Page Optimizer Pro
Get the free course that Kyle mentions on POP. Check it out! I’m an affiliate, so thanks for the support.
Welcome
Doug: Hey, what’s going on? Welcome to the Doug Show. I’m Doug Huntington, and today I have my old buddy Kyle Roof. How’s it going?
Kyle Roof: Oh, great to be here. Good to see you, Doug.
Doug: And it’s been a while since we chatted and I, I think back and I’m like, oh, we always have a blast. And we’re like continuing to talk after we hit stop.
And it is good to see you. You’re actually, you’re back in the US for a little bit, so we’re in a, a favorable time zone right now.
Kyle Updates and Business
Doug: But what’s been going on with you in the last, whatever, six months or year or whatever you want to share with me about professional stuff going on?
Kyle Roof: I was gonna about to talk about my personal life.
I’m back in Phoenix visiting fam. It’s been great. Uh, think about moving houses. No, I’m not actually. I love mine. Um, but yeah, no, things are good. Uh, professionally things are great. Uh, uh, the age old question is SEO dead. I think our last six months have been our best six months that we’ve ever had. So, uh, um, yeah, business is good.
Uh, running page optimizer pro, the software is, is my main focus. And then, um. Traipsing around the world. Speaking at conferences is the other part of what I do. So yeah, things are good. Life is good.
Doug: Very cool.
Is SEO Dead Again
Doug: Yeah, and I mean, we all have an evolution in our, our business life probably, and I have sort of a maybe around a five year time span of interest in certain areas slash you know, I, I just keep moving on to sort of different stuff and.
SEO got a little bit harder, probably, at least for me, around 2020 ish or so. Just everything took a little bit longer and I was kind of getting out of day-to-day activities. Number one, I sold some of my sites and was kind of moving on to like a content creator. Type role for the most part, still had my foot in a couple different areas.
In probably in 2022 to 23, AI was taking off a little bit more. And that made it even more of a disconnect for me and my skills, or at least the skills that I had in the past that were helping me rank sites and content, um, or AI content specifically, like made. Made it even more difficult and I wasn’t super interested in learning.
So from my perspective, I was like SEO was kind of dying. That said, I was still using search engines for a lot of stuff, and I also know small businesses, especially e-comm kind of sites. They still needed customers and they weren’t necessarily impacted in the same way. So I’ll just kind of open it up and like, SEO dead, we’ve probably heard the same like, uh, sky is falling situation every few years over and over again since maybe like 2010 or maybe before that.
So
Kyle Roof: where do we stand now? Every six months. Every six months Since 2010. I think s. Has died. Um, no, it’s actually, I think we’re past the panic of AI shutting everything down because all, all AI can do is enhance your skills. It can’t do SEO for you. It can’t really do anything for you. It can enhance the skillset that you have or you know, somebody else’s skillset and you can guide it in a direction.
But you have to have the expert that understands the output. And if you don’t have that expert that understands the output, it’s just gonna drive you into the ground faster than you would do on your own. And that’s, that’s. That’s all that it can really do for you.
AI Search Adoption Reality
Kyle Roof: But what I’ve found is that, um, with AI search, which I’m still not a huge believer in, um, just yet, but, um, in, in terms of adoption, but AI search is just really doing SEO again, it’s all right.
It’s, it’s, it’s even more the wild West than the wild West that we still live in, in, uh, in SEL.
Doug: And as far as the adoption, I mean, I was. I was like, ah, I’m, I’m still using, um, Google or some other search, or I’m going to a specific place like YouTube to look things up. And it kind of slowly happened where more people were being shown AI results in Google and when it really hit home.
My wife started saying, Hey, I’m seeing these. And she was like, I don’t really trust them. And then she started trusting ’em a little bit more, at least for certain things. So what about the adoption? Like, are people actually using AI search and it’s different? Um, when you look at. Google AI results, which I’m maybe not using the exact right term, uh, that we see in the results there in the SERPs versus someone on chat, GPT or Claude, where they’re like, help me plan my trip to Dallas or whatever.
Kyle Roof: Right? So a joke that I made in, in, in, uh, some of my talks last year was that I came up with an index on adoption. Uh, and I call it the Dad and Janice Index, but my father and his wife. And then I showed just a screenshot of a conversation that I had with them on, on the thread that we’re on. I was like, Hey, do you guys use, um, chat GBT or any kind of AI search?
And Janice writes back, we don’t know what that is. Yeah. Should we be doing, should we be doing it? And and the reason they bring that up is ’cause they’re in their later sixties, but they’re, uh, semi-retired. They, they. They’re upper middle class, they take two or three trips a year. They’re constantly online.
They buy stuff online. They’re very comfortable doing all that. Um, and they have, they’ve never even heard of Chad Chip team. Like, it just, it wasn’t any like, it just, I mean, they’ve heard ai of course, sure. But they had never had heard about, you know, that there’s, uh, something that they could use as a search engine and they could sign up for it and go do it.
That’s something that they had never even occurred to them and they still don’t use it. I’ve, I’ve now educated ’em a little bit more so at least they know what it is. But, um, they have zero desire to, to go and use it. And then I was like, well, what about like the AI features you see in, in Google? Do you like those?
And they’re like, what, what, what, what features? They don’t, they’re not even aware that those things exist. And so I think we’re a long way off from mass adoption nerds like you and I are easy to. Adopt things quickly. We like it, you know, like, like we like that kind of stuff. But, um, I think this time last year, um, AI search was less than 1% of all search, and I think it’s doubled.
So it’s like, like it’s like almost
Doug: 1%
Kyle Roof: now. Exactly. I think that’s where it’s at, so.
Doug: Okay.
Kyle Roof: Really, I think a lot of it is, is especially if you do like client side work, it’s a lot of vanity metrics for CEOs. They want to see their stuff in there, or you have vanity metrics for business owners. They wanna see their, their stuff there.
But, um, in terms of, I mean, of course there are use cases where people, your audience is using that, like your marketing to marketers, then Yeah, I would definitely, I would definitely be an AI search. But for the, for the average human, uh, I, I, it’s not, there’s not much there, I think in terms of, of return on investment.
Doug: Okay.
Echo Chambers and Marketers
Doug: Yeah, it’s so interesting the sort of echo chamber that people might be in and not realize, and I think maybe again, nerds like you and I, and we, uh, we have a. A lot of people are skilled in different disciplines, but I think like you and I have a self-awareness and and listeners out there, if you’re watching or listening to this, you fit into this as well.
I hate
Kyle Roof: to break through. You’re also a nerd if you’re listening to this.
Doug: Exactly. Exactly. And you’re also self-aware, which I think is a very good quality. And I was chatting with, um, a friend recently and I think, yeah, Matt Vei, it may have even been on a podcast. I, everything blurs together in my world. I think it was on a podcast, but he was referencing YouTubers that he, he watches and he was like, oh, it’s, they’re so big.
Like everyone’s talking about it. And I’m like, I’ve never even heard of 80% of the people you’re talking about. Just like Alex Hermo, which I don’t find too. Entertaining or helpful. It’s fine if other people do, but not for me. Exactly. But anyway, Matt was in a little echo chamber where he was thinking that is the conversation that’s going on in the world with everyone.
And I’m like, I don’t even know who those folks are, even though they have millions of subscribers and followers. And that sort of thing. And same thing in my personal finance world. It’s like common knowledge stuff is not common knowledge for most people. It’s just specific people that happen to geek out on a specific topic.
So I can see how we end up in a spot where we’re like, everyone is using this all the time, and I’m behind because I’m not. It turns out like we’re on the, the leading edge of this. Okay. So. As far as, um, people using AI search, it’s very small from what you could tell. That said, it is. On the sort of front of mind for marketers and marketers that are maybe leading a team at their bigger company or whatever.
So how does that fit in? Because like there is a, there is a market there maybe outsize for the number of people that are actually actually searching, but it is kind of moving in that direction. So like how can marketers or people working at an agency kind of position themselves and spend time a.
Proportional amount of time that makes sense from an ROI standpoint on, you know, AI search versus SEO or is there a big difference? So kind of a broad question, but I’ll just tee it up for you.
Kyle Roof: Um, one thing I was gonna say is, is we’re all very tribal. And, and like, people are like, you don’t know Doug Cunnington.
You know? And then they’re like, and you know, you have people like, like of, in my mind, of course, everybody knows who you are, but then I’ll be talking to people, I’ll bring you up. And they’re like, who’s that? And I was like, what? Yeah,
Doug: yeah. You
Kyle Roof: know, that’s crazy. That’s insane.
Building AI Kyle in POP
Kyle Roof: Um, but you know what, where, where I think it.
Again, using it as a tool, you can, um, uh, build kind of avatars of yourself and how you would, um, break down data or how you would approach a problem. And that can really speed up, um, the, the approach that you have to doing, um, any of your work. So I’m, uh, building an AI version of myself, uh, to be, to actually use within our software.
So you could ask me a question. You know, like, how would Kyle approach this? And what’s really crazy is we’re, we’re probably about 85% of the way there on it. The answers are good. Like, it, it, it’s, it even talks like me little bit like, which is really kinda weird. ’cause it’s, it’s, you know, it’s, it’s, it’s seen my videos, it’s gone through my transcripts.
It, it picks up and we all have our own cadence, how we speak, and, uh, maybe words that we use in, in a, in a unique way. And it’s starting to pick up how I actually talk, which is really, really crazy. And then, um. But uh, on top of that though, I have to tell it don’t go outside of your knowledge. Like, don’t guess.
’cause it really wants to guess. And so like, when, there are times when it’s not really sure and it can go through its knowledge and there’s not a good answer there. It’ll just make something up and then it’s wrong. You know? And that’s, that’s, that’s the problem that it has, is that it, it’ll go and it’ll try to just make something up.
It’ll guess at an answer, and the answer is often incorrect. And that’s more than anything. What we’re working on is, is trying to, um, uh. Have it be truthful and say it doesn’t know the answer. But then on top of that, we’ve got a version that would kind of be kind of client facing or, or user facing. But we’ve also have, have an internal facing one, especially for the marketing team.
We’ve given it our style guides. This is how we like to do our marketing. These, this is the, the tone that we like, you know, giving an examples. And then when we have a new feature, the marketing team actually goes in and has a conversation with it and says, okay, we need to put together five static ads and we need to do this.
And we also have, uh, created an AI version of, of me. Uh, this is actually not real me. This is, no, I’m kidding. This real. Um, but, uh, and then they can actually, because marketers, uh, you know, the marketing team doesn’t always understand the features the same way the product development team will, right now, I can explain it to them what it is.
And on top of that, I can help create the email copy. ’cause we’ve already got that inside of, of the tool. So there’s, there’s a lot of, you can kind of almost like customize the search or customize the data output. You’re going to do the search anyway, now you can run it through some filters and get it in the way that you want a lot faster and, and your productivity will go up.
Doug: Mm-hmm.
Hallucinations and Counting
Doug: It’s so interesting with the, um, hallucination issue because, I mean, obviously the, the big LLMs do this as well, and then you double check. You’re like, Hey, that doesn’t sound right. Are you sure about that? They’re like. Oh no, actually that was completely wrong. It’s the opposite thing. Yeah. And it seems like such a simple thing to not have it fuck up in that way.
And yeah, it’s like obviously you’re working on it and other people have observed this and you hear it in different industries, but it’s not solved yet, but you guys are getting closer to it.
Kyle Roof: Yeah. Yeah. It’s just a big thing of, of training it, like having a hard rule. Like don’t go outside of your knowledge.
Doug: Okay.
Kyle Roof: If you don’t know the answer, say you don’t know the answer. And it, it’s something that, that’s like built into it. ’cause it, it gets twitchy, right? It’s like, but I gotta, I gotta give it, I gotta give an answer. No, but it is like, if, if, if AI Kyle doesn’t know the answer, then AI Kyle doesn’t know the answer because real Kyle would be truthful and say he doesn’t know the answer.
And you know, that’s, we just have to tell it to do. Um. I was gonna say something though. Oh. Like we have the hardest time with our AI writer, where it’s like, initially we’re like, okay, use this term four times anywhere you want, and it’d be like, done. We’re like, okay, there are three in there. And they’re like, oh, you’re right.
Good catch. Let me rewrite that. Okay. Done. And it’s like, there’s five now. Like we told you four. And then you even ask it, can you please check your work real quick? And they’ll be like, good. They’re like, like, no, no, you’re not good. I see this many. And you point out and you’re like, oh, good catch. You’re right.
Doug: It’s, it’s so crazy ’cause like the one thing the computer should be good at, it’s fucking counting.
Kyle Roof: Counting.
Doug: And
Kyle Roof: it
Doug: boggles my mind and I’m not, it has gotten better, I think. But I, and I know some of the, um, you know, more recent releases, they’ll have sort of the thinking process where I, I think it does go back and double check itself and then iterate upon that.
But you still end up with those same issues again. Mind boggling. And I. It seems like the simplest thing to fix, but it’s
Kyle Roof: not one would thing. But like, I guess there was like, it was demonstrated that like, count the number of R’S in strawberry and it can’t do that because at in the tokenization of how it does the words, um, it’s not doing it in the same way that it’s not really counting is the problem.
That’s, that’s the biggest thing is it’s actually not counting.
Why AI Content Fails SEO
Kyle Roof: But, you know, another issue that we ran into, so I’ve been running a case study for the last year and some change is, um, getting a baseline on the LLMs. Like, what do we get without any prompting? And so we’ve been doing it just as a baseline with the most recent and relevant LLMs.
Um, give me a thousand word article on what to see in Rome. So there should be plenty of, of information out there, uh, within its corpus or within the knowledge that it has to provide that. And, um. In the last year and a half, they have not improved. And the thing that’s really nutty about it is the, the biggest problem is the LLM is bad at language.
It leaves out contextual terms, terms that give meaning to what you’re talking about is its weakest point. And it’s like the clue is in the name you’re large language model. And like, so like if, if you can’t even get language right, we can’t expect it to count. You know, like that’s, that’s the, that’s the thing.
But the, um. Some of the conclusions that we’ve drawn recently into that case study are that Google doesn’t have to do anything special to find AI content. And I don’t think Google particularly cares if it’s AI or not if it, if it adds to value to the conversation. But I think there’s a presumption that AI content does not ’cause It’s really just a facsimile of, of somebody else’s work.
And so all they have to do, and I think they did this at the end of last year. Um, was just turn up the dial a little bit on, um, contextual terms and the value of contextual terms within the ranking models. And they can sort out all the ai because what happens is that Google just doesn’t know what that content means.
They don’t, they don’t understand it. And so you’ve launched 400 pages of AI content and then your site gets hit. It’s not because it’s ai, it’s ’cause Google went and crawled and said, I really don’t know what this is. Where to put into the index. I don’t like the counts that it has. And then as a result, um, the site gets hit and then the AI content gets removed from the SERPs.
Really, it’s a elegant solution for them to take care of what they perceive as a problem and do it cheaply and with the, the tools they already have.
Doug: And I’m a little bit familiar with that case study. And one of the, I mean, one of the core issues, aside from the context, but this. They’re related. You ask for a thousand words and you don’t get nearly a thousand words.
Like what kind of results are you seeing
Kyle Roof: that’s gotten better? But the first time, um, only, I think we did like say nine LLMs and pick your top ones. We did it. Um, only three got over a thousand words. Like, for our purposes, again, this is, we’re looking at it through the lens of, of, of marketing, you know, digital marketing, SEO.
So this is for our own purposes, right? But, uh, for our purposes, going over a thousand words is no problem. Going under is problematic, you know, because we need, we need a certain, uh, word count. We probably told the client we need this, or, or we’re shooting in border. We realized, you know, this is the kind of content we need to put out that Google is going to like, and, um.
Uh, only three got there. They’re getting better at that now, for sure. That’s maybe the one area of improvement, but when it comes to the, the what we need for ranking purposes, they’re still terrible. My favorite one was, um, uh, a llama. Um, it literally was like going, going, going, and then it, this little blue line popped up at the bottom.
It was like, dude, I am tapped out. I cannot write about this anymore. And I was like, oh, sure. You know, surely this must be, you know, I really want, it was like 430 words. Like, it’s like I’m no expert on Rome, but I’m guessing you can write a little bit more than 430 words on what to see in Rome. Uh, yeah.
No kidding. You write, you can write that. You can write, you know, people have written entire essays, you know, on the Coliseum alone or the Vatican, or you like, yeah, there’s plenty, but it’s literally, there’s a thing at the bottom. It’s like, I honestly can’t write anymore about this. I, I have done all I can.
Doug: Yeah.
Kyle Roof: Fucked at the bottom. Good job. Good work.
Doug: Oh man.
so we’re talking about. AI being used to create content, how could people do a better job?
And you mentioned the, um, you know, the context and basically word count sort of being off, even though there are improvements. So what can people do to use AI to write their content and actually have, you know, something worthy to be published and help on the SEO side? Huh.
Kyle Roof: You know, you and I were, uh, chatting before, uh, we, we started recording.
It’s as though Google pinged me and they were like, we’re gonna do some major core updates. How can we help pop? Like, because it, it’s something that we we’re actually very proud of and something that we do with the tool is we, we don’t guide you towards keyword stuffing. You know, it’s using contextual terms, it’s using variations of your term appropriately.
’cause that does a lot of the heavy lifting and, and that’s what you need now. You don’t have to use my tool. That doesn’t hurt my feelings if, if you don’t. But you do need something that’s gonna give you counts. On the contextual terms, and it’s going to, um, uh, look for those terms that, that are needed, that support your, your main concept.
And if you’re not, you can’t do it by hand. Um, that’s, that’s one of the things for sure, and even really good prompting doesn’t get you there either. I know a lot of people are like, oh, this is really good. I prompted this really well. And then they show me the content and I’ll run it through Popin and it’ll score so poorly.
So it’s like it reads well and it gets the message across that you want. But for our purposes and what we need for ranking and for what Google to like the content, it just needs to be more, um, contextual term rich, you know, LSI, if you like that word. And, uh, it’s just not there. So you need to use something that is giving you those counts or you do run the risk of getting your site dinged.
Uh, and uh, I haven’t yet to see any of those sites really recover.
Doug: And that said, you know, I’m thinking, you know, we’re talking the AI side of someone’s, basically, let’s be generous and say they’re writing an outline and then they’re having, they’re not,
Kyle Roof: go on
Doug: super generous and then we’ll swing on the other side.
’cause I, I have a couple friends that just started Substack and I know they’re writing it themselves. ’cause it sounds like I’m sitting there chatting with them from that perspective, like. Going back to the old days when people used to write by hand, like, does that usually fit well from like if you ran that human written content through pop, does it often check out where it’s gonna do pretty well?
Or do you run into the same normal issues? No, you run
Kyle Roof: into the exact same issues. So, you know, I do some consulting on the side, and I would say in the last three months I’ve looked at 12 sites that were hit. And all of them have this problem that they’re, they’re just so light on contextual terms, but two of them were handwritten.
Like they, unless they were lying to me, but I don’t think they were lying to me, but they were like, I actually wrote this content. You know, like, this is me, or like my assistant who’s sitting right over here while I’m taking this call, wrote this content. It’s not AI at all, but it’s just getting caught up in the wash, you know, where.
It doesn’t. So again, it gets to point, Google doesn’t, can’t tell if somebody wrote this, you know, like that. You know, that, that somebody physically started with a pen and paper and then moved to, you know, their, their, their, you know, their Google doc. They, they, they don’t know that they can only do it by the counts of the terms of there.
So if you are, if you’re light in the way that you write, it’s gonna appear to be AI to, to Google. And it’s not that it appears to be ai, it’s just, it’s just content that doesn’t understand. It just doesn’t meet the mark of what they’re expecting for that type of content.
Doug: Okay. And I, I could imagine just because of how someone writes a particular style, maybe they wouldn’t use some of the, again, contextual LSIO, other words that are often around, ’cause that’s what we’re talking about.
It’s just
Kyle Roof: right
Doug: in pop. There’s a lot of data that has been analyzed for what else is ranking, and that is where those specific word counts for certain words or semantically related words.
Kyle Roof: That’s right.
Doug: That’s where it’s coming from, right?
Kyle Roof: Right. That’s right.
Contextual Terms Explained
Kyle Roof: So you’ve got your, for lack of a better term, an LSI type list, but then you’ve also got say A an NLP type list.
Um, we have two we can tap into Google’s N-L-P-A-P-I, and then we have our own, uh, NLP libraries that we tap into as well, to get that context, that meaning those terms on the page that, that explain to like, you know, the example I give all the time is if you’re talking about a kitchen, you could say words like, um, sink, you know, stove, refrigerator, and then you’re actually talking about the, uh, action.
An actual kitchen. You could talk about, you know. Uh, you could start talking about remodeling terms, you know, and then you’re talking about remodeling a kitchen. Or then you could start talking about like, family and love and, you know, and holidays, and now you’re talking about that the kitchen is the heart of the home in all three, all three of those.
You’re talking about a kitchen, but you’re talking about it in a different way. And it’s those contextual terms that allow the meaning or the understanding of what we’re actually, we’re not just talking about a physical kitchen here, we’re talking about. The kitchen is, is is the heart, you know, and it’s the, it’s the metaphorical heart of, of, of your family because that’s where most people gather.
And, um, and then, so it’s an important part of the family dynamic. Um, that’s a completely different conversation than a remodeling conversation, but you don’t know that unless you’ve got the right contextual terms in there. And so the idea is, Google’s looking at that article you wrote about kitchens and it just doesn’t even know what you’re talking about.
It’s just ’cause it, it can’t identify where you’re going with this and, and what the purpose of this content is.
Doug: Gotcha. And can you go over the difference between LSI and the NLP
Kyle Roof: and Sure.
Doug: What people need to be aware of? Yeah.
Kyle Roof: NLPI think it’s more, it’s, it’s better to think about as things, concepts, you know?
Um, and then the, uh, the LSI or the actual words that, that go on a page. So when you look at like Google’s N-L-P-A-P-I, so you can tap into it and Google will say like, okay, I think this is what you’re talking about. And then it’ll say like, you’re, according to us, we think you’re missing these words. And you might go look at the, at the text and the word is actually there, but that’s not good enough.
Just having the word on the page doesn’t match than semantics or what, what the NLP, the language processing that Google’s looking for, because it needs more contextual terms than to understand the concept. So NLP is more conceptual, uh. Con conceptual, physical things is kind of made, it’s, it’s a, now it’s a person, place or thing, an organization, but in concept form.
And then the LSI, the contextual terms are the ones that give the meaning that, that allow those semantic, uh, ideas to be brought forth within the, within the text.
Doug: And you used to be a lawyer, right?
Kyle Roof: I did,
Doug: yeah. I still am. Yeah. Okay. Fair enough. Um, the, well, and I was gonna say, lawyers, if people don’t know, it’s like.
The, all the words mean so much. So it’s like, it is funny, like going back to the specific, like this word means this in this context. So like it does go back to
Kyle Roof: Oh yeah.
Doug: That kind of
Kyle Roof: thing. No, my, my, my, my training was perfect for this.
Doug: Yeah.
Kyle Roof: Like my, my, my training got me right to, uh, understanding language in a clear and precise way.
Uh, and then also the, the, the value of meaning. Like, there’s interesting words, mean very specific things in very specific situations, uh, unless you modify them with other very specific words. And, uh, yeah, that’s been great for me as a, as an SEO.
Doug: Tricky. Yeah, I was always bad at, uh, most of the stuff, most of the things that you just mentioned, so.
SEO Growth Opportunities Now
Doug: Alright, so as we’re, um, thinking about. Growth areas, growth possibilities. Um, a lot of the audience, either they were like previous, um, SEOs or current SEOs, but a lot of folks were deep into like affiliate sites and content websites. Mm-hmm. Kind of moved on, or those sites like dropped away or they stopped working on it, but they still have the core skills.
Some people are looking for side hustles or things that they could kick up. Because of the skills they already have. So do you see any specific growth areas for SEOs or marketers where there’s kind of a good opportunity for them to get rolling?
Kyle Roof: Well, e-comm and and local business have not been touched in years by any recent Google updates.
So lead gen, which is a form of affiliate marketing, is still, I think, alive and well. I was never in the lead gen game. I never really got it. Um, but I know a lot of people that are doing it and doing very, very, very well. Um, and there’s a lot of good, um. Uh, CRMs out there now too, that sink in very well with websites that actually maybe make lead gen a little more possible than it was, uh, before, like a go high level, for example.
I think that they’re just built for that, you know, and then, and those kind of tools are, are, are really, really good. Um, uh, e-comm will always be there,
but I’ve seen more and more affiliate sites come back in in the last year or so.
I think Google killed off who they wanted to kill off a few years ago. And I, and I’ve seen some affiliate sites, uh, come back in.
KGR and AI Search Insertion
Kyle Roof: Now I was dumping earlier, earlier, I was dumping on AI search, but if you see a niche. Where that there is an audience, I think it is still quite easy to get insertion into AI search.
So there would be opportunity there. And you know what’s really working well for, um, uh, insertion is keyword, golden ratio. It is working extremely well because the, the way that ai, uh, likes to do searches, it does query fan out. So it takes a main concept and then it breaks it down into. Smaller pieces, related searches essentially, so they can understand what you’re driving at and it can start to answer all the questions.
And that’s exactly what KGR gets you into is getting into those terms. So you have your, like your main page concept and then I write supporting pages and KGR is a big, uh, filter for me when I’m doing that. And turns out doing that will actually get you inserted into the LLMs as well.
Doug: Interesting. I haven’t, um, like I said, I’ve not been doing like SEO, like boots on the ground kind of stuff.
And every now and then someone will mention they saw you at a talk and they’re like, oh, Kyle’s still mentioning KGR. And I’m like, no, shit. That’s crazy. I should do a site just to test it out.
Kyle Roof: It’s working so much better now. Like, like it always, like, I think it always worked. I think part of the reason that it didn’t work is it was missing one step.
And that’s kind of like how strong your site is.
Doug: Yeah.
Kyle Roof: You know, the, that was kinda the missing step. And so I combine it with, um, avalanche theory. Um, and, and k and I think we, the last time we were on, I think I spoke about this, it was working great. Then it’s even working better now. Interesting. Like it is working so well right now.
I’m, I’m actually, I’m crushing it
Doug: really? I, yeah. I was like, oh, I, I should find just something small to, um, you know, do a little experiment, get it, get people talking about it and going back to the, like Google killed off the affiliate folks that they, they were wanting to, I think. Well, it, it was kind of my world and a lot of my peers where we were creating content, encouraging people to start sites, but now that no one’s talking about it, there’s just fewer folks out there.
It’s just the hardcore folks that were like interested in it and they wanted to keep going with it. Or maybe they had a big site and they’re only recourse is to like do what they can and pivot and they’re making it work. But yeah, it’s kind of interesting now that there’s not a buzz around it. It’s like, ah, it’s not a big deal.
There’s not new folks jumping in thinking they can, you know, saturate the whole sort of industry. So,
yeah. But you can, you, you definitely,
you
definitely can.
Very cool.
Wrap Up and Where to Follow
Doug: Alright, I think we covered most of the stuff that Yeah. I had on my list. Anything else on your mind or things that people should consider, um, thinking about this, um, SEO and an LLM world?
Kyle Roof: No, I mean, what we went. You can’t do it by hand. And, and even if, even if you think you can, you can run into problems that you, you don’t even know you have. So the biggest thing is you need something that is doing counts for you and, and, and you are, you are benchmarking your content on, on the, on the amount of contextual terms you’re using, I think is mission critical.
After that, do whatever you wanna do and I think you can, if you wanna get into that side hustle, I think, uh, affiliate side hustle is probably on its way back.
Doug: Very cool. Alright, Kyle, if people wanna learn more and follow along, where should they go?
Kyle Roof: Yeah, just head over to to Page Optimizer Pro. There’s actually, I have a free course on there, academy Page Optimizer Pro, and if you wanna learn about, about what I’m doing in seo o that’s, if you like the sound of this voice, you can get another three hours of it.
Doug: Very cool. And then, yeah, I was gonna say, and there, there may be, uh, an AI version, so you could just keep asking, you could ask it unrelated questions and just hear like, Hey Kyle, tell me what you had. For breakfast, and then there’s, there’s not an A SMR version, but maybe, maybe someday. I don’t
Kyle Roof: know. That would be hilarious.
Like smell of vision from, uh, from Willy Wonka.
Doug: All right, man. Good catching up and uh, we’ll talk soon.
Kyle Roof: Thanks so much. Appreciate it.