In this episode, we unpack how AI search is reshaping discovery, why traditional SEO isn’t enough anymore, and what it actually takes to show up inside AI-generated results. If you care about visibility in the next era of search, this conversation will change how you think about marketing.
https://page2pod.com - AI is transforming how users discover information, shifting from traditional search engines to AI-powered answers. In this episode of the Page 2 Podcast, Jon Clark and Joe DeVita sit down with Baruch Toledano (SimilarWeb) to break down the latest insights from the Generative AI Landscape Study and the 2026 AI Brand Visibility Report.
They explore how AI search differs from Google, why prompts are getting longer and more complex, and what it takes for brands to appear inside AI-generated answers. From the rise of niche publishers to the growing importance of your entire digital footprint, this episode is packed with actionable insights for marketers navigating the next era of search.
🔍 In This Episode
• How AI is compressing the internet into a single answer and changing discovery
• Why AI prompts are dramatically longer than traditional search queries
• The surprising rise of niche publishers and long-tail content in AI results
• How SEO and AEO (Answer Engine Optimization) now work together
• Why your entire digital footprint matters more than ever
• The role of brand recall and branded search in AI-driven journeys
• How industries like fashion are winning in AI search visibility
• Why Reddit and consensus across sources are key ranking signals
• The growing importance of content depth, FAQs, and follow-up questions
• How AI and Google complement each other in the customer journey
This episode reveals how marketers can adapt their strategies to stay visible in an AI-first search world.
If you found this episode valuable, subscribe to the Page 2 Podcast for more insights on SEO, AI, and digital marketing.
💬 Comment below: How are you adapting your SEO strategy for AI search and AEO?
🛠️ Resources & Platforms Mentioned
• Follow Baruch Toledano on LinkedIn - https://www.linkedin.com/in/baruchtoledano/
• Baruch Toledano's Tech SEO Connect presentation: https://www.youtube.com/live/4whGj6qpZWY?si=EvTsXPFJmMD6QHPP&t=497
• SimilarWeb - https://www.similarweb.com/
• 2025 Generative AI Landscape: The State Of Gen AI - https://www.similarweb.com/corp/2025-generative-ai-landscape/
• 2026 Generative AI Brand Visibility Index - https://www.similarweb.com/corp/2026-genai-brand-visibility-index/
Jon Clark (00:00)
Search engines have spent decades organizing the internet into pages of links, but AI is doing something very different. It's compressing the entire internet into a single answer. And that creates a new question for brands. If AI is summarizing the web for users, how does your brand make it into that summary at all? Welcome to episode 111 of the Page 2 Podcast, where we uncover the strategies, systems, and tactical decisions that move brands beyond page two and into real visibility across search and answer engines.
Today we're joined by Baruch Toledano, VP of Digital Marketing Solutions at SimilarWeb. Baruch has spent more than two decades building the technology at the intersection of product, e-commerce, and digital marketing, from leading product at Magento to serving as Chief Product and Growth Officer at Conductor before joining SimilarWeb. In this episode, we're digging into two major research efforts from SimilarWeb, the Generative AI Landscape Study and the hot off the presses 2026 Generative AI Brand Visibility Report.
The data reveals just how different AI discovery really is. Prompts are dramatically longer than traditional search queries. Answers are built from dozens of sources. And the brands that show up aren't always the ones that historically dominated search results. Baruch walks us through what SimilarWeb is actually seeing in the data, from the surprising industries gaining visibility inside AI answers, to why publishers and niche sites are suddenly appearing everywhere, and how traditional search and AI search are now feeding into each other in ways
marketers haven't had to think about before. At one point, Baruch says something that really stuck with me. Your entire digital footprint matters, not just one channel. And that idea really sits at the center of this whole conversation. Because if AI is reshaping how people discover information online, then the ways brands build visibility and authority may be changing just as quickly. If you learned something new today, take a second to subscribe to the Page 2 Podcast. Leave us a rating or review and let us know what resonated. We'd love to hear your thoughts.
All right, let's get into it.
Jon Clark (02:01)
Welcome to episode 111 of the Page 2 Podcast. I'm your host, Jon Clark, and I'm joined as always by my partner at Moving Traffic Media, Joe DeVita.
Joe (02:10)
Hi.
Jon Clark (02:11)
we're joined by the VP and GM of digital marketing solutions at Similarweb, Baruch Toledano. Welcome to the show.
Baruch Toledano (02:18)
Thank you and what an honor episode 111 so many ones I'm just excited.
Joe (02:24)
We've got
Jon Clark (02:23)
You're right. Good luck. That's a number
you gotta play later.
Joe (02:26)
Ooh,
we've got, we have so much to talk to you about today. We're there's no way we're to be able to pack it all in in this hour. Thank you for the hour ahead of time, but our audience is digital marketers across all different types of channels and tactics. I assume most of them know what SimilarWeb does. It's a leader, a leader in the space for competitive intelligence. Could you maybe give a
30 second overview of how you guys collect data and why your data is a little bit different than your competitive set.
Baruch Toledano (02:56)
Yeah, so SimilarWeb is basically an internet digital intelligence platform. As you guys are pointing out, we have a lot of our known about competitive intelligence and we use a variety of sources, including of course, real user data or what some people may refer to as their contributory network to be able to sort of understand movements on the web and on apps as well as on marketplaces
to provide organization with the data they need, particularly marketers and decision-makers, but also businesses to make better decisions, I would say, about their campaigns, their benchmarking, but also knowing a lot more about the consumer and their interests and trends that are happening. We actually have lots of different solutions
to, of course, on page two, SEO managers that covers the entire suite of things, both on SEO, AEO, but also we have a lot of things around ad intelligence and app intelligence as well. So it's a variety of data sets that allow us to really have a comprehensive point of view for any business or marketing teams, both in fine granularity as well as in an aggregate.
Jon Clark (04:06)
Good. we had Pat Reinhart, on our show last week, maybe an old colleague from Conductor. and we talked about some industry data on that show as well, but today really excited to dig into the 2026 generative AI Brand Visibility Index, which is sort of hot off the presses. Before we get to that, I was, I was curious about your background. So you started developing chips originally, right? And then you sort of moved into, I don't know,
Baruch Toledano (04:11)
Yes.
Jon Clark (04:29)
marketing enterprise type software from Magento to Conductor and now SimilarWeb. Is there something about marketing that really hooked you or was it really the opportunity to work with such large data sets that marketers often do?
Baruch Toledano (04:42)
I think, you know, in one's career, you're probably looking in some decisions you make at a given time and like what were the decision there. I want to say that like in the early 2000, when I became sort of like hardware engineer, I thought of the web a little bit as like this little child from engineering and not so much as a real engineering challenge.
What I found actually is that I built a lot of skills on system design, on breaking problems, especially in chip design, I would say it's very much like understanding sort of like ⁓ things that happen in parallel, both on the logical, but it also has a lot of physical constraint because
the chip has to work and doesn't get heat up and stuff like that. So it's like a variety of things that are happening. So it's very complex moment engineering. And so it was a natural thing for me to apply it also from software. The way I transitioned actually to marketing and large data sets and that actually happened to me mostly, I would say at Magento. And what happened there
it was like almost like a rush of building websites and e-commerce in its sort of like early days, but doing it for a very disruptive way. And what really came apparent to me is that all these websites are going to come on the web, but who's going to find them and how are they going to be found? And what's going to make these websites really shine? And like, as we know, a lot of websites are built like the physical world a little bit
but really we consume stuff online differently sometimes. And so I think we're definitely with the AEO GEO revolution, we're experiencing a whole new experience with the web, which is not confined just to the way we used to in the physical world, like retailers that have different departments and stuff like that that are organized on the website. So that got me really excited and marketing became this intersection for me
of really marrying between what the digital experiences are and what goes in people's mind and psychology, which is not something you can encode in a chip. So I was like, this is great. Like, this is the kind of challenge I love. So that's how I jumped into.
Jon Clark (06:42)
All right.
Very cool. I mean, it's so interesting with all of the things that are happening in AI and all the news around like the video chips and all that sort of stuff. You're sort of seeing these two things converge, you know, and being able to do more with better chips in terms of the data and analysis that you're working in now. I did want to, before we get into the study, I did want to talk a little bit about the Tech SEO Connect, which I was
Baruch Toledano (07:05)
Yes.
Jon Clark (07:12)
really bummed to miss last year. ⁓ It's on my goals of conferences to go to this year. And there was one slide that actually prompted me to reach out to you to come onto the that was around the nuances of SERP features and how Google repositions them depending on the query. Sometimes you do the same query and they're in different spots.
Each one has a unique impact to zero click performance, which when you sort of step out of it, you're like, yeah, of course that makes total sense. But I had never really thought about it in that level of detail. And so that was just so fascinating to me. In that data, were you guys seeing
any of those certain features like, I don't know, images or video, or people also asked that were out, that had an outsized impact on that zero click.
Baruch Toledano (07:58)
a tough question because I can't hone in on two particular subfeature per se, but I will say to you, someone who's been working in the SEO industry and building technologies for SEOs, I can say to you that the dynamic by which Google constantly changes the SERP
and hence, from there, what we see, what people clicks, which is the advantage of SimilarWeb, seeing, OK, this is what they search for, this is what they actually clicked. And so it gives you a whole different perspective about how much Google is changing constantly and also how much it's testing with the user of what information they want to consume in the form and the format that they believe is right for them.
Jon Clark (08:23)
Right.
Baruch Toledano (08:40)
And I think that, and this is what we see, and this is why we built such a complex algorithm to basically constantly evolve based on what we balance between what we see people do and versus what we're seeing on the SERP. I wanted to give the audience a little bit of an insight to what it takes to actually build this type of like very small metric clicks and zero clicks,
and I think sometimes we as SEOs walk into it like, it's a zero click. What's the big deal? And like, no, it's a really big deal. And it really has a lot of nuanced way of thinking about it. It's definitely we see the standard ones, as many of you guys have seen. Google is putting a mix of those when it's news. There are some predictability, of course, of the things you'll see when you see images.
Google feels a little bit like Amazon when you're product stuff and allow you and hone in to the particular For us, it's more of a matter of how much can we give it to the person to be able to make better decision about both estimating what they're going to get out of it, but also recognizing the implication of the format that drives more attention.
Jon Clark (09:33)
⁓
Got it. Yeah.
Joe (09:53)
You have you, you're just mentioning something all SEOs know is how often Google changes their search and your result pages. Like if you've worked in SEO for even more than one year, you know, like the change is constant. I listened to your presentation from Brighton UK and you make a nice comparison about how slowly Google is changing compared to these new LLMs, like the magnitude of change.
Baruch Toledano (10:04)
Right.
Yes.
Joe (10:20)
I don't think you gave it a number, but the scale of change is way faster than anything that we're used to. It's like you call Google the stable channel right now. It's crazy.
Baruch Toledano (10:32)
Yes,
Yeah, so at the time, I must say they were the stable channel and they still are to an extent. This is generally speaking, I think that like what the AI did to us is it introduced a new form of interacting with information where Google gave us a bit the trust and the relationship we built through the
organized results of how they're actually showing us. And we like that predictability. Also, as you can imagine, changing in Google have sometimes a big stake for Google itself and how users are expecting it. So I think if they make a change, the swing is very different than what you see. It's not to say that they don't make a lot of changes. It's just saying that they are making those changes. They're just very careful about walking these
very fine line behind changes, where some of these new LLMs are walking into spaces where they might make a change that might be actually triggering an incorrect answer or some form of hallucination. And Google, I believe, take it a lot more seriously when it walks to certain topics in certain areas before making those very drastic change. And you can understand why they want to disrupt more
how consumer or how users are actually seeking information on the web.
Jon Clark (11:48)
Such an interesting nuance that I hadn't thought about before either, which is Google sort of gives you access to data, right? You can put it in a query and it sort of funnels all the information up to you. The LLM really allows you to interact with that data. It's like, okay, here's all this data. Now, you know, take it one level deeper into something. That's really interesting, which is maybe a good segue. So the 2025 landscape study,
Baruch Toledano (11:59)
Right.
Jon Clark (12:14)
I wanted to ask a couple of questions there, which maybe ties into this. So I think one of the things that was super interesting is the prompt length versus a traditional search in Google. I think it was somewhere around like 60 words for a prompt versus maybe just under four for
Did you revisit any of that data as part of the 2026 brand index or are there any other sort of internal data points that are saying that is holding true or maybe even getting further beyond that range?
Baruch Toledano (12:44)
Yeah, so I would say to you that it's getting even more nuanced. And I actually don't see it slowing down. I actually see it increasing. I think we'll hit some plateau because some people don't want to put the whole life, full life story on every prompt. They make the assumption that memory and context is retained within the LLM. But I think that one of the things that we see
Jon Clark (12:51)
Mm-hmm.
Baruch Toledano (13:05)
is that people have so many nuanced things that they're asking. And I actually went a little bit to like dig even further to give you some examples. ⁓ And I thought maybe I should share some of those with you from just like two days of information. I was like, my God. So for sure they're designing their travel, they're assessing their eating habits.
Jon Clark (13:19)
Yeah, sure.
Yeah
Baruch Toledano (13:28)
Now, every week we have a different word or a different brand that we're interested. This week is about yogurts. So what I found about the interest of yogurts were like insanity of how much yogurt cooking and yogurt in diets and yogurt and yogurt to kids. And it's like, I don't know where this whole thing, but every week we are looking at different dimension. And so what I'm basically trying to say is then I see a lot of people sort of like creating songs,
putting songs. But also I wanted to like see how much Taylor Swift, for example, is coming up in all these like different prompts. Because like just I'm interested in like different things. I have a 13 year old daughter and like this is top of mind for her. I'm like, so I wanted to Yeah, so I wanted to read you one from one person saying I have a really good friend named Lauren whose birthday is on the seven. I don't know what to get her. I know she likes Taylor Swift. So I wanted to get her
Jon Clark (14:08)
You're a Swiftie by default.
Baruch Toledano (14:21)
like some merch and there's a snow globe on eBay, but it's only going to arrive about five to 11 and I'm only living on the 10th and I'm going to Pakistan. getting a lot more depth, I would say, about the and about the choices. And I don't think it's going to slow down. And by the way, there's other people,
she keeps going and it's like, this is how much money I have and here's what I'm willing to this thing. And when will it come? So I think when we are, and then there's another person asking about Taylor Swift in a whole different way. I'm applying for a marketing internship and the prompt asked me, who is your dream brand client and why can you help me craft a polished answer using Taylor Swift as an example, focusing on long-term rollouts, fan engagement and strategy.
So that's a marketer sort of point of view that is asking question and sort of like want to position themselves. So we see this very nuanced type of context and language context that people are bringing. And I don't think that this is going to slow down, but What I do think that it is actually an opportunity for marketers to not only think about just the sheer informational queries that we used to look for, the what, the why, the how
Jon Clark (15:10)
cut.
Baruch Toledano (15:38)
and things like that to compare all these things. But actually what people are asking to do in many cases it's a preemptive, like, things that we don't even always think about how to organize this type of content. So for example, like if you are, you know, looking for gift ideas, like how did they look at gift ideas through a lens of an artist, not necessarily through the lens of just comparison of one type of merch and a similar brand with the same type of merch that they may have.
So I find it to be like not going to slow down.
Jon Clark (16:10)
Yeah, that makes sense. I mean, one of the hardest things with a lot of this stuff, whether it's creating content or finding a gift is like just figuring out where to start. And I think that's why, you know, affiliate sites tended to do to do so well a few years ago, because it was like, okay, here's a list of 10 things. You know, they're all the best. So it gave you a starting point. LLM sort of give you a whole new dimension to being able to do that. I think the
Baruch Toledano (16:33)
Right.
Jon Clark (16:35)
maybe the interesting piece of that is even though you might be doing research there, it still seems like based on your data that a lot of the conversion actions aren't necessarily happening within the LLM. Those are still sort of heading over to Google in terms of conversion rate, like that. Is that something you anticipate as these LLMs start to include more and more e-commerce functionality or is that something where
Baruch Toledano (16:57)
Mm.
Jon Clark (17:01)
you know, someone's still gonna validate their choice on the Google side.
Baruch Toledano (17:04)
I think that we're going to see a mix of both for quite some time and both on not just exclusively for late stage final navigational type of queries versus informational type of queries. I don't think it's completely exclusive. Right now we see two main things. Yes, more information in AI search
and more transaction increase in some transactional navigation type of queries as a result of many times where they were primed before through AI search coming into But didn't sit around and Gemini within very few, I would say weeks, this thing became like a very, very powerful informational type of an LLM. And so I...
Jon Clark (17:44)
Yeah.
Baruch Toledano (17:48)
I don't think we're seeing an AI mode also on the way. It's been rolled out a lot more now. We're seeing ⁓ based on the previous study we did, like we saw AI mode, but not as much as sticky, but now we see AI mode even used even much more. And so I don't think that we are exclusively going to only see the two. I think they are complementing each other. And sometimes they may choose to start in one place and then go to
to Google, or we might start to see conversion happening on site on the AI search when you will be suggested adoption as a result of the context you provided to what you want. And I believe early adopters may actually lean into it as well and actually start buying. So we're watching it carefully. We're also watching the ads that are going to come carefully because we want to make sure that we give this information to users.
Jon Clark (18:33)
Mm-hmm.
Baruch Toledano (18:36)
For now, yes, there's definitely, we see a growth in navigational and a lot of AI search. And it's really important to look at both of them.
Jon Clark (18:43)
Definitely. I think one thing that sort of stood out from the landscape study and maybe aligns to the brand index study is around the age group cohorts. You know, so usually older demographics are a little behind in terms of, you know, adoption and things like that, but it seemed like AI users over the age of 45 were sort of over indexing. Do you think that plays a role into any of the Brand Index Report
Baruch Toledano (18:54)
Mm-hmm.
Jon Clark (19:10)
data? or is that purely looking at how brands are positioned in the results? is there any overlap there or are they pretty distinctly sets of data?
Baruch Toledano (19:20)
So we basically look at results we're seeing overall at scale. So we don't necessarily have an attribution to who is this persona behind it per se. Always we have, we know what they are, but we don't necessarily have sort of something to say about do they get
Jon Clark (19:26)
Mm-hmm.
Baruch Toledano (19:39)
separate different results for because they are from a certain age versus another another age. I believe that's my own opinion that over time the LLMs will become even more sophisticated in this scenario because I think that we in different ages we want to consume information differently and also I think that trust when it comes to those things is built differently for different people for different age group. I could see it we build on things sometimes that we took for granted.
You know, certain brands are standing out for us because that was more how we became, so we want to keep seeing it. And we have an inherent trust, at least in my age group, and where younger people might see actually an older brand being, almost like a turnoff sometimes and how did they evolve and, and they have to prove to them that they actually doing good in the context of, of what this brand is standing. So I think that AI and personalizations of answers definitely plays a role.
I don't see it in the data per se when we are actually giving a snapshot to give, okay, these are the brands that are showing more, but these are the brands that are showing less as a result of looking in like age group or demographic in that context. Or we're basically pointing out that this is becoming more ubiquitous technology rather than just early adopter or young generation that is actually we've seen in previous year.
Jon Clark (20:54)
Got it, got it. So in regards to the you guys chose sort of like a binary score evaluate the metrics. So like mentioned or mentioned, probably the easiest way to approach it. But did you guys look at sentiment at all evaluate a mention in terms of a list of 20 versus 10? Were any of those metrics as well, maybe a future study?
Baruch Toledano (21:06)
Mm-hmm.
So our approach, first of all, has been sort of like you have a set and you want to know whether they are in there or not. So in a way, we're approaching it first from the binary question of saying, are you in the answer? You're not in the answer. Let's start with the basics. And I think this is important because you can't even talk about sentiment if you're not even like mentioned yet. So I think from this sort of context, I would say that like,
this is why we believe that first let's start with the mention. And then from there, I think there's a whole slew of information about in what context were you showing. So what was the prompts all about? What are the topics? And then, you know, sentiment is an overlay that in cases people want to look, but they always have to look at it in the context. So you might be showing a lot, but the sentiment may be negative or, you know.
You might be showing a little and you're just like fighting to be shown. So I think those are like two different sort of things and combining the two, sometimes in an organization, I would say that like even different people might be handling the problem, right? If it's a PR problem of, of sentiment issue, then they might be taking it. If it's like, am I showing or what are the sources that influence even what shows up in the AI answer?
is maybe an SEO person or a content marketing person or social person might be going and actually improving where they're being even shown in the context that they care about. So I think this is why we took an approach of like, okay, we're going to show you first, are you there or not? And when it comes to sentiment, as you know, it's much more nuanced, you know, even saying negative or positive, it's really in the eyes of how you want to look at these things. And so...
this we have this capabilities but of course in the platform for the topics you care about of course and the brand you are interested in looking into.
Joe (23:06)
I thought this study had a little bit for everybody. There were some insights there. If you were working at an established brand, you might read some insights and say, OK, this is good for us. And then if you were maybe working for an up and coming startup, you might say, OK, this is good advice for us. There was a little bit for everybody in there. The one thing I really enjoyed was you used
Baruch Toledano (23:18)
Right.
Joe (23:31)
a comparison with your visibility ranking and your branded search volume. I thought that was pretty unique, I never heard. It's a, it's a nice way to compare two things in the kind of digital space. Why did you guys choose that above some other known metric?
Baruch Toledano (23:46)
Okay, so let me maybe start with the fact that branded search is the best indicator, together with direct traffic to your website, the best indicator to brand recognition that is done for behavioral measurement attitudinal, yeah, me sending a survey and asking people what they think about my brand versus what they actually do.
So from our perspective, it was important to sort of like seeing how much recall this thing has in one's mind and how prevalent is this thing also when it comes to AI search. So we took the branded sort of like search in a way that is like, that's in the enough people's mind and let's try to compare the two and how they're actually showing.
Jon Clark (24:20)
Mm.
Baruch Toledano (24:30)
And it goes back to my previous point in the, in the, other question, which is like, you know, branded search, we see an increase in navigational, which means we see a lot more branded type of search. So like, did they come from AI search first and then they came here? So for us, it was a really important thing to point out that marketing, especially in, in Google, they so prevalent and this fast growing ChatGPT and other LLMs that are starting to show up.
We have to be able to look at them side by side and have them together because we think that users are, and we see with behavior of users, they are still going to Google to do parts of their journey. They're still making the final comparison on the price sometimes in Google. And so maybe they've been educated, but they are now there. I think it raises another big piece into this entire thing, which
Jon Clark (25:15)
Hmm.
Baruch Toledano (25:22)
gonna see cost of CPC going up on branded search because, fighting on the last mile. So, you know, that's become an important piece into this into this sort of dynamic. Yeah, this is something that I find that is essential to look at the two parts. It also points out that there's an opportunity for many other brands
that don't have a name to how to think about it. And this is something else we try to bring in the study that is not just the brand. And the truth is, is a lot of the sessions we see have nothing to do with brand to begin with, has to do with the bigger questions and information before even the brand is showing up.
Jon Clark (26:02)
I have so many questions around the conversion piece, because as a marketer, it's the hardest part. Are you guys maybe in the SimilarWeb platform, not necessarily in the study, but are you doing anything to connect those things? So for example, someone comes in on an LLM, they do a search, you can see all that, which is unique compared to other, like Google Analytics, they don't know that that's happening in the LLM.
Then someone comes in and maybe they're, you know, they do a brand search, come in on a paid ad, which now looks like a different source, right? Are you doing things to sort of connect those attribution, you know, mismatches?
Baruch Toledano (26:29)
Yeah.
Yes, but I will say that they're tough, right? So let me maybe try to give you a window into it. So we, we have a metric called the brand recall. And the brand recall is the ability to say within a window that is bigger than, than Google analytics window of sessions, you know, like what is the level of recall over time
Jon Clark (26:41)
Sure, yeah.
Baruch Toledano (26:58)
that we see from users? just so we understand some ideas about brand strength and brand momentum over time. We don't today necessarily go one level deeper, which is potentially where we might take it, which is like, you saw an answer of this brand showing up. How much longer from that moment you went back and put maybe a branded keyword inside Google
in this context. It's definitely a lot of interest, but it's a really hard problem to solve. It's not a trivial one. I want it to, but it's definitely not a simple one. But we do have a lot of these brand recall. This is why we bring, I genuinely believe we are in a place where marketing is being challenged on many fronts where the first time all marketing roles have to work much, much closer to each other
of how they move the needle. We're going back to like the basics of marketing, and I think this is actually... it's a good thing. You know, your entire digital footprint matter, not just like the one channel, one channel that is speaking and AI is just like amplifying it.
Joe (28:01)
It feels like the most exciting time of my entire, I think we've all been working around the same number of decades and it feels like the last couple of years have been the most exciting for my entire career. I would get, I know you two both have to agree with me.
Baruch Toledano (28:07)
No.
I think so.
Jon Clark (28:16)
100%. Yeah. ⁓
Baruch Toledano (28:18)
of us. And yeah, it's, it's,
a, it's a lot of fun. Actually. It's becoming harder, but it's also becoming, it forces organization to really stand and say, what are we want to be known for? What do we want to be good at? and not you, you can't meet everyone everywhere. So it's really go back to the core of what this company wants to be known for about the product, their audiences and what problems they want to solve.
Jon Clark (28:40)
You going up. That is something you're already seeing in your data or is that sort of a prediction out of this brand study? I mean, it makes total sense, like, you're trying to generate...
Baruch Toledano (28:51)
We see a little
bit, we see a little bit of it at the preface of it. With that being said, I must say that the pie of the overall search in its behavior of ground two. So this is the other piece. We're kind of like, they're not mutually exclusive. And so, you know, whether Google will apply more, you know, keywords to bid on that may be even in different that they traditionally didn't,
is, you know, might be another way where they actually making, you know, a difference. For the time being, we've seen some increase, but not a major one. What we see is a lot of fight of brands protecting their brand because it's the one channel that is like, sometimes people take advantage and, you know, organizations wants to really protect those brands
well, and they want to monitor what's happening with it.
Jon Clark (29:39)
Meaning are you're seeing increased budget on branded terms or increased like compression share?
Baruch Toledano (29:43)
All right.
Increase of organizations bidding more on branded terms. So they're willing to spend more on the branded terms because they want to make sure that they're protecting their asset. This is their asset. It's really important for them. So.
Jon Clark (29:58)
Got
Joe (29:58)
You
know, we haven't really seen an uptick in conquesting. We're working with about 40 clients right now. We provide performance reports for all of them. We haven't really seen an uptick in conquesting from competitor brands, but I think that'd be something to start looking at a lot closer now.
Jon Clark (29:58)
it, got it.
Baruch Toledano (30:16)
100%. Yeah.
Jon Clark (30:17)
I wanted to talk a little bit about the category. So the index report covers, I think it was six categories. And you mentioned earlier about like LLM models getting better at personalization and sort of aligning their responses to you as a user. Did you happen to notice any changes based on the industry? So for example,
did ChatGPT have a propensity to act as a financial advisor for like finance related topics I don't know, be more informational for news related or publisher related content? Are you seeing any nuances like that within the data or is that not something that you really got a chance to dig into very deeply?
Baruch Toledano (30:57)
We haven't organized per se the particular sort of like sector, I would say. But we do know that there's a lot more careful approach to areas like finance like things we've seen also with Google, with Health, and type of those advisers that are like really...
even the LLM are taking a careful approach into what answers they're giving. But ⁓ in general, what's fascinating is that it really depends on how much time you spend with AI. And we've seen a lot more convenience of asking things that traditionally maybe were seeked elsewhere are now there.
So news and media is a really interesting one where we've seen a where people are just like interested in news or point of view about news in this context and because they can pack the entire topic in an easy digestible way. So it is an interesting sort of like news in general is its own sort of dynamic on its own because the content is their product.
They also are working in the background to work on deals with the LLMs about their stuff. But yeah, it's definitely a place where we've seen a growth accessing AI search. Yeah.
Jon Clark (32:08)
I think Joe and I are probably gonna ask you the same question here.
Joe (32:11)
insight in the news and publishing section of the study
for evergreen content.
A news organization doesn't make its money on evergreen content, they make their money on the breaking news. I, I think we only work with a couple of publishing clients, but I think that's where the organizations are focused. But the opportunity might be around evergreen content. Is that only for smaller publishers? You think that that is advice you can give generally to, ⁓ to the news industry.
Baruch Toledano (32:39)
I think in general, what we see is that many, many publishers are gaining a great podium because they're addressing a lot of questions that people have and they have such a diversity of content or specialization in certain areas that the brands don't always give the same level of attention. And so in fact, I think they see a renaissance because they're becoming like a lot more
prevalent in these places where it's no more nuanced type of context. And they are a great source that LLM are tapping into as a way to sort of like bring that. With that being said, some of it may be addressed for an evergreen content in the past. And some of it may be like, what's happening right now in Trump tariffs, or I don't know, the next conflict that people have or the stock market as it stands. So I think there's a room for both.
But it has to be good and it has to be genuine and there has to be enough external sources that validates the connection of you to that topic. It cannot be live in isolation because AI value more. I would say that enough consensus exists about you as a brand about this topic. And so it's not just that your domain authority, but there needs to be a bigger consensus around it,
go and give you the answer.
Jon Clark (33:57)
Makes The more times that they see you referenced or see a data point, they're going to trust it more. You know, the model's going to adopt it. The news category has always been a fascinating one for, I think us to talk about generally as an industry. We had John last season. One of, one of my favorite discussions. But I think anytime you talk about publishers, the
Baruch Toledano (34:13)
I love John.
Jon Clark (34:19)
nuance that you mentioned, which is they're all trying to develop these individual relationships with the LLMs. They want to be paid for their content, which is totally reasonable. But sometimes that impacts their visibility currently. Did you guys account for that at all? Or was there an easy way to do it? I don't know, looking at robots.txt files or stuff like that. Or was it purely like, we're just going to report on what's visible and not necessarily who's missing. Does that make sense?
Baruch Toledano (34:46)
Yes, it's the latter, right? Like, first we tell you what's going on. Like our perspective know what is going on. Like this is the most important thing. Now, the reason we looked further to see what the other publications are actually showing up, and we see it in all categories, by the way, it's not unique just to news, is that There was something that stood out to us really, and when you dig deeper in who gets a lot more AI answers,
Jon Clark (34:52)
Yeah.
Baruch Toledano (35:09)
and not necessarily as a brand, known brand name, you see a lot of media outlets. You see a lot more opportunity in the long tail that exists. And I think that organizations, even brands that adopt more of this sort of approach of looking at themselves, not only as a, as this is my bread and butter, my products, my services, but actually becoming to an extent publishers for the entire journey, actually benefiting from it
and seeing like a great... because here's the thing, when you put ⁓ a query on AI, AI makes a lot of effort to give you a compressed, but a full answer that also addresses and almost like compresses your entire buying journey. So if you've got the products to sell, it's well organized and you also got a lot of ways to, contextualize how these
products are being used, not starting with the product, but starting with what you are doing in your life and where does it fit and what's currently trending for you, we see a success for them and I believe that we'll see more CPGs are investing in there as well as fashion brands because this is one area that we've seen some success.
Jon Clark (36:15)
You had an expert, Ethan Smith, you know, he's been on some really good webinars with you guys too, but he, I think, one thing that was super interesting to me was he mentioned, you know, to be successful in the future, you're going to have to be thinking about publishing content you wouldn't typically publish for SEO. And I think that aligns perfectly to what you just said. You really have to think like a publisher and sort of cover everything, even though it may not be, you know, SEO driven
Baruch Toledano (36:18)
Yeah.
Correct.
Jon Clark (36:40)
at the outset. I was also really curious about the individual surface, or I guess the platforms themselves. You know, they surface different brands for maybe the same prompt or the same query. Is that, are there any nuances there that you can talk Just in terms of, I don't know, maybe a certain LLM treats different brands, I don't know, more
different than say, ChatGPT which maybe surfaces known brands more or something like that.
Baruch Toledano (37:07)
I think it's a big question also for all SEO. Is good SEO good for AEO, you know, is good AEO good for SEO? It's like a very sort of like common question. So I will say, yes, if you're a known brand, you probably are going to be in the top 10 brands that are showing up in that study. There are some nuances, as I said, within each industry a little bit, like the news that we just discussed. But in general, I think the question is more,
Jon Clark (37:12)
All
Baruch Toledano (37:31)
how much room and diversity exist? And yes, the top brands are important and we should learn from what they're doing. But I also think that we need to look at what we try to show in the study is what we call the momentum and the ones that decline as well. And so what is interesting is that it's not a finite that's your destiny and this is who's going to show.
⁓ And actually what we see is that some of them are actually coming out of nowhere and doubling down and keep investing and they get rewarded, but some of them are not or they used to, and now the LLM algorithm has changed to an extent. So it no longer gives them the same podium that they used to see. The study that you see is, is combining everything we've got from ChatGPT, like OpenAI, AI mode, Gemini and Perplexity.
With that being said, we have seen some differences between the different LLM where ones may have a different preference than the other. But because it's so hard to just manage all the differences of these, we try to give a point of view about all of them. But people who want to see even more detailed, ⁓ in Perplexity, I'm more known about this, or in AI mode, I'm more known about this. We have that capability.
But yeah, about 70 % of the prompts that we've seen from people are mostly coming from from an adoption perspective, not even like a...
Jon Clark (38:52)
Yeah, I think
Totally, yeah. I think back years ago, we used to get questions from clients like, do we optimize differently for Bing versus Google? And it's like, well, you go where the volume is, right? And ideally, you'll see movements in both. I feel like that's sort of the methodology here as well.
Baruch Toledano (39:03)
I know. Right.
yes, but it's important to see the nuances, right? They are evolving and it's important to see the adoption, 100%.
Jon Clark (39:17)
Yep. I wanted to ask a question about the, like SimilarWeb as, as a data tool. So you guys provide, you know, citation analysis, prompt analysis, sentiment. Based on this brand study, if you know, you had a client using the platform, is there one of those that you would sort of point to as a leading indicator of, Hey, these efforts that we're doing for AEO is,
Baruch Toledano (39:24)
with it.
Jon Clark (39:40)
you know, it seems like it's leading us toward success. Is there a way that you guys are thinking about those data points, you know, differently in this age of AEO?
Baruch Toledano (39:49)
Yes. So Reddit definitely is I would say, disproportionate point of view about what is a source. So definitely the second piece I would say is the...
Again, the consensus building, right? So is the consensus about your topic and about the brand in this relationship is actually matters more when Google, you know, we have a system in place, right? You rank higher. You are, you are more known for that. Like it's very simple. Google has given us a good judgment in these things. so we have to basically look at more sources to be able to determine sort of like the brand affinity to a topic.
We also are digging these days deeper into the source of citation that exists inside the answers because we need to really help our users to process so much information. There's so much to do that it's hard for them to do it. And I think this is exact role of where computers should actually give those insights.
With that being said, we don't want to take away the knowledge that the organization has about itself. So all we can do is surface where these things are, giving you the recommendation, or finding the alerts when things are deviating from the normal. So those are the key things.
You know, you want something to generate content, fine, go to ChatGPT and we'll generate the content. I just don't think like, like we could do it. We could do it well. It's not where I think at the end of the day, we don't take from the people. What we want is to give them the attention and direction to where their efforts will matter more when it comes to their own, to their own brand and to their own sort of product and what they want to be known for.
That's an approach we've been taking. Do we see a change with some of our client? I would say the fundamentals remain. Good SEO is good for GEO. Like don't put it down. Like your site has to be indexable. It has to be able to read it. You have to chunk the information properly for human as it is for machine. It's no different. Like people have to be, people don't want to read your long stories. They want to read it in different ways.
Jon Clark (41:35)
Yeah.
Baruch Toledano (41:50)
And then the other piece that we have noticed quite a bit is always think about the follow-up question. We see a lot of great pages, product pages that are showing up more when they are covering the entire thing. So the FAQ, the return policy, the fabric that this thing is made from, how to wash it, how to use it, so where to wear it. The more you cover this, the better we've seen you show up.
Jon Clark (42:11)
Such good advice. Think about the next question. That's good.
Baruch Toledano (42:14)
Yeah.
Joe (42:14)
John will tell you, I don't, oh man, I'm gonna make fun of myself for not being fashionable somehow. But I thought the section you guys had on fashion was the most interesting because there's not a lot of differentiation. Brands have to market themselves to be different, a pair of blue jeans, to me, there isn't a lot of
Baruch Toledano (42:18)
Ha ha ha ha.
Jon Clark (42:20)
Haha
Baruch Toledano (42:23)
Yes.
Joe (42:33)
difference in blue jeans across the hundreds of companies that make them. But you guys show like this is probably the most competitive industry right now. You can be helpful to people in thinking about how to put outfits together and you give them information about materials sourcing of materials. It's like there's a huge opportunity for a fashion brand to just break out of this pact right now.
Baruch Toledano (42:57)
Couldn't agree with you more. And we saw it in one, I think one of them was like, how, what, why, or one of these, these comp, these brands. And it was like, why is this brand winning? And when you're looking at the structure of the content, almost a combination of editorial and e-commerce in one place, in really what top of mind to people. And it goes back to our beginning conversation, which is like, website do not have to be structured per se, like a department store.
They have many types of experiences that they can bring. And I think they're doing a good job and it's easy for the LLM to just like, I can index them properly. They have the right context on the embeddings of the words and the relationship between lifestyle and what to wear. And so it's like very much like right there for them. So it's, it's, it's gaining the benefit of actually being more visible in many ways. The other piece I think we found that was last year, I don't know if it's still happening, but like,
they were like really niche website, especially in travel. I think one of them is called Rome 2 Rio. Rome 2 Rio, I don't know if you know them, but like it's like programmatic SEO that tells you the best way to get from point A to point B, right? Like, do you take a bus? Do you take a train? Do you take a bicycle, right? Like what is the best way to get from point A to point B? And they do it through many, many things, right? I want to travel between Toledo and Madrid.
I want to travel between... So it gives you all the option. And so all of a sudden it was over indexed because a lot of people who came to travel and wanted to create a travel itinerary, gives them a full travel itinerary referencing this is the best way to actually go and give you the answer. It's like fascinating, like this website nowhere to be seen in Google of a sudden, like a whole thing in ChatGPT.
Jon Clark (44:34)
Yeah, it's really, really incredible. Will Critchlow just put out a, I think one of the first studies that SearchPilot did, or I guess AB tests. There's been this, you know, historically, like everyone wants the pages to be leaner and less content, more visual. And the test was pretty simple, like add more content and see if LLM visibility improved or, you know, leave it sort of trimmed up. And of course, like the more content, right?
product features, all the things you were mentioning, the test sort of blew it out of the water. It makes sense. You're just giving them more to be able to return. Yeah. Well, Baruch, this has been amazing. I love digging into data like this. Do you have time for a couple of rapid fires? All What's the habit from your Magento days that still shows up in how you lead today?
Baruch Toledano (45:04)
Yeah, more to work with. Yeah.
Sure, let's do it.
Okay, so I think that the two thing that I take from Magento is like community matters. Community matters because you can build the best product, but it's the best product in your eyes. It's not the best product of what people actually want to use and where they needed to apply it in their job. And so I learned this like really closely at Magento,
where it's an open source, where people tailored it in ways I could never believe. And it was something that I carry. And this is why I work with many of these different influencers, because I want to hear from them what they need. And they are a pinnacle at pushing the envelope in many ways. But we also work with users that are just at the beginning of their career. So to me, community matters because it's really important.
The other piece I take a lot of course is scale matter, you know, bigger organization, more complex way of looking, bigger website is a reflection of the same thing. And so yes, I learned a lot of lessons from complexity there and I bring it to my team when we design and we think who is the person on the other side? What kind of websites are we have in mind? They like it because it gives us a direction of whom we're targeting the capabilities that we're building
and what we want to be great at.
Joe (46:32)
What's the next data set you and your team want to go after to make the tool even better?
Baruch Toledano (46:37)
I don't know if it's a tool, but it's a data set. It's the word of mouth. So if you could tell me what, what you told, what, what, you know, Joe, you told John before, if I could get into those, maybe that would be the other way to actually expose, because that's the best form of marketing still to this day. Something that is in your conversation is that it's a very powerful thing. There are many data sets that we want to bring in. There's no shortage of those.
Joe (46:54)
Listening to Alexa devices or
Baruch Toledano (47:05)
I think the question for us is more, this data useful and how would we adopt it in the context of what people do day to day and will it change their decision? So that is the key thing that stand for us. We adopt AI and how we work as well, quite a bit in everything we do. Like we analyze a lot of calls with clients to understand what really, what are the themes that show up. And we anonymize them, of course, we don't want to necessarily expose, but like,
this is exactly the kind of things we could have a lot of data that, but this, this, this data helping us is the key, is the key. And does it help to our users? is always my question. But yeah, if you give me word of mouth, I still think it's the best.
Jon Clark (47:36)
Right.
I think Elon Musk is creating a brain chip that might be able to help with that at some point. What was the most surprising industry finding in the That as you were looking at the data, you were like, wow, I didn't expect to see that.
Baruch Toledano (47:46)
Hahaha!
I think the fashion, the fashion is to that to me like, you know, Apple is in electronics. So that was kind of like obvious. I will see them at the top. That's very clear. But the fashion is, it's always been a competitive space, but it's gotten so interesting how you can disrupt it. And I think this showed us less of a, yeah, Nike is up there, Nordstrom is up there.
Jon Clark (48:03)
Mm-hmm.
Baruch Toledano (48:17)
But there's so much room for so many more and there's so many nuance that people have that I was like, that's why there's so many types of brands out there and they're all competing for very nuanced thing. And seeing them doing it online, it's also a very fascinating thing. And there's a good reason why this is a much more, I'd say distributed category because of the needs of what people.
Joe (48:35)
you
Jon Clark (48:38)
I thought that category was amazing.
Joe (48:38)
Can you share a piece?
Can you share a piece of advice someone gave you earlier in your career that you you'd like to pass on?
Baruch Toledano (48:46)
Never give up. If you give up it's like the success is not happening at the first trial. I think it's everything that I've seen. I do it with people. do it with the way we build always a way to solve something. We just have to put our mind behind it. And so I don't give up and to me, sometimes the collectives is bringing a lot more
approach to those problems that we have. And so if I don't give up, somebody else might have the answer and I bring them in and I don't try to like the same people sort of addressing the same problem, because I'm in the business of solving problems. So that's what we do. So I don't give up.
Jon Clark (49:19)
I love that one. I tell my daughter that every day. Never give up. Never give up. this has been so much fun. Let our listeners know where they can find you online and if you're going to be speaking anywhere next.
Baruch Toledano (49:22)
Never give up.
Yes, so SimilarWeb. You can can find me on LinkedIn. You can find me, I'm not very, very active, but I'm active sometimes with things that I think are newsworthy. I don't want to fight for, fighting for everyone's attention is difficult. You can find me in a bunch of places, BrightonSEO, I go to Tech SEO Connect, I was there. Amazing group, amazing. The SEO conference there is just like fantastic.
And Mike King upcoming SEO Week, I'll be there too. Not speaking, but definitely meeting some of the folks. So yeah, would love to connect with people.
Jon Clark (50:02)
Yeah, hopefully we can connect in person here, either at SEO Week or Tech SEO Connect. Well, thanks again for joining us on the Page 2 podcast, it was a pleasure. And for those listening, if you enjoyed the show, please remember to subscribe, rate, and review. We'll see you next week. Bye-bye.
Baruch Toledano (50:06)
Amazing.