B2B SEO: The #1 Reason Brands Actually Get Recommended in AI Search — with Gaetano DiNardi

we've been saying all along for years about trying to educate the 95% of the market that's not ready to buy. And AI SEO is kind of the opposite of that now. Exactly. AI SEO is the opposite of all the things that you guys push clients to do. This is like a thing that like every company is struggling with. When I tell companies, look, we have to eliminate top of funnel entirely. They can't believe it. We have to go all in on the bottom of funnel. There's just no way around that. And it's a very tough pills as well. Hey everyone, welcome back to GTM Live. We're really excited for the guests we have today. And if you remember some episodes back in the spring, when we got on the event circuit for this year, this is someone who we said we were definitely going to have on this year. So I'm glad it's coming full circle to make that happen. We have Gatano Denardi. What's up? Yo, Gatano, thank you for being here. And I want to give some context because you're an OG. And I want to make sure that everyone knows exactly who we have this heavy hitter on the show today. So if you've been in B2B marketing for more than five years, I would say you've seen his fingerprints somewhere. This is the guy who took sales hacker from a scrappy blog to acquisition. He helped Oregon go from zero to hundreds of thousands of monthly visitors within the span of a year. And he advises dozens of B2B SaaS companies this month every month, really in demand. And right now SEO in GEO, his company is marketing advice.io. So I also just want to say again, like I said, throwback to anybody who's been listening to this show across the iterations over the years. Gatano, you know, correct me if I'm wrong here on the year. But I'm one of those people who's been listening to, you know, the Chris Walker show for so many years now. But I don't I don't necessarily predate back to a new and Chris first had the state of demand gen calls. And was that in 2019 or 2020? You tell me, but I was one of those people who I wasn't listening, you know, I wasn't listening at that time. I wasn't savvy to it. But I would always scroll back like down into the podcast archives and say like, where does this go? Like how far does this reach into the past? And I remember hearing some some sessions that you guys did. And it's just amazing to like now know you, you know, in person. And so thanks for thanks for being here today. Thank you both. I really appreciate the intro. Very warm fuzzy intro there. And yeah, you know, I go pretty back with Chris. Yeah, we started it as COVID was unraveling, you know, even before COVID was unraveling, we did a session in Miami that was really cool. Just a deep dive, you know, living room style chat that we published on all things growth marketing and demand gen. People really enjoyed it. So we realized we had something cool, no real goal in mind other than to put out great content and, you know, track the audience. That was really it. And then, you know, COVID happened, then we started doing demand gen live. Every week it surprisingly grew, started kind of small. And then I remember we were getting like hundreds of people joining live sessions every week. You know, it was a mix of marketing knowledge plus entertainment, you know, we would air out the laundry of our grievances, you know, with the industry regarding things we were annoyed with and stuff. And a lot of people found it relatable. Also, there wasn't much to do during COVID, you know, events were completely shut down. It all went to digital only. So I guess we became one of the, you know, most exciting things you could tap into during the COVID era. And then from there after that, you know, it all just took off like a resource. So yeah, I think those were some fun days and believe it or not, I still get client flow from it. I still get people saying, man, I still, yeah, I still get people saying, man, you know, I thought of you. I remember you from the Chris Walker, you know, a demand gen live. I said, wow, that was like five years ago. And it's, you know, so, you know, the analytics are crazy. Sometimes I go back and look, obviously, because this podcast is like still riding, obviously the same listenership in the height of that show. I think it was averaging about 25,000 listens per month, which is pretty wild. Yeah, it was like just crazy sometimes to go back and look and you know, see what people are, we're listening to you at the time. So really cool. The reach was surprising on that. I mean, 25k a month is pretty wild. That's the thing. You just never know who these interviews are going to reach. And you just never know when it's going to boomerang back around. So, you know, what a special moment in time thinking back to, you know, the 2020 era and you and Chris were took, you took the opportunity, you know, and you took the risk and we're all the better for it. So thank you. We'll definitely link some of those throwback episodes in the description for anybody who wants to listen, who hasn't yet. But one thing I really admire about you, Katana, before we get into the content today is really framing around, you know, as a professional, how you continue to evolve to, you know, meet the market where it is and stay with your finger on the pulse. And you have really, yeah, you've, you've, you just continue to like evolve and get better and better and better. And that's why you want, you're one of the most elite experts in the SEO space, right? As we talk about GEO SEO. And we've been seeing this as a problem that a lot of our customers are facing around what questions, you know, to ask. And so, we definitely wanted to have you on and I'll let Carolyn chime in about that. And yeah, what kind of sparked us to bring you on now? Yeah, totally. Well, I mean, I'm sure you see this, Katana, but I feel like there's just so much, I don't know, like content fatigue around the topic of AI Surger, you know, SEO or GEO or whatever you want to call it. But I'm also, also just seeing just how profound of a challenge it is for customers too. And obviously, like being in the data of Pasetto clients, I see firsthand what they're seeing, which is, hey, organic search is declining. Hey, AI search is either flat or it's not growing. Like, what are we doing wrong? And unfortunately, a lot of these companies also have AI SEO agencies doing this for them. And then, you know, when they run into this challenge of like, well, why aren't we growing? The response that most of them get is just, well, AI search is changing the landscape. And I feel also too, like majority of the content is just really around, well, AI search is just changing the landscape. And it's just, you know, what do people in a CMO role who don't have their, like, you know, fingers on the pulse of what actually to do when, like, on the keyboard, you know what I mean? And so I guess that's where I'm coming at it from our listener, the lot of senior level people, like what WTF is happening in the world of SEO right now? Yeah, that, I mean, you basically nailed it. That's the biggest toughest question that exists right now is our traffic's going down. How do we know if this is working? Our agencies are just telling us that while it's AI search, disrupting the landscape, that, that is a oversimplified true, I mean, that is true, but it's a very oversimplified explanation and it doesn't scratch beneath the surface. And so that leaves CMOs and really any marketing leader flying blinds some extent. If you don't have somebody that knows how to really diagnose dissect what's going on and explain it in super simple terms, it's going to be difficult. And so, I mean, I don't know where you want me to start exactly, but it's a very big topic area. I can really slice this apart in many ways, but I don't know, maybe to start, I can just explain to you the most simplistic breakdown of like what actually is happening, what actually works and what to think about it all. Yeah. Okay. Do that. And then I have a immediate question right after that for where I want to go next. So yes, tell us the 30,000 foot view of the world. 30,000 foot view. Yes, traffic is eviscerating, but just put that to the side for a moment. Just skip that for right now. We'll come back to it. Because if I go down that rabbit hole, it's going to open up a lot. It's a massive can of worms, but we'll just put that to the side for now. Okay. Cool. If you put that to the side, and you just think about the ways in which you can achieve any kind of visibility through LLMs, it's very, very, very basic. There's just three main ways. The first and most obvious way, the easy, and this is like the easiest, most trackable way is direct referral traffic from an AI search platform. So if you're completely in the dark about this and you just want to get the very basics, you can get this tomorrow today in any, you know, analytics account. You have the referral sources AI search. You can see, you know, from chat GPT, Gemini, Claude, etc. How much referral traffic are we actually getting from these places? This means somebody clicked from a chat interaction to a URL in your site, and they landed on your site. From there, you can also know. What happened? Did they convert directly on a page? Did they spend a lot of time on the site, you know, through extended duration session? How many pages did they interact with? What pages did they actually go to? Right? So just that alone is like easy, peasy step one. If you want to just have a basic understanding of how much traffic is coming from these systems every month, where does it go and what does the traffic do? You can find that out very, very easily. That I would say is like for anybody in the dark, easy step one. Start there. Now let's go into the next portion. So that's very what I just explained to you is very direct response, the linear easy, right? Now the next one is really tough because there's none of that. Now you have the zero click answer, answer recommendation, right? So this could be vendor short listing, vendor short listing, discovery, narrowing, the vendor exploration inside of a chat interaction, whether that's AI mode, you know, you ask it, you know, what is what are the best third party risk management solutions? You know, you'll have in most cases the vendor short list selection from the AI tool and there may or may not be links to those vendors from those answers. Most of the time there are no links or the links that are present are from other kinds of pages. So from there, you are now in this sort of dark SEO funnel because the AI platform listed a set of tools and options for you. It's also going to offer follow-ups. Like it often wants to know, like, tell me your situation, what kind of company are you? You know, are you looking to do this, are you looking to do that? Are you thinking Bivers is built? How many, you know, organization members do you have? How many seats will you, like, it's going to, you know, prompt you to continue the discussion so that it ultimately can get a better understanding. Like, it already kind of knows all that stuff through personalization and memory context anyway, but it really wants to dial it. So if you search a very basic question, yeah, it's going to give you the options, but it's going to push you to give more. But in any case, what ultimately will happen most of the time is that people won't click those links. And if they do click the links, the links generally supplying the answer set for those questions are not always the vendor pages themselves. It's, you know, the most of the time I see that it's a Gartner page or a Forester or maybe it's, you know, some other kind of review site or analyst site or some other kind of like affiliate site, maybe anyway, then they go search Google directly for the brands and then they'll start searching comparisons and reviews and pricing. So the way to think about this portion is that AI search is the discovery layer, then they go to Google to verify. They pretty much will only go to the actual website of the vendor once they're well informed, like at the very end of the process, like that's the difference. The discovery is happening elsewhere. The narrowing down is happening on AI search. The validations are happening through Google search, you know, comparisons and pricing and reviews and other places. They only really go to the vendor website at the end. They're not really spending too much time like sifting through like vendor content pages and stuff like that. That's the stuff that's getting used to serve like the bots, right? And so you have this dark SEO funnel happening where there's no attribution really to any of this. That's why you see so often now like conversion paths being so simple. Like when someone's ready to convert just homepage, request the demo, you know, pricing page, request the demo really quick because they already know what they want to do. So that's the other phase of AI search. We talked about direct referral traffic. We talked about the zero click, direct SEO funnel, and there's a third portion as well. Yeah, it's so relatable. We actually got, we were on a call earlier today where this came up around the dark, the dark, you know, channels. And you're you know, bringing this to light as one to keep top of mind. And it's just so interesting because I noticed that the status quo is like, we'll reject like, oh, we can't measure it, you know, the status quo will reject, you know, there's this knee jerk reaction to like, oh, well, then how do we attribute, you know, how do we attribute it and how does that show up in our attribution model? And so, Carolyn, I'd love to hear from you about, you know, the stance that we take and collectively as well around how to measure, you know, some of these things that at the end of the day, you're not really going to get a direct measurement for every single question. I'm going to, I'm going to jump in with that in a little bit, but I'm definitely, I mean, what you're saying, Gitano definitely reinforces what I'm seeing, which is when somebody comes to your website and they, you know, fill out a form, they book a demo and they become pipeline, like those always convert the best, obviously, better than any other source because people are raising their hand and they want to buy, but usually what companies are wanting to do is understand what got them there in the first place so that they can do more of that. And it's really common for me to see that, you know, 90% of the time, there really is no footprint for those people. They come to your site, they convert, and that's that. And there's really no like, you know, series of touch points that happen before that. And I think that a lot of that feeds into how people are, you know, doing their research, no, just in these untrackable places. Is there another sort of tier on top of that one? I think that was the second one, right? There is. There's another aspect. In my opinion, this is the lowest value aspect, but it's the aspect that gets talked about maybe almost more than anything else. And that's the aspect called citations. You guys probably hear this all the time, citations, a citation. All it means is that the AI crawled your content that it ingested your content and that it used your content, your URL as a source for serving the information, like it used your page to get the information it wanted to create a portion of the AI answer summary. Now, what is the problem with this? The problem is that you may get cited, but not recommended a citation without the recommendation is like, you know, having a backpack to put your money in, but no money, right? Like it's like you go to the bank, you walk in with your backpack and you walk out empty, you know, it's like that's the way I view it. Now, there's a lot of different opinions on citations. And by the way, citations can be heavily gained. It can be heavily gained. And that's another reason why I don't like it. But the problem with citations is really that you can be cited all over the place. But if the AI just uses your page to extract the information and it doesn't use that page to create a recommendation that reflects your brand, it's almost pointless. So like a great example of this is like, let's say, you know, let's let's say you are, I'm just going to use random brands, all right? I don't have any affiliation with these brands. I'm just going to use these brands as an example. Let's just say that you are fresh desk. And you create a page that says, here's the top 10 alternatives to Zendesk. And your goal is to put yourself at the top for best Zendesk alternatives so that the LLM's find fresh, fresh desk. And they will use that page to recommend your brand in the answer. Now imagine that your fresh desk and you produce like the perfectly optimized page. And let's just say you build backlinks to it and the contents like decent. And over time, you actually get it to rank really high in traditional search. So then what ends up happening is AI mode starts citing that page more because it's like, all right, this is a high ranking page. We're going to use this as a basis to extract information. If it passed our Google traditional ranking system filters, we have to assume it belongs here for a reason. But now, if the market doesn't believe that fresh desk is one of the most credible alternative options to Zendesk, meaning like there's not a review profile that supports it. There's not, you know, a G2 profile supplementation that supports it, that other list pages out there aren't talking about fresh desk as an option. When it comes to Zendesk alternatives, that there's not a set of verses and comparison pages out there that are comparing Zendesk to fresh desk. All that sort of supplemental digital footprint is absent. There's nothing you can really do to jam your way into that AI answer. Like, so this is the thing. GEO is a category level eligibility problem. It is not an SEO ranking problem. That is why you can rank your page number one alternative to Zendesk, but the AI systems, if they don't believe that you deserve to be there, you just somehow game the traditional search ranking algorithm, which is still possible to do. You'll rank, you will get cited all over the place, but they'll just take all the competitors off that page. They'll mention everybody else except you and thank you, nice doing business with you, you just spend all that effort on content and we just scrape it, serve it up and we're not going to recommend you. So that's the problem with citations, everybody's making this big fuss over citations, like citations to me are almost meaningless. So, you know, summer making the argument that like, you know, if you see all over the place in a search result that it's like a digital highway billboard. You know, the argument that summer making this is a stretch in my opinion, but they're saying, you know, the more they see your brand, it's like a billboard ad, you know, it's like a highway ad. And they're more likely to recall your brand down the road, you know, so, so just getting the brand impression there is great. I don't know one CMO out there that's interested in hearing the impression report. Are you saying then that people tend to like over index on just creating a shitload of content to try and get cited without realizing like the other variables to go into that. Yes, that's exactly it, that's exactly it because I have clients right now that are getting 95% brand recommendation rate for their category. But guess what guess what guess what their citation rate is for their own first party content, guess how much the AI tools are relying on their own first party content to recommend them 95% of the time for a high volume competitive category. Take a guess I don't know like 15% like 5% yeah, how and why is that because it's all derived from third party sources like there's a PC mag affiliate page and another affiliate page from a company called black and white zebra. And a couple YouTube videos and a gardener page that's like driving everything. So that's another thing that we need to actually point out. This this this claim that oh 95% of AI searches off site it's third party that is very, very misstated like that that is very it's misleading and I think it's just flat out deceptive because now it makes marketers think oh we don't need to worry about on site content. We need to do 95% of our efforts on third party marketing that's not the case like it's all category dependent and it's all search intent dependent like for this particular search, which is employee monitoring software best employee monitoring software. To do is crack a couple of sources you need to basically get a relationship with PC mag and paid to be on their affiliate side that you know you need to find the other affiliate listing out there from black and white zebra you need to pay and bid to be on there. You know so that that's just like two pages if you get on those pages and you build up your review profiles on the corresponding G2 and gardener pages and that's all you need to do so this idea that you need to you know get thousands of mentions and you know work with you know geo vendors to build citations and stuff like it's it's all garbage. Okay, and would you say that like most geo and SEO agencies tend to operate that way yes, sadly they do i've been told on calls because i have rescued customers from dealing with these companies like i've heard. We don't have ours right you said that yeah yeah we talk about that so maybe let's use that because I feel like this this one customer that I think you're starting to work with and the near term sort of like fell into this trap or you know like problem that. You know a lot of companies are experiencing but they had an SEO agency so like when you sort of jumped in and you know evaluated their site that you know they're an enterprise software for I would consider like a pretty mature category. What did you see right off the bat that maybe they're current agency or their former agency I should say was doing that was not the right approach well I think that in certain cases. What you need to do is actually very cut dry and it doesn't need to be this big complicated thing like for this particular site I look that it is a very simple straightforward site there's a you know couple. Categories couple product pages that really drive everything a lot of branded search most of the blog content that was created was just high funnel informational stuff so that's another thing with AI search by the way. AI is destroying top of funnel see that's what's happening with traffic we need to just go there for a moment. Because AI is actually the normalizer it's the great equalizer so most B to B SAS strategies for SEO historically we're focused on just driving as much traffic as you can. What happens to drive the most traffic is higher funnel informational traffic because those searches have higher volumes see the lower funnel you go. The lower the volume gets so traffic volume becomes less of the thing as you start getting more dialed into bottom of funnel so what I basically see almost every time when a new company comes to me and they say they're unhappy with their agency. It's the mere fact that this agency hasn't figured out how to get them focused on bottom of funnel demand capture. They're still doing this bloated model they're not taking into account that producing informational thought leadership on corporate blogs is completely dead. You need to move top of funnel now to rented land your own land is not a place for top of funnel anymore top of funnel is moving it is moving to YouTube it is moving to social media. Is moving to LinkedIn posts right it is moving to substax it is moving to X is moving to be high news letters you know it's moving to all these other places like your corporate site is not the place to do you know perspective content anymore like you know those short corporate blogs that is like a perspective from somebody internally it's like a 500 word post that should have just been posted on their LinkedIn that's the kind of stuff that like is dying. Anyway I feel like I'm going too much on a rant on this but obviously it heats me up because if you're if you're investing a lot of effort in top of funnel SEO strategies today it's just not going to work. So that's where we're at wow I mean that's a lot it's a it's a big shift in the market and in some ways I almost feel like. It can be a little counter intuitive to what we're telling people to do in the way of like more brand advertising as well which is just casting a wider net around. What it is you do instead of just focusing on demand capture do you know what I'm saying like I feel like in some ways it is kind of counter intuitive to what we've been saying all along for years about you know trying to educate the 95% of the market that's not ready to buy. And SEO or AI SEO is kind of the opposite of that now right exactly AI SEO is the opposite of all the things that you guys push clients to do. And this is this is like a thing that like every company is struggling with like when I tell companies look we have to eliminate top of funnel entirely like they can't believe it. I'm like yeah we have to we have to go all in on the bottom of funnel there's just no way around that and it's a very difficult to swallow when I tell them look what what we need to do to begin. We need to do narrative control that means we need to do all of our comparison us versus vendor XYZ pages and get those in tip top shape we need to do alternative alternative listing pages. Similar to what fresh desk did to send us because the example but. And what we need to do those pages for is not necessarily to just promote ourselves as the best. We need to do it for narrative control. We need to be able to push our unique differentiators over and over and over again. So that LLMs, you know, they essentially you have to stuff the ballot box with LLMs because these things are statistical probability machines. Right. So, you know, like I use like a AI visibility monitoring tool for all my client work and I have all my clients plugged in. And the way that I explain like that portion of AI that recommends the brand is look, you have to think about this in terms of a percentage or what percentage of the time. So if you ask any AI tool the same question 100 times manually, you're not going to get the same exact answer every time, but the goal is to have a majority stake. So like, you know, 90 out of 100 times your brand is going to recommend it and the narrative is consistent. That's what you want. It's never going to be 100% but I show them by topic and by category. Here's the recommendation rate over time across platform platform platform platform platform. That's the way you measure AI search. So, you're not going to get that though, unless you track bottom of final prompts. Can you give an example of that? Yeah, yeah. So let's just say let's use third party risk management, imagine you you track and a prompt tracking tool, something like, you know, what is third party risk management, it's not going to recommend any brands. It's just going to explain the information and it's going to define it and it's going to give you, you know, some supporting explanation around, you know, the concept of third party risk management. It's not going to it's not going to say on its own. And also, if you're interested in seeing vendors here are the best options, it doesn't do that. It might do that occasionally, but it's rare. So if you track a bunch of informational props. It's you're going to see your brand recommendation rate be like 5% of those. But your competitors are also going to have a 5% recommendation rate because it's just not the kind of question that would trigger a vendor response. So if you just search things like what are the best third party risk management solution, what is the best, you know, supply chain risk management platform, so on and so on and so on. And you track a couple of variations of those things, you tag it by the category or the topic level. And you have a set of data that's worth working with because you have questions that you know the tool is going to have to go and reply with vendor options. And you want to know if you're showing up in those vendor options or not. And so it's only worth tracking bottom of funnel prompts in your tool and it's only worth doing bottom of funnel that that's the whole point and you also need to align it to your category strengths. That's another thing that's changed about SEO and B to be marketing in general is this idea that like you could stretch and reach as far tangentially as possible to connect a long chain of dots to your product use to some topic. It just doesn't work anymore like I don't know if I can think of a good example of the top of my head, but maybe we'll use the employee monitoring. And if you have a solution that is state of the art best in class and their flagship offering is employee monitoring solutions. But they also kind of offer endpoint monitoring where they also kind of offer compliance monitoring. These topics are like really outside of the realm of relevancy like you can maybe make the argument that we can drive traffic for that stuff, you know, to attract the ICP. But that's that's dead now Google doesn't want you doing that they want you staying within your home stay within your realm of relevancy don't stretch and reach a stray don't topically drift. That's why it's hard for companies that are like coming out with new products to like reposition and rebrand and you know go go to market for different product lines. It's really tough because it's kind of like you're you're a big company, but you're starting over again because you have to rebuild that whole digital footprint. And I see that all the time. So well, yeah, I mean like when I hear all of this, it kind of makes sense in my mind, but it feels like a lot to like wrap your arms around if you're not if you don't have, you know, maybe not the resources or the technical expertise. So it kind of just like begs the question of hey, if I'm a VP of marketing or a CMO and I'm not seeing the results I want to see in the way of SEO, like where the hell do I even start where do I start. Yeah, well, the first question is how do you know you're not getting the result. Yeah, let's look at it the way like our, you know, most of our customers would look at it from. Okay, I'm not generating. I don't want to say leads. I'm not generating as much inbound. I'm not hitting my inbound targets. I need to get my inbound volume up. That's one. When I look at my analytics by organic search is not growing. I'm investing in SEO. Why is my organic search not growing. And even if I look at AI search, you know, my CEO, and I'm finding I'm not getting recommended. Like I would say the combination of those three things combined would usually indicate we're not doing something right with the SEO. Sure. All right, we have to address all three of those. So the first one is that traffic is going down. We look at our Google search console traffic is getting killed. All right, almost every single time that I've done this. I have found that see B2B SaaS websites have a ton of clutter in their traffic reporting, like a ton, like I got to just pull this up. I don't know if I can share my screen on this if I should share my screen. Let me just tell you guys about the maybe I'll just tell you guys about it. Yeah, yeah, go for it. So I just audited a really, really big company, massive company, this very complaint here, our organic traffic that is down 50% year over year. That was one of the chief complaints. And we're not talking small numbers here. We're talking like we went from a million a month from Google search to 450K a month in Google search, like we're down massively. We have no idea what the cause is, how this is affecting business, like all those questions, like coming up and like it was just panic, right. All right, first thing to do break down the traffic decline by country, almost nobody looks at this right off the bat. I saw, you know, 100,000 visitors lost from Pakistan, 100,000 visitors lost from Japan, 100,000 visitors lost from India, 100,000 visitors lost from Philippines, almost all of that stuff. Like out of the like from the geos that mattered, it was tiny. It was like, all right, you lost 500K traffic, but 400K came from geographies that you don't care about. Were you aware of that? Oh, wow, no, I didn't know that. I was like, all right, is India, Philippines, Pakistan, are these strategic priorities for you? No, not at all. Okay, there's one. Next. Did you know that the majority of pages that lost traffic were careers, certifications, community, downloads, events, forums, support, videos, things like that? Did you know that that's like we're almost all your traffic last came from? No, we had no idea. All right, great. Do you know that when we apply all those traffic filter exclusions and we isolate two United States only that you are only down 1,000 clicks from last year? Did you know that? They're like, no, I'm like, and it gets even better. Do you know that if we were to isolate to non brand only and if we were to isolate to content pages only, do you know that you are only down about 100 clicks from last year? No, we had no idea. So just think about that. They went from full panic mode. We're down 600,000 visitors a month. Things are collapsing. We know we're screwed here. I come in audit this thing and say, guys, you're barely down. I'm surprised that they don't know that already. They don't know that they don't know that they don't know how to look. They have amplitude. You can find this out from amplitude. Just ask it and it'll tell you has this agentic feature. Google Analytics now has what are they calling it? They advisor the little the little assistant there. You can ask it. Nobody knows how to look and ask in filter. Okay, that's traffic. Yeah, so that's traffic. I guarantee you companies are not doing this level of triage. I guarantee it. Yeah, you know what though? I know, you know, folks probably really like value. your perspective on all of this, but I feel like you could just literally export all your Google analytics and like jump it into like clawed. And it'll tell you in two seconds. Yeah, for sure. Okay, sure. So that's that's traffic. The next one would be what was the next one I had said? I said, you know, high intent inbound leads though, they're flat. They're not growing like we want them to grow. So I'm not saying that this is going to be the case for everybody, but I'm going to say that it could be the case for a lot of people. You have to think of the overall channel mix. And you have to think about how these channels play with one another. And you have to think about the the household. That's what I would frame it as the household, you know, so if somebody in the house is really doing really well financially and somebody else in the house is doing really poor financially, you know, and then somebody else is doing medium, right? Is the household in a positive net game? Or is it a net negative for everybody the whole thing, right? Yeah, yeah. So when they say high intent inbound leads are down, you have to isolate. Is that from organic search only? What happened with direct? What happened with paid sources? Are we overall trending net negative because of channel performance across the board is experiencing these drawdowns? Or is it just SEO or organic search? You know, a lot of times they don't even know that. But I'll go and look. So when I report for clients, I report on total handraisers every month from organic search. This is the channel that I'm owning for clients, so I'm focused there. Yeah, it makes sense, right? So for this particular client, I just produced a July report and it showed, whoa, something happened in July. July tanked. Like we fell off a cliff. We went from like 350 qualified handraisers from inbound from organic search to like 190. So we dropped by a huge amount. Next, I said, all right, that's really fishy. Like, I haven't seen a drop off that big ever for this client. What the hell is that? So I go and look at traffic. And you have to, when you have a drop off this big, what you have to go do is look at the pages that normally drive a lot of conversions. The first clue is the homepage. For most B2B SaaS clients, almost always the homepage is like top of the list. Sometimes it's other kinds of pages like pricing or just get a demo or maybe a solutions page is really strong, but generally, it's like in those buckets. But I see homepage traffic from SEO fell by like 50%. Very, very unusual. Very unusual. They didn't even notice. So I'm like, all right, what is going on? Let me go check paid search. What do you know? They decided to start bidding heavily on branded search through Google ads. What do you know that paid search efficiency has skyrocketed? What do you know that like cost per lead, cost per conversion is down by like 50% clicks from paid search skyrocketed. Conversions from paid search have skyrocketed paid search expansion by 70%. You know, massive spending. So now you have this interesting question on cannibalization versus incrementality that nobody's thinking about. If you just if you didn't go and check paid search and you resent that SEO report, you're going to look like a fool. You're going to look like a stumbling, bumbling idiot because you're going to present that something tanked. You can't explain why everyone's going to be scratching their heads and you don't even think to go look at other channels. So that's that's what happened with this one client. And so they still don't really know, you know, is this a overall net positive? Is this incrementality or is this cannibalization? I went and gave my perspective because I actually went and checked the auction insights and I showed, you know, some data that some competitors have gotten really, really aggressive against our own brand bidding. And this is a company that historically never believed in bidding on their own brand. They didn't believe in brand defense. They never wanted to do it. All of a sudden now they flipped the switch and did it. But this clearly showed almost imperfect unison the drop off in SEO and the massive efficiency gains in paid search. And so if I didn't go and check that, I would have looked like a fool. Right. So you know, it's it's not the answer every single time, but like a lot of times there, there can be, you know, an isolated answer. But other times there may not be. Other times it's just like, look, the deals are stalling in the pipeline because of bivers is build. I'm seeing that with a lot of clients too. And they don't like hearing it. But that's another reality that, you know, has to just be accounted for. Yeah, that makes total sense. Great, very cool story around that. I think that's an interesting one. The last one. Okay. CMO is seeing or anybody for this matter, it's like, we're not ranking in the chats. Why aren't we ranking? And first instinct is we got to go do more what like programmatic SEO. We got to go get more content on the site. You're saying that might not be the best play, right? It could be you want to get referred from other third party sites. Am I here? Am I understanding that correctly? Is that right? So kind of. Yeah. So, you know, this whole thing of I searched the CEO searched X and we didn't show up. There's so much diagnosing that needs to go in that. We don't know what the CEO searched. You know, we just don't know, we have no context. We just have no idea. Plus remember, you guys explained that it's a statistical probability machine. You know, it's not going to be the same every time, but let me just show you guys how this actually works in practice. I think I can show my screen and we'll just do a little exercise here. Yeah, yeah. Sounds great. All right. So I think I can just pick Chrome tab, chat GPT, and we're in there like swimwear. All right. So let me know if you guys see this. Yeah, yeah, we got you. All right. So let's do what are the best employee monitoring solutions on the market today. I'm just going to give it you know, something basic and open ended like that. And this is just going to be a reverse engineering process. So you guys can understand how it actually arrives at these answers, what actually happens and how people listening can think about the diagnosing process because I'm sure this comes up a ton and people probably don't know how to actually diagnose it. All right. So it came up with its vendor list of options, right? And it's going to give you best by default, best for this scenario, best for that scenario, blah, blah, blah. And then it's going to go at the end and say, I would actually buy my evaluation order would be this, this, this or this, depending on, you know, if this is your need or if that's your need. But this is actually not even that interesting. What is the most interesting is to go to inspect element and to get this part of the URL here at the top because we are going to use that to extract the network stream information. Because within the network stream information, that is where chat GPT is actually breaking out your question into many, many smaller questions. And we can see what it used to construct the sensor by debugging the network stream data. Now control F queries. All right. So let's copy all this information. Go back here. Let me go full screen. Now you want to ask chat GPT, please extract all of the sub queries used to construct this answer. I like where this is going. I've never seen it before. Yeah, from the network stream. Break it down by branded search versus non branded. Then give me a list of all the safe URLs used to form the answer summary. All right. Now we're going to paste the network stream data. And we're going to hit answer and now you will know what did it actually search to construct this answer and what are the safe URLs that it used? What is a safe URL just in like cohorts? Yeah. In non nerdy, non technical terms, it's basically a source. It's basically URLs that it deems to be credible and relevant that it's using to either create the answer or validate things in its training data that it wants to be, you know, duly certain on. Got it. Makes sense. Okay. All right. And so you're going to see that it doesn't just take a question that you ask it and take that question and go answer your what you want based on that one question. It interprets your question and breaks it down into, you can see right here, eight smaller questions, eight smaller searches. All right. Now you can see, you can see chat GBT zero non brand search. There were zero, it no purely non branded category only searches. Why did it search active track, teramine hubstaff active track workforce, you know, features 2026 features 2026 pricing pricing pricing pricing basic standard premium enterprise starter grow team. How did it know to go and do all that like people in their minds think of it as a different way. Does the category level search and then, you know, it finds vendors the same way people find vendors through Google. That's not the case. And then you see the safe URLs. It has all the vendor pages here, so where it's gone, gone that data for wow. This is where it went to go form the answer now in turn, now here's the question of third parties, right? You hear that? Do we need third party mentions? Well, in this case, the only third party that's relevant is Forbes advisor. And that G2 G2 categories, G2 comparisons, the verge. That looks like some sort of a PR related page, but that's the only one in here. You know, so, so you can see that that whole narrative of like, we have to do offsite marketing. And, you know, if you want to appear in a search like this, all you got to do is figure out how to beat the game on G2. Wow. So now, so now we ask it, how did you know to search for those specific brands instead of doing a broad category level search? You specifically looked for Terra Mind, Active Track, Hub Staff, you know, insightful, etc. How did you know to go and look for those brands? What caused you to do that? I mean, it's so basic, but nobody's taking these steps, right? So when your CEO searches X thing and we didn't show up, this, you need to take that person through this workflow. That's why they pay Gitano Denardi the big bucks. I mean, it's so basic, but people aren't doing it. So anyway, look, these brands did not come from web search results. They already knew these, right? So it said, so where did the names come from? Here we go, training data. The models pre-existing learned association with the category employee monitoring software. In other words, the model already associated the category strongly with those brands. It doesn't see, so what's happening which had GPT is they have training data. They have what I think is called internal knowledge graphs where they know, all right, Active Track associated with this topic good for this. That's our mine associated with this topic good for that. The reason it knows this is because of digital footprint that it crawls and it analyzes over a long period of time. So therefore, you need to lean into your category strengths. You need to align with the right categories. Your product pages need to reflect the language used by market demand and what buyers are saying. You need to build out this category level association with a thing in order to be recommended. There's no amount of tricks that are going to change how this works. You can't do lllm.text and answer boxes at the top of a page and schema. That's another big one that I got a debunk as like a big deal. You know, you hear of all these tricks and hacks, but meanwhile you can't use hacks to manipulate training data. It doesn't work. So so chat GPT does this as a anti spam mechanism, right? This is why that fresh desk versus Zen desk example I told you about, you know, if fresh desk doesn't do work, they'll have the digital footprint. They won't they won't get default search like this. So you can see that chat GPT knew those brands from the models prior learn knowledge. And then they use that knowledge to seed the search. The brands were not discovered through category level search. So it already kind of knows who belongs there. And then it just uses other pages to validate what it thinks that is the difference. That is that is what is is changed between SEO and GEO. It's all that right there. Wow. Okay. That's amazing. I learned a lot from this, but I have one more, I guess I have two more small-ish questions. One, how long does it take to actually develop that? What did you call it training? What did you call it? And just for simplicity, say, call it entity association. Okay. How long does it take to actually build entity association for one realistically? That's my first question. All right. So it varies. If you're already kind of there, but you're not a category king, let's just let's just say you're not a category king, but you're maybe a category challenger. I would estimate anywhere between 12 to 24 months. If you are now if you are completely, let's just say you're a category dog, like you're just a brand new, you're brand new, like you've got no digital footprint, you've got nothing. It's going to take years. It depends on also how mature the category is and see, but that's also why you need to pick a lane. See, you need to carve out a lane. You can't just, let's just say it's, you know, AI app development. You can't just go for that. You're, you know, you're going to get killed. You need to pick something, something within that, which is narrow, achievable. What's your wedge? You can't just go for the biggest, broadest positioning. Yeah. I think that's one of my favorite takeaways, honestly, that I didn't, that I didn't really understand at all, being a positioning nerd myself. And I think it's really great to think about it, you know, like that, like what, you know, what is the lane and how are you being creative? So yeah. Thanks, Katano. And thank you so much for giving us the tactical examples too. So, you know, if you're not in front of the screen somewhere where you can actually watch, for example, on Spotify or YouTube, definitely go, go check out that video if you want to see the live example that Katano just walked us through. Super, super helpful to help tie some of those pieces together and, yeah, thanks, Katano, for all the tactical recommendations. I know we could go on forever, so let us know. Let us know audience what you want to hear more of. And when we should bring Katano back on, but Carolyn, was there anything else before we wrap up? Yes, I have one wrap up question because when we met, I think it was back in February, April or whenever, whenever we met up in Florida, you had said, this is not for everybody, not every company should prioritize SEO. So I just want to hear from you, like people who are listening to this, who should sink the time and the investment into prioritizing this and who should just say, hey, you know, focus on this other stuff first. So I would say that if you don't have your digital footprint, you know, well defined, if you don't have a well defined category build up plan, right? If you don't have a well defined digital footprint, I think that you're better off doing other kinds of marketing. So generally the idea is if you, you might be too early stage, like there's this idea that like, oh, if you're early stage, you can actually hack it faster. You know, you can just, you just get on other lists and, you know, use. uh parasite SEO to to rank through LLM's like it's not going to work. It's not going to work because there's no third party corroboration. There needs to be corroboration out there. So probably what you need to do if like you're really early stage or you're just too early stage is maybe founder led marketing. You need to maybe nail your your lane and make sure all your bottom of funnel pillar content is established on the site and is in strong condition. You you need to build up reviews. You definitely need to do that. You need to determine what gardener page do we need to be on? What G2 category is most important? How can we start ramping up reviews and credibility there? Um what about what about YouTube? Remember I said top of funnel has to move to rented land. You can start doing a YouTube channel and YouTube series, right? Like that's going to help a lot as well. Um you know write guest posts like I write on search engine land. You know I do it once a month and it just helps with overall credibility. You know I get my point of view across like maybe you need to do stuff like that. Uh events like the I think you basically need to have um at least 1000 sites backlinking to you. Wow. Okay. Yeah. If you don't have at least 1000 sites backlinking to you like quality sites, you're probably too early. Great. I think that's a good great threshold. Okay, awesome. So we do have a lot of folks in this community who are mature, huge and my benefit from people like you, Katana, so where can people learn more about how to work with you if they want to do that? Yeah. Um go to marketing advice dot IO. Um go to my sub stack which is sub stack dot marketing advice and you'll find me there. Amazing. Simple easy. Yeah. Great. I love your sub stack and everybody here should go check that out because it's so informative. So amazing. Thank you so much. That was awesome. I learned a lot. I know this audience is going to learn so much from it as well and as always so good to see you. Thank you. Thank you both. Thanks Katana. Talk to you soon.