetooth speaker connectivity, right? Now, I might be looking for how I can connect all of the Bluetooth speakers in my house, so I don't have to find each individual one. I can just play Shakira on all of them. I might be thinking about how would I do that, but I've only typed in Bluetooth connectivity. And they might be like, "Oh, Bluetooth works like this, and this, and this, and this, and that, and that, and that."
Mordy Oberstein:
Yeah. Not what you wanted.
Crystal Carter:
And I'm like, "Y'all, that's not what I'm interested in." And so sometimes what's on the SERP is more aligned with what I'm interested in. And sometimes it's not, but I think that the nature of the beast means that sometimes it doesn't align, and it'll be interesting to see how they get better.
Mordy Oberstein:
Yeah, we digressed, by the way. I was trying to get into the Snappy News by saying, "You should try..." No, that's my bad. I did that. If you're trying to keep up on the news on AIOs, you always want to check out what Barry Schwartz is doing over at SE Roundtable, but until then, check out this week's Snappy News. Snappy News, Snappy News, Snappy News. Two from Barry Schwartz, both from seroundtable.com. First Google AI Overviews not linking to sites hit by helpful content update. This is from, per report, this got started by Lily Ray who reported on X, "Hey, when you type, who is XYZ's site?" And insert site affected by the HCU, the Google AI Overview result seems to not include links to that site, yada, yada, yada, yada. Glenn Gabe tested the same thing, and he says, "Yeah, it looks that way." He said even for the sites that, quote, "bounce back a bit with the August core update, and they don't show up in the AI Overviews for their own brand site."
It's fascinating. It feels to me like that's like a glitch in the matrix, especially if it's a website that yes, got hit by one of the previous core updates, or the initial... Not the initial, but the last of the HCU. That should be the name of a movie, the Last of the HCUs, back in September 2023. If it's your own website, if people are particularly specifically Googling, "What is name of website.com?" I feel like you should show up there regardless of what Google thinks of you algorithmically in terms of ranking for informational navigation... I mean informational, commercial transactional searches related to your queries. It's a navigational search. Would you not rank navigationally? If someone typed in, "What is blah, blah blah website.com," you would show up number one regardless of any algorithm. I mean, unless you're like manual action kind of stuff, but so why not in the AI Overview?
Then there's a little bit of speculation on X about Barry's own website, seroundtable.com, which I check out each and every day, and which you should too. And if you want, you can also get a dose of seroundtable.com on the Wix Studio, SEO Learning Hub where we have It's New, our daily dose of SEO News with Barry Schwartz and Greg Finn. Plugging is over. Barry's site didn't seem to show up in the AI Overviews, which is just weird. I did a few of these searches related to seroundtable.com myself. Sometimes Barry does show up there, sometimes he doesn't. Sometimes it's just his X handle. Sometimes it is the actual website. I found one where you search for what is seroundtable.com. So then Barry does get the link in the AIO overview, but Apple gets the featured snippet for some kind of search engine round table thing that they have in the app store that who the heck knows. That's not this. That's not the real Search Engine Roundtable. Barry is.
But for example, if you search for what is search engine around table website, seroundtable.com does not show up in there as of the recording of this recording. The recording of this recording, that's not redundant. Only Barry's X profile shows up, so that's interesting. So again, it just seems wonky, for lack of a better word, or glitch in the matricky, for lack of a better word. Anyway, on to the second article. Also, again, reminding you from Barry Schwartz on seroundtable.com, which you may or may not find in an AIO overview near you. DOJ may break up Google as remedy to monopoly ruling. I see this on It's New. First, Ben Affleck and Jennifer Lopez break up, and now Google. How many breakups can I handle? I can't. I just can't.
This is reporting... A lot of it was reported on CNBC, which Barry links to in the article. I don't want to get into the whole 32-page document issued by the DOJ. You could read that on your own time because that's not snappy. Anything 32 pages is not snappy. But Barry did pull out some highlights and Barry writes, what is proposed is not just a break-up, which could split off Chrome, Android, and many other things. "But also," says Barry, "Google will make available its competitors' data within its search index and models. Google will make available its AI-assisted search features." Again, whoa. "Google to make available its ad ranking data, policy regulations about retaining its data, limitations on its partnership with Apple and others," that does make sense, "and more."
So some of that is just like, "Wow, just take everything." It's a full-on garage sale. We're taking everything. So that's really interesting. I'll link to Barry's article in the show notes, which links to all the other links that you need. If you want to deep dive into 32 pages of this stuff. I do not. The minute I hear law and legal, I'm out. So I'm out. That's this week's Snappy News. I'd like to see AI Overview Barry. You ask Barry and it just spits out a summary, but the summary is two words.
Crystal Carter:
Yeah, right?
Mordy Oberstein:
Reads.
Crystal Carter:
Not new.
Mordy Oberstein:
How do I connect my Bluetooth speaker, AI Overview Barry? With a plug.
Crystal Carter:
Not new.
Mordy Oberstein:
Not new. With electricity. Complete brevity. Brevity AI Overviews by Barry Schwartz.
Crystal Carter:
And efficiency.
Mordy Oberstein:
Efficiency. I had a pocket in my shirt. I put my phone in it. For reference, go look at an old episode of It's New for that. Which brings us to our follow of the week, and our follow of the week is someone we already mentioned this week, so you should be following him anyway by this point. You should have listened to the episode and clicked follow on this person's profile because we mentioned their name. That's a little much. It's Rich Sanger. Rich was one of the folks who, first off, we met him a few times at our New York City digital marketing meetup at the SEOFOMO meetup we did in New York City at the Wix Playground.
So Rich is a great guy. I think he's a Red Sox fan, so I won't hold that against h