Does Google Penalize AI Content?

Transcript: Does Google Penalize AI Content?

This transcript of Does Google Penalize AI Content? comes from the video's captions, lightly cleaned. Click a time and the set jumps to that moment.

Critique of AI content by Neil Patel

0:00 Welcome back folks to the David Quaid SEO show. This is episode six. In this show, we got a really, really great topic. Why Neil Patel is totally wrong about using AI content. Neil's done some videos lately that we think he's got some basic things wrong. We're going to give you a load of info. I've got my great partner and SEO expert, the king of SEO. I've got David with me. So, David, would you like to give
Our audience a quick intro before we go into the meat and potatoes of this show?
Thanks, Jonathan. Yeah, I've been doing SEO since I guess the first time I tried it was 1995. But I've owned primary position since 2002. And, I love all things SEO. I care more about SEO than money. I'm I'm in it for the engineering like most propeller heads.
I love it. So what we're going to be doing, like I say, Neil produced a video where basically he was saying that if you use AI content without it heavily human edited, you are going to have some problems. And we're going to go through the some of the key points of the video. Some other people kind of pled in with similar content. We're going to discuss it all. But before we go into the main part of the show, we've got a couple messages from our sponsors.

Sponsor messages

1:37 We will be back in a few moments, folks. WP Tonic is sponsored by Ninja Forms. And let's talk about one of their powerful tools, conditional logic. Conditional logic is the clearest example. Every rule for the whole form lives in one panel instead of being buried field by field on client work. Build speed and low maintenance are part of your margin. A form that is fast to configure and that a client can run afterward without coming back means less friction during the build and less support load after the handoffs. Ninja Forms, powerful and genuinely easy to work with without being dumbed down. To learn more, go to ninjaforms.com. WP Tonic is sponsored by Jetpack. Managing WordPress sites shouldn't mean stitching together a pile of disconnected plugins. Jetpack brings security, performance, publishing, growth, and AI tools together in one WordPress native toolkit. Use what you need. Backup, scan, boost, social, video press, stats, and more. All with modernized dashboards built into WP Admin for your site or your client's sites. See what Jetpack can do.

2:58 Jetpack.com.

Debunking general AI myths in SEO

3:00 We're coming back. So, in the show notes, David, that I sent to you, I think. I don't know if you want to do intro about what your general view or we should go into the key points that I listed. What do you think we should do?
I think we should talk about the my general view.
Yeah. Go on. Off you go.
So my general view is that Google remains content agnostic. The last white paper that they produced on Wednesday this week says that engines including AI engines cannot be preoccupied with individual document. The only content we have from Google on it was from the DOJ's book of evidence from the DOJ versus Google court case that they won that Google explains to their executives they don't understand content and Neil pulled together a video and I've been trying to give him benefit of the doubt. I don't want to be I don't want to come across as someone who's jealous. I think he's been very successful. I think we were talking about him a month ago before a podcast and I said how much I respected admired him. He's done very well as a businessman. He's done very well as an SEO. When you search his name, it comes up everywhere. You know, he's way superior to me. But this last video just really the trains the wheels came off the train for me in a big way. First of all, I think he conflated watermarking and identification through heuristics. And he his whole video is like Google is watermarking your content. Which is it's just it's just not really what's happening because not a lot of people first of all will use Google AI to write their content.

4:49 And so I think he's conflated or he's confused a whole lot of terms and mush them together. I actually think like a lot of people said in the comments on the video that he may have used AI to write the script.
I don't know. Well, my take on this, David, is that this is just a general trend and you just find a lot of [snorts] SEO experts, people that are big influencers, they are saying the same basically. And I used to believe it quite a bit myself. I don't believe it now. So basically I think what he's saying is this. Maybe you can explain what this SynthID from Google DeepMind is because I think he was really you know you just touched upon it watermarking but Google's got their own and I think a lot of people they don't directly say it but they hint at it indirectly that Google's got a technology that they actually can look at the content and know that it's AI generated and they will penalize you for using AI content and it seems to be around this SynthID.

6:26 So maybe you can explain what SynthID is.
So I think this is where the conflation is happening, right? I think and also you're you're leaning into confirmation bias. In other words, you you think you've seen sites penalized for AI content. So when someone says it's happening, you've got a predisposition to believing it, right? And so thanks for demonstrating that to the audience because I think a lot of people are afraid of that, too. I think there's a lot of people out there who've been using AI content. They see it not performing and they're there's also a lot of things happening in Google's world where it's just playing with its standard mechanics like tightening authority. And so if it tightens authority and you lose clicks, you get the staggered waterfall effect. Joy Hawkins was saying during the week on one of her ex- livives that it's been a long time since she's seen a manual penalty handed out or even a penalty. It's a long time since I've seen a penalty handed out too, but I've seen a lot of scaled content penalties and maybe about 30 or 40 this week in the last 5 days. They are a cliff wall, right? Your content goes up and up and up once it passes over the heuristic zone. Like in other words, Google's decided it's got enough there's enough data and you fall into that. It's an instant straight down. It goes from here to here. There's no staggered down, right? A staggered down means that you're losing traffic to rank position or you're losing traffic to AIO or other people are pushing you out. And as you lose clicks, you're losing authority and your all the pages that are interlin start to fall together.

7:57 What Google and I guess this is preceded by Anthropic are saying is that they can watermark their content not by putting in a visible watermark into the text like people have been doing with images for a long time, but the way they explode, right? So if you take like I think I've described LLM content to you before as it's a bunch of concepts nouns and adjectives that are tokenized that live in a mathematical world. So almost like every idea is a polygonic shape and if you want 50 words they can explode that to 50 words. If you need a thesis and it's going to be 50 pages with 500 words a page they can explode it to that. And all they're going to do is drill down between all of the topics and add more words. Right? And perhaps add more word cit or what however they're designed by looking at how the English content is derived and by re-imploing it converting it back tokenizing it back into LLM language if the concepts all land in the same spot however it's tokenized or however it looks like an LLM can tell that it was produced by another LLM it's very unlikely that Anthropic shared that process with anyone else it might be very possible that if you to retokenize English content or any human language content back into tokenized language, you could see a pattern. So there's a big difference between can and does. And this has been one of the things that has been haunting Google from day one is people have been fabricating can and does. So people have been saying Google's got a patent very this is very recent and this is actually something that Neil and other SEOs have leaned in heavily on is saying that Google can tell if your content doesn't add anything right if your content doesn't add anything it can see first of all in order to do that for every page and every concept it would have to reiterate across all content it would have to run a process like that it can't just ingest every page and go this adds more or less than the last page I ingested because they're not done in order. The second part of that problem is that people have been doing things like when Google became a registrar, people were inferring that age was important to a domain and Google could see the age because they're a registrar and then Google stopped being a registrar. Google became a registrar because they wanted to have their own.

10:22 Google domain name down the road and people have inferred that Google can see when you last date stamped your page. And this constant inferring of Google can and Google does is has always been incredibly dangerous to us. And I whenever people talk about SEO and they keep saying it changes so much and I keep thinking the it has is because I'm not following every potential possibility that if Google can do this, therefore it does. Right? That's the first problem with this. The second part is that while a lot of people have grown to distrust Google or say that they lie and I'm on the fence about it, they do have a document that says we don't care about AI content. And I've seen enough AI content deployed to know that they don't. And that content isn't in any way dewatermarked or done in a special way to avoid that. Yeah, it's edited because some of it's, you know, for sites that it needs human editing. Some of it. But I don't think that they care. At the same time, if you had a single sentence from an LLM, that would be very, very hard to tell that was written by an LLM because it's just one sentence. It's not going to carry enough if you retokenize it back. So, if Google was to do this, it would turn into a horrible game of whack-a-mole. So I think Neil Patel honestly is doing what I accuse the geo agencies and geo specialists of doing is just playing into that confirmation bias and feeding people what they want to hear. Like saying for example like one of the worst comments that he made was that if you put an author bio, Google will start to rank your page higher. U it's just more rubbish, right? Like why would Google rank your content higher because you have an author bio? Authors pe special people that don't tell lies. Maybe the TSA should allow people who have an author bio to use TSA pre, right? If they're that special. It just it's just complete nonsense, right? But I think that marketers and content writers you know there's there's a person who I don't want to name that very common in talking about eat and he's a very big proponent that Google likes to see where recognizes the author ID but that person doesn't rank for anything everything right. I don't think Google sees an author ID and goes like, "Oh, well, this is written by David or this is written by Jason or this is written by this guy. Must be good. We must rank it." That just doesn't make sense. It's very different to I recognize this author, so if someone searches their name, I can show the latest articles written by them. That's surface data.

12:53 That's completely different. But the video that Neil put together was one full of AI typos. It was full of the very same advice you get the poisoned well advice which is like eat is critical to SEO and eat
Just so we you know I'm going to play I'm going to play the devil's advocate here right
But he but Neil's got evidence though David
734 pages
Yeah 744 client articles
On client websites and from that he's got clear he says he's got clear evidence the whole market just vote half the content is now written by a but only gets 20% of the reward Neil says he says that the content that is heavily edited gets the majority of the traffic and the clicks he and through this 740 articles He sees a clear pattern. So what do you say about that, David?

14:06 So it reminds me of the parable of the fish where,200 people saw Jesus feed all these people with the fish, but there's no list of names of the people who did it, right? It's just a claim. One of the things is why I think so many experts lean on sort of making claims is because they know that people will believe it without question. And that's where we get into dangerous territory. It's when we let our guard down, right? We're not naturally good at asking questions. I get in trouble all the time for going, I don't believe that. Explain it to me. And people say like, oh, you're you're being argumentative or you're being obtuse. I'm like, no. If you've done this research and you've got 700 pages, where are they? I can I can show you so many reasons that make sense why 700 pages dropped in traffic or dropped in visibility or whatever it is measuring because that's a bit vague that I would see all the time without it being AI or human written content. And if you've gone to this trouble of doing this test, then why isn't it published? And what one of the videos I made early on this channel was with Jake Hundley who did
An experiment.

15:20 You spoke to him as well.
Yeah. Great point.
It took him very super guy, great guy, just such a nice guy to talk to and he's been a great friend
And he came to Edward and me early on when he was talking about the schema test and it took him months to set up. It was involved 39 companies. And when you when you see these when you see people just nonchalantly talking about how we tested a million pages or 500 pages or 50,000 pages, you begin to realize that, you know, they're doing this every week, right? You see these posts coming out all the time. It took a to help Jake. He went to a number of SEOs who tore apart his methodology and said, "This is this is just doesn't make sense. You've no control group, nothing." These are important things because otherwise you can have you can just stand up and say, "Well, I did this and this happened and therefore I'm great." And I actually think that's what happens 90% of the time. If you've got the evidence, then show it. If you don't, then it doesn't matter. Anecdotal evidence is the worst kind of evidence because you can make up and say anything because nobody can prove you wrong. I think it's so bad that it's almost a danger in itself that if you all you have is anecdotal evidence or just a summary of evidence, then go away with it because if you're not going to let anyone touch it, then it's it it presents a hazard to the SEO community because nobody can look at it. If you can't be honest about it,
And just say, "Well, the data is over here and the data told me this and I'm going to tell you what the data told me.

16:54 I'm sorry, but I don't believe you. And you might have a good name and a good reputation, and I'm not trying to sell you that. I believe you should have it, but I believe you should let other people test that data. So to recap saying that there's a clear pattern there's so many possible different factors in that assump assumption unless you provide how the test was done. The specifics of the websites give a lot more detail really it becomes irrelevant because there could be so many other factors that are
That are affecting the data basically. Is that the gist of it David?
Yeah I mean you a blogger yourself we I think nearly everyone who watches this channel has a blog or worked on a blog at some point.
Yeah.
Every post you post doesn't get the same number of views. It doesn't they don't all rank. In fact, if you look at Google Search Console and you look at your top 10 pages, the falloff between the number one page and the top the bottom page could be 90%. And that's just within the top 10 pages. So now if you're throwing 700 pages out across all these domain names, where's the control? What if the pages started to rank and stopped ranking? An average view from that of losing 20% is a meaningless number. I I could show that. I mean, that could happen with content that was written by Elvis Presley. That has more to do with topical authority. What topics did you pick? That it started to rank and the clickthrough rate nose dived or it started to rank or it didn't even rank, right? Or it ranked for five keywords and stopped ranking. Those are much more plausible things that happen in reality.

18:42 And just because I think the problem is that just because someone famous said it but then didn't provide the data I don't just automatically believe it and given all of the other mistakes in what he said like for example telling people who like to believe that having an author bio makes Google rank your content higher and so if five people put up their hands and say yeah I added another bio my content went up then that makes it true and first of all it's Google have said that's rubbish. But there's so many people who've tested it and that's the one of the other problems in critical thinking and our lack of critical thinking and our lack of foundational really functional literacy is that people never test the corollary. In other words, do the opposite. And most people can do the opposite, right? Like if you put an author bio on and your pages went up, take it off. I mean, if it if it drops, you can put them back, right? We don't we don't live in a world where Google's like keeping a checklist and an audit of every time you change a blog post. And people don't do it and they say they're afraid to do it, but I don't think they want to do it. And I think if more people tested, we would have less of this sort of like blind sheep effect where people are just following these sort of like I call them ivory towers for lack of a better word, just throwing out breadcrumbs of knowledge. And people saying thanks for doing the testing. Why not go test it yourself?
Right, let's move on and I will be the devil's advocate. Right.

20:16 Sounds good to me.
Good. Cuz basically Neil says that Google is really looking at who said this and do we trust them? And then he goes on about these 20 minute human layer that if you add 20 minutes or hour to article it's going to make it much more trustworthy to Google and that and that's you know Google needs proof that a hum was involved in the writing of the article add something that AI couldn't know cut the filler Titan in every claim. Freshness is important. I think that you might agree with freshness. Not at all. All right. And Google needs a real author, which you've just talked about. So, he says that adding these elements to the article that proves that a human touched it. And I think you've you've been consistent in saying that Google doesn't really care. Basically, it really Can can we see can we see the steps of delusion along this way? Right? That you that first of all, you've now given up. Never before in your life have you ever been told that an author is somehow infallible, right? But that but people are willing to make this massive leap of faith that if Google sees it's a human author that it believes it. I mean, people have been faking authors for eons, right? But but let's say it's true. Does it mean that I can't make a mistake? Does it mean that I can't manufacture terms? Are we saying that authors are infallible? That because it's a real author, Google believes it.

22:06 Like it's just nonsense. It's just absolute and and again, it plays into the EAT card that again I think people are happy to hear. We keep forgetting every time I enter a conversation in Eid, we keep forgetting that Eid applies to YMYL and that YMYL is incredibly narrow. And it doesn't include Nike and it doesn't include doctors or chiropractors. I'm sorry. I' I've from what I've seen of eat of YM, it's very very restricted. Whereas
We're gonna we're gonna talk about that probably in the second half of the show, but you know you know we've we've had Lily Ray with her study of 220 websites and I'm not sure you know Neil says that she supports what he is seeing as well. So it's very difficult for people when the they have a consistent message that's been delivered to them, isn't it, David?
Well, so again, let's test it. First of all, Neil says that her study supports what he says, right? It's not Lily Ray saying it. And so I'm just going to leave Lily out the picture because
Right. And and if she if she comes back and says she does support it, that's fine as well, right?
But first of all, again, 240 pages control group. So I actually went to have a look at the article that he linked to or that he spoke about and in it Lily writes that
The tools that test it that there are a huge amount of gray areas that where the tools would be inaccurate or where it wouldn't actually support it. And we can look it up during the next break and we can read it out. But when I started reading I was like I'm not this isn't but also what she was talking about wasn't necessarily all of the things he said right? He was just talking about human edited content versus non-human edited content. There is no way that an AI and this is his myths are just are just built on top of other myths and they're very similar to the myths you get if you just casually ask any LLM how to do SEO, right? They just regurgitate you should write good content, you need to pay attention to eat because it's algorithmically important to Google, which it's not, right? And because there's so much content out there that supports these myths, Google's two documents or three documents on it are statistically in insignificant and so they fall away. And so the LLMs can't actually decide what's true and what's not. And so that I think Neil's entire argument is written by an LLM because it sounds like the same nonsense that I get from an LLM. And I think if you look at, you know, the other state, if you can read that sentence out again that he said, it was that you can add bullet add points or to add data that an AI can't get from somewhere else. Anyone can make up data points, right? I mean, people make up data points in Reddit all the time. It gets indexed all the time.

25:10 I made up fun comments fun data points and put them into a blog post and guess what? It's ranking for the king of AEO, right? Just because other pages mention a stat doesn't mean it's true. Right? I'm I'm very sure you could do an exercise where you could determine that the capital of Paris is of France is Paris because it's out there in so many documents. But what happens when we get into opinions? What happens when we get into observations? What happens when we get into beliefs? There's no facts. You can't fact check that through consensus. And consensus
Yeah. I think in the second half we're going to what you think Google really is looking for and what practically people can do to improve their SEO. But I'll just put this another SEO influencer that I like I like some of his content and I listen quite a bit is Cale Caleb Yu Yuku probably butchering his second name. But he kind of specializes a bit more in local SEO, but he did a video that had elements of Neil Patel's in my opinion video where he was talking about he used the term template quite a lot and I've noticed a lot of other people talking about SEO on YouTube They talk about template and he was saying if you use the kind of same template produ to produce like 20 30 pages different areas of a city so you're covering the whole city a bit of local which you don't you don't deal with but he was talking about if Google senses you're using a template it's and then he talks he goes into EAT eats quite a lot. And that Google's looking for expertise, experience, authority, and test and trust.

27:34 And in the end it had a lot to me it had a his video had a lot of elements of Neil Patel's video that Google makes a judgment call that it and it's looking at this eat and I think I think you've explained before that this document was and I might be totally wrong here and I'm sure you're going to correct me, but I think you've said in previous interviews, previous podcasts that you've been on that this document was developed for human assessors and they for actual real humans when a manual penalty was done or they were looking at a website for various factors and a human was looking at it. But I the reason why I added this to our show notes, David, cuz it's all kind of one big pattern to me that really confuses people.
Yeah. And it like I said, it's something that people want to believe in. Someone blocked me on Reddit because I was trying to point this out and which is like, you know, again, why is your why is your ego involved in something that's fact-based, right?

28:59 But let's just break E down very quickly. Eat is applied to queries that are YMYL, right? Not to content, not to whole sites, right? This idea of sitewide categorization would require human editors, but this is park that eat is what Google thinks people look for in YMYL topics. It doesn't apply to everyday generic sites. That's fact one that people just and SEOs who are listening to this keep forgetting it, right? I never see YMYL when people are talking about eat. Also, I see people putting law sites under YMYL. And I do I do now have a lot of experience in local SEO. I've been getting back into it. And I' I've recently edited 7,000 pages across 10 sites. And I'm feeling quite confident about it again. But let's just get back to eat. Eat is what Google thinks that people who are about to do something that's lifechanging that presents an something very different to their money.
So I'm just this is this thing money in your life.
Your money your life. Yeah.
Yes.
So for example and James Dooley was talking about the only site that he seen which was a betting site which didn't have contact information. In other words, there was no recourse. So if you if you went and placed a bet or you put money into the site and you lost it, there was no way for you to get the money back. And the minute you put the address on, the site was allowed back.

30:26 And he said that you could tell by looking at the Google refer. So basically all the human eat reviewers were logging into a portal. They visited the site and if they then noticed that required information was missing, they would put a flag on your domain and then a Google quality engineer would penalize it until it was rectified. Right? Much like they'll do with Google ads if you break a policy in Google ads. I've checked hundreds of domains to see if that referral was mentioned in the last three years and it wasn't in any of them except one which was a co vaccine technology site and that was a site where we took content and moved it to a subdomain and that page would not show up in Google anymore. It would show up on the parent domain because the parent domain had specific links from the CDC and another site NIH I think. Otherwise once we put on a different domain and I tried the same content in a couple of my other sites and if I searched for the brand name the page would show. If I searched up for the I can't remember the co protein that we looked up it blanked it. If I added the brand and the co protein it would show up again and then I realized that's what YMYL looks like. Whereas I think a lot of people use Y and Wild to make it look like they're doing something hardcore or something that's very difficult to do or it's harder to make it rank. Eat is not something that's algorithmically driven.

31:53 What they're saying is that in some cases people may look at it and here's the worst conflation. A lot of this is being pushed by copywriters, by authors, by people who do writing for a living. They're not people who have the firstirhand experience or expertise in a subject. They're typically writing that on behalf of other companies. Right? So, so they don't have experience or expertise. And Google can't see trust. Just because somebody makes a claim about experience or expertise doesn't mean it's trustworthy. And so,
To be honest, it I I actually had like a little bit of a spat with two people on this on Bluesky, John. I was like, I I hate that they've left it so vague and that they almost seem to relish in the vagueness of eat, right? And it's it's really been blown out of proportion because I think people are trying to find something objective and I can all I can tell you is if you're worried about eat you're so lost. It applies to so few things and it applies to the humans reading it not to Google. Google sees YMYL pages the way they perform and the way people bounce as a way of testing whether there's eat in it. But thinking that putting an author bio or adding claims into your content is doing eat and then saying your PageRank higher from it is just absolute nonsense.

33:13 Right, let's just recap and then we go for our middle break and then in the second half we talk about what you know about what's happened in the last couple years, why it's happened and what is Google really looking for. Which I think is be helpful. But just to recap, your basic position is that Google really doesn't really care what how you produce your content because it basically it system really monitors to some extent what happens afterwards. It hasn't got a system that monitors what quality is because it that's impossible to judge. And when you really when you really think about it, you're totally right. How could they make a large you know this is a whole I just want to put this to you before we go for our break is this is the total contradiction of the industry at the present moment is that it's saying that this is enormous sophisticated AIdriven monitoring system that knows about quality but in the same breath you get a lot of the same people pushing through listicles on large learning models that can that you know use hacks that were used 10 12 years ago that can actually move large learning models to recommend clients articles or their services and products. And there's a total contradiction there because if Google's using the same technology, large learning models to build this enormously sophisticated system that can tell quality, but in the same breath, you can manipulate a large learning model by a simple listical. That's a total contradiction, isn't it?

35:18 It is. And Google put together a white paper. And it's amazing that all of this junk we're reading about LLMs and their capability is being fabricated, right? Because OpenAI isn't talking about it. Anthropic isn't talking about it. Perplexity isn't talking about it. Google put out a piece saying how difficult it is for LLMs and that LLM aren't going to be the future of AGI and how they're not expert re search tools. And it's a very short document. I shared it on LinkedIn last week. A lot of people replied to it and found it interesting. There was an interesting single line in it that said we cannot tell if a single document is good. We can only look at the totality of documents or a set of documents that fit in a mathematical set of heuristics. Right? In other words, a set of patterns that we outline, but we don't know if the individual documents in there are good or not. In other words, Google is disavowing any ability. This ability that's being ascribed to Google that they have LLMs that can tell if content is good or accurate or not is complete nonsense. And one of the things I think we're terrible at as humans is is deciphering between subject subjective content and objective content. It was actually Neil deGrasse Tyson who helped me think about this in a quote that I read that he said Google is effectively a bias confirmation engine.

36:34 But basically, it's not just about quality or facts. And everyone, and I can see it in the comments already, people are trying to boil everything down into like encyclopedia content. But if you're writing a blog article about who's the best CRM, that's not an objective point of view to you. If a CRM can't do X, Y, or Z, it may be an objective failure to you, but that's subjective. Most of the content that people put out there is marketing content, opinion, observation, experience, or strategic thinking, right? Like, here's how I think I could fix a sales organization, right? Think about the content that consultants put out there. This isn't objective facts. This isn't something Google can test. Unless these LLMs are actually out there building corporations of 50,000 people and testing the tools and then testing if one corporation is better than another, which they're not, then I'm sorry. I don't see how they could be doing. It's just not possible. I think people are dreaming because they just want to oversimplify everything, which is a perfectly normal human trait.
Yeah, I think we're going to go for our break, folks. When we come back, like I

Sponsor message

37:42 Say, we're going to be looking at be asking David what really Google is really looking for and what are the real factors that have affected people in the last couple years. Should be a great discussion. We will be back in a few moments, folks.
Updating plugins and themes shouldn't feel risky. That's why over 300,000 WordPress [music] users trust WP Rollback. With just a click, you can safely [music] roll back any plugin or theme. No manual file uploads and no downtime. And with WP Rollback Pro, you can download [music] premium plugins and themes. Tools like Elementor, Gravity Forms, and more. Track every roll back you do with detailed activity logs. Protect your workflow and take the stress out of [music] updates. Learn more at wprollback.com.

The mechanics of SEO and Google

38:38 Coming back. Want to point out if you're joining us live, we do this around 12 noon Eastern Standard Time. You find us on YouTube, you find us on LinkedIn. And you got any questions for David, we will at the end of the show, answer those questions. So, join us live. Now, David, I think the first show was fantastic. I think we killed a few dragons.
I think you got I think you got your armor on,
You know, St. George against the dragon and I think you're gonna get a bit of flack for this, but I think it was necessary. Now, over the last
Over the last couple of years, I'm going to let's remove AI overviews because that's a whole that's a whole other area of what why Google's done that and the effects of AI overviews. Because I do see that as a in a way a totally different bucket. But before AI overviews increased their effect as Google introduced them wider and wider, I think a lot of people saw a reduction in traffic, saw a reduction a lot of people saw reductions and this was around Google's high in authority. I don't know how you would word it. How do you see what Google has done over the last year, 24 months?

40:28 What things have they changed and why do you think they changed them? David,
I think that Google has changed proximity of similarity between words, right? Like their semantic engine. I think that Google have changed.
Can you explain what you mean by that bit first?
Sure. So in the original PageRank Google described a system or Larry and Sergey described a system based on the similarity between two pages would allow the votes from one page to pass to the other page and people have misread that to Helen back as wrong right like people have invented sitewide authority like your site is in some site category. So this site is in psychology and this site is in carpentry and this site there's no such thing there. You cannot put websites in a single category. And categories themselves are human are subject to human definition. Right? If you go to
If if I'm working for a client that's in cyber security but they're in dark web monitoring they're not going people that are in say net sec ops or dev secc ops aren't going to see them in the same category. And these categories keep moving right every six months these category definitions change.
[snorts]
So the first thing we have to understand is what Google understands by related and relevant. And it just comes down to simple keywords, right? Like train holiday USA has more in common with train holiday Canada than car factory USA has in common. Even though they're both about the same country, right? Mhm.

42:01 But because there's two words similar between the USA and Canada term, it's more relevant than car factory USA cuz car factory only shares the word USA with train holiday USA. And that's just as simple as it is and it's just a page level system. And so how Google allows authority to flow where trust authority is coming from like the originating source or the nearest seed be it the New York Times or CNN or whatever we're not 100% sure common crawl does the same thing they're an alternative search engine that a lot of LLMs are using as Google blocks them out if people are thinking that Google's going to interfere with something like block
Can I interrupt just for a second so because I really think this is important and it's linked to the first half of the show. So what you're fundamentally or I think what you're fundamentally saying is that Google isn't a quality engine as a lot of people were we were talking about in the making a using some magic to make a judgment about the quality of an article. What it is a relationship engine. Themselves call it a reance engine. They call it but I've just used the relationship the relationship of different phrases
So it can make some sense and some judgment about what should be shown am I am I on the right track
100% it's it's a relevance engine a relationship engine yeah
Not a quality engine
Nothing can be a quality engine
Well that's that's what I think a lot of people getting on, aren't they? For various reasons,
Even the even the people I think that I speak to that believe that Google isn't a quality engine today or yesterday believe it could be one tomorrow, which I find just as heartbreaking. But it's not a quality engine because quality
Is subjective.

44:10 Subjective.
I joke about golf, but because I always say there's nothing there is no article you can show me about golf, even if it's about a golf course being torn up to build a pink parking lot. Cuz I don't drive either. So I don't care for parking lots. But the thing is that whatever you think is a priority to you, like a sense of humor, someone else will put another priority on. And we don't even understand why we have these priorities. We just do, right? And our priorities in life also change. And so one way to actually demonstrate the sort of transit nature of it is if you say the let's say you you remember the first time you read about something that you're now incredibly passionate about. It could be WordPress, it could be SEO, it could be hiking, trains, whatever it is, right? If you go back and read that very first document, you'll probably be disappointed. One because maybe it doesn't necessarily convey the same or invoke the same emotions in other people, but also because it's very basic, right? And so the very first page you read about how to do SEO, when you go back and look at it after 10 years and read it, you'll be like horrified and go like, "Wow, this is so basic." And actually, you know what? 50% of it's wrong. However, the first time you read it, it was the most amazing document. It was the best thing on that topic. And maybe it still is.

45:33 Maybe there are but along the way documents that were amazing will be less amazing or if you no longer even like the topic right because you had a bad experience or something else happened that document may now be the worst most horrible document you could read. So this idea that there's this sort of like fixed objective quality standard for the human race is isn't there except for you know machine generated garbage which was the predecessor to LLMs right so believe it or not long time ago 20 years ago there were people that were just producing words and filling pages with words and they were just garbage you couldn't read them it's horrible to read it just it's like you the can apple phone button right it's just it's just a collection of words and that Google can detect it's very good at detecting but that's about it and and I don't think there's anything wrong with that and Google have never claimed that right it's a claim that's been foisted upon them by people who want to push that opinion it doesn't exist anywhere there's nowhere in Google that says this is a quality standard the eat guide is not a guide it's not Google do not say use it they say this is a guide that because our systems can't do subjective decisions we give it to people to make it pretend what they're actually doing is testing their subjective viewpoints and so yeah cosine similarity may be one way of looking at relevance engines
So as I said I you use a different term but I said relationship engine so we got that so what so moved on so it's a rel I use this term relationship engine but what changed a couple over the last
Last few Mhm.

47:16 Yes.
I think so there's there's two parts to it. One is let's call it back links or nearest seed linking or distance or vectors or whatever you want to call it. That's your primary entry into an index. Right? So if you're starting as a website, the reason you don't go into the top of the website is because the index is because there's other websites with more authority than you. The second part is Google's phenomenal and almost anal capability of testing how people look at content and whether or not they have to go back and search for content again. Were they was their intent satisfied as some people I think oversimplified.
Hopefully the audience isn't going to get angry because I got told off a couple weeks of interrupting you. But I'm doing it not I'm doing it
To be this intermediate SEO person that's that's talking to a true expert. And I'm being the voice of that intermediate person. I think I'm intermediate. I think you would. Maybe not. I'm asking. But Oh, lost my friend there. But let's get let's get back to it. Jonathan, I think what Yeah, I've lost my friend. I apologize.

48:39 I lost my friend.
I might I might help you get back to it. So, what I was saying is there's two parts. One is the authority part that helps you get into an index and then two is the customer satisfaction, right? That's the thing. That's the thing that you when you've been talking to Edwards and Edwards has been talking a lot about it which I I think is credible is that they're not as we were discussing in the second half the first half of the show this all knowing AIdriven machine that understands quality blah blah. But what it does understand is how people are relating in to the material. Are they going back to the index because they didn't find the answer from the article? It's giving signals that people are not getting what they're looking for from that. It definitely knows that. That's what you're saying, isn't it?
Yes. And that's what they said in PageRank and that's what they said in the DOJ's. What the DOJ said in their just in their in their evidence, right, is that they're very good at measuring that. And I think that they changed how they measure that, right? And I don't think they measure it with Chrome or with you know, analyt G4. I think they changed one I think that you know what an algorithm looks like, right? Like if you have seven times something divided by something, you change one number and your result goes up, right? You could change a fraction and that could have a dramatic impact if all the other numbers are the same, right? And so I think that they changed and tweaked some of the impact or the waiting that when people bounce or when people click back faster demoded a lot more pages a lot faster.

50:32 And so where you had people that were ranking here and were very were plateauing and then suddenly they dropped down. That's to me that's because Google and I'm just this is purely hypothetical. I can't test it. Right? That's where Google gave sort of said we're we're moving from these metrics to these metrics or we're putting more emphasis on these metrics and that's why some people saw a demotion in in pages.
But it's only analytic not analytic anidotes.
It's only data. Yeah,
It's only at my own energy. A lot of people Well, let's link it to Hub HubSpot, which is become a classic example, hasn't it?
And Nike
And I think you met I missed out. I have to listen to your recent discussion with Edwards about Nike. But this HubSpot is has become a classical example over the last year, 18 months. They had you got this thing called domain authority which I don't know what the way you said that Google really looks at page authority but you got this thing called domain authority as well
Site authority yeah
And I think you do believe in that to some degree don't you
There is a site authority right
Site authority yeah
That's what new pages inherit so new pages inherit a site authority and that's because the domain already has pages with authority so that has filled it up.

52:01 There's so many ways we could go with this because I'm tempted to ask you how page and site domain authority link, but we'll go off on a tangent. So, let's get back to HubSpot. They had they were writing top of funnel. They were writing all sorts of stuff that really in my mind had nothing to do with their core product.
Nothing to do.
Absolutely. Yeah. A lot of the stuff they were writing had nothing and they were getting traffic,
Right?
And now their traffic has been decimated to degree. How much of that traffic actually converted into actual sale of their core services and products? We can only guess because that they have been saying or third party people have been saying that their sales have gone up. I don't know how true that is. But a lot of it was very top offunnel and really didn't have a lot to do with what they were really selling. I think you would agree on that is what and they've seen a massive decline. What did Google decide to do that affected that decline? David,
So I think I mean I don't know, right? These are things we can't know, right?

53:27 Yeah. But we can we can get
You can summize a much more educated summise than I can
That they weighed less on the site authority and started testing what if we put other pages here and what if those other pages perform the same on these metrics but outperform on these metrics and would we would we still show HubSpot and HubSpot lost out? Did Google make an a consc conscious decision to say HubSpot ranking for too many things? No. I think it's a case of factors. One, what is how how thin is your authority being spread across these topics? And which of these topics are other sites performing better than you if we change the metrics we're looking at? And I think those are the changes. And I think that's why so many people saw things that they were ranking for turned upside down and then as their pages started losing clicks, their site authority falls with it and the authority that those pages pass on fall with it.
So it's a kind of waterfall effect.
It's a waterfall effect because Google can't really release an update that impacts more than 5%. They don't allow.

54:45 That's their threshold, right? If more than 5% of or that's 5% was a specific number that Matt Cutts used to talk about but if they if an update ends up changing more so basically they keep a copy of the entire web they rerun all of the most popular queries and if the indexes change if the if the ranking positions are changed by more than 5% they roll it back and so each roll out is probably affecting that harder and harder, if that makes sense. And so, while it had a massive impact on HubSpot, it also had a massive impact on much smaller sites that isn't as visible,
But it also affected or decimated a lot of news sites. I don't understand why news sites got so decimated. It also decimated a lot.
Decimated a lot a lot of Well, I thought I thought we'd keep our site my site out cuz I Right. Go. Fair enough. You brought it up. I wasn't going to bring it up. So, there we go.
It's a good example. Your site, you know, one of your top pages is about GoFundMe, right? Because you talk a lot talk to a lot of communities, right?

56:00 Yeah. A nonprofit community, learning management systems because they it's all kind of interlin. You know, you want to build a community for nonprofits. There's also internal training. There's a lot of overlap, isn't it?
Let's flip the table here. I'll I'll I'll play the devil's advocate, right? And I'll put you on the hot seat. So when you say they were they were interlin, talk to me about the concepts of linking. Like so you've got GoFundMe, you've got WordPress and hosting. How is GoFundMe and WordPress linked?
Well, that's a good that's a good question because they're they're not directly linked.
Yeah,
Conceptually they're linked, right?
Conceptually, in my mind, they're linked because I have a deep knowledge. Well, I think I do. I hopefully people would if they went to the website, they could see that. I I don't talk about cooking. I don't talk about car maintenance.
Yeah, it is diverse, but it is linked to the sector that WP Tonic is involved with, right?
Whereas I would say they're quite disparit. And what was interesting when
I think I think Google thinks that as well [laughter]
And when James was talking on the show with me and Edward about the difference between SEO and AEO, he said that one of the things he found profoundly interesting was the way that we both saw that bridging in a previous episode I did with Edward about starting for with SAS payroll systems and how he started down the ladder. And so if you look at your site and how your authority has been diluted is that there's very little much very little bridge from the pages that do rank to the pages that are important to you.

57:52 Oh that thank you for saying that's oh that's help that's been very helpful. I'm getting a clearer picture now David. Thank you so much for that.
Yeah sure. I'm sorry it took me so long to help you with it but that's effectively what's happening. I get there in the end, but I'm old dog, so but I don't give up. I just
One of the top podcasters on YouTube. Don't give yourself a hard time, but that's effectively what's been happening. And so your site effectively is getting hit by sideswipe by those ripples from how Google is tightening and controlling that authority flow. Google's not going to come down and tell companies that they can't produce a listicle or that they can somehow research all the listicles or that they know the listicles are self- serving. But what they can do is change how that listicle performs or if that listicle could be tested against other content and people prefer the other content then the listical page will drop and I think that's what Google has been tweaking. And then when I talk about I'm not interested in the veneer of SEO, I'm interested in the mechanics of SEO. That's that's precisely what I mean that Google's looking at.
How do we really know that we had the right metrics to test if people were happy with the content,
Right? So, another factor was a lot of people were saying for quite a while, and I think it was linked to the first half of the show, that Google wasn't looking at links so much. There was a there was a lot of talk that Google was looking at the actual content.

59:32 That they, you know, they weren't looking. I've never really concentrated on external because I never wanted to buy links and I just and I didn't want to send out emails. But as you pointed out in previous shows, there's building relationships and sh and sharing links with your group of friends and people you know. Which I've never done, which was a mistake. But the other factor is internal linking I think and it's something I've did and he's and is still doing but I didn't do it very well. So I think I think you've talked about internal linking sharing authority is important and maybe having one or two links and really thinking about how you do your internal links. Is that something you think is really still important?
Yeah, it's absolutely vital. You know, for example, how else will Google know that we're doing the show together unless we link to each other, right? And so, internal linking carries authority just like external links. It's just that the internal link doesn't have as much novelty play in in but it does distribute authority across the pages and that's how you effectively can take a page that talks about GoFundMe and link it to WordPress and hosting for example but you have to build that relevance into that blog post as well right that blog post has to move away from just talking about GoFundMe but also into WordPress and hosting and then it passes that component that topical author authority over.

1:01:26 I think
What Google's very good at is it actually can see where a link will fit on a page if it's being blocked by something else. I've heard of people saying that if an like if a banner or an image blocks that link, the pages that link sometimes links to actually drop in rankings. And Google has a click probability score and if it doesn't think that people will actually click on the link, it won't actually let it pass any authority. So, I think people have to be very intentful and judicious about internal linking. If you think it's just about building hundreds of links inside your pages and you see more crawling activity, that doesn't mean you're doing a good job by PageRank standards. To kind of finish off, do you think the way that Google looks at page authority and site authority, do you think it is attitudes and its responses have been affected in one way by AI? Because it is obviously it's well noted if you use programmatic age production SEO you're churning out hundreds thousands of pages that this mount AI does exist but people got people got confused with mount AI with just producing like a over a couple of years a couple hundred pages But we're talking about hundreds and hundreds and thousands and thousands of pages. But my question is linked to this but in a more indirect way it's become easier to produce content in some ways. You know, to produce written content, to use a large, you know, I think you should be talking about something you actually know about and you should read it because a lot of that content I found has a lot of factual mistakes in it. And you I think just to respect the reader, you should edit it to some degree and read it over and add some other material to just respect for the reader basically. But that's your brand, right?

1:03:49 But that's that's my opinion. But do you think Google I'm trying to I'm thinking how to word this David you know because there's so much more content out there. Is that what a lot of people have been talking about Google indexing but not really showing it? You know, you know, showing up in search console that they've indexed the page, but they've chosen not to show it because a lot of people saying that's indirect comment that you've used AI because it's saying that well, yeah, you've written this and we've index it, but we don't think it's adding anything. But that then you get back to the first half of our show is making a judgment call about the quality and it can't do that. Know that you
So that whole that whole argument is a lot of hot air would you say?
Yeah.
Again it comes back to conflating right. So confusing and sort of like compressing problems together. So in the scale so scaled content is almost universally associated with pSEO, right?

1:05:02 You need pSEO to do AI scaled content. You need AI to write it, right?
Can you can you tell the audience what you mean by pSEO?
So basically pSEO is programmatic SEO where you're basically
That's also where templates come from, right? Going back to the question you asked earlier about templates.
And so all almost all scaled content uses AI content, right? It's an almost one:1 ratio.
Yeah. And so people who are copywriters who want to defend their pipeline in their industry are using that to skew the optics and make it look like Google doesn't like AI content and that's wrong. The second part is that a lot of pSEO also uses templates and that's where one of the ways Google will clamp down on scaled content and Google is absolutely clamping down on scaled content and it's not hundreds of thousands of pages. It's not even thousands of pages. 100 a few hundred pages on a brand new domain is enough
On a brand new domain. Right.
And I don't think Google will stop or give up on I've I've someone left a comment on Reddit. I made me chuckle that you know Google should get with the times scaled content is going nowhere and I think they should get with Google because Google's not going anywhere and scaled content is not going to work out. Now there are definitely like look there are lots of examples of pSEO at work at scale and there are lots of SEOs who are good at that and I use it a lot too and I enjoy it. It's how it's done. But coming back to that statement like Neil Patel made at the start that Google needs to see that you've added something that other pages haven't. That's not something a search engine can look at.

1:06:36 It really isn't. So that's so really that's the crux of this whole discussion and why Neil Patel's video was so I won't say damaging but so wrong. This is the crux of it is that Google cannot make a judgment call about the quality of a page
And it starts to make a uniformal web where some people are enforcing their structure or their pref preferences and it doesn't work that way. I'm sorry. You're allowed to make spelling mistakes or there are certain areas where you might have an let's go back to this sort of like expert expertise thing. And by the way, Google doesn't require that you're an expert at all. It says if people are if people have to make a decision about content like for example, should you take this drug or should you get a payday loan from this site, they need to have trust and they need to know that they're reading from an expert, not Google, right? Google can't validate that you're an expert and being an author doesn't make you an expert. I'm sorry that's stupid and delusional, but it doesn't actually require you to be an expert because there are lots of things that people write about that they're not experts in like travel, right? How are you an expert in travel, right? That's just that's nonsense, right? And so I think people just keep conflating and twisting and trying because they want to make it real. They want to force their standard. Want a uniformal web. But if you look at advertising, when you and I were kids, right, typos weren't allowed in newspapers. Now, when companies are talking to kids or talking to certain demographics, they're exploring a world of misspellings and grammar grammatical mistakes because they want to speak the local language, right? They want to speak in that tone. Google can't come along and enforce that, right? So the minute you start trying to put a quality standard, you're applying a template to the web and that's just a terrible idea, right? And it's the wrong idea and it's not going to happen.

1:08:36 And Google would end up banning lots of pages for lots of reasons. Real experts might come from another country where they speak a different language. Right? The person that came over to the United States after World War II was a Nazi who built rockets in Germany, but he was the only person in the world who knew how to build rockets and jet engines. Now, if he writes something, is he going to be as good a writer as Stephen King? Probably not, because Stephen King doesn't write books about or factual documents about rockets. Does it make the German scientist less of an expert because he doesn't write like Stephen King? No, it doesn't. So, it's al So, I just I understand that people want to simplify things so they can understand it and nobody wants to listen to me. It's much easier to listen to Neil Patel and stick a author bio on your blog post. I'm just trying to tell people that just start using a little bit of critical thinking for just a little bit of time. It's not that hard. But these are not things that Google looks for or rewards. And that everything isn't eat it applies to a couple of YMYL things. And when you start applying that critical thinking, all of Neil Patel's video just falls apart into complete rubbish. He makes a claim about 740 pages yet doesn't link to them. He makes the claim that it proved that they all lost 20%, yet he doesn't show a single chart or graph. If people are making claims and people are telling you things and people are contradicting just basic common sense, it's okay for you to question and then say like why?

1:10:11 What are they selling me?
Are they selling me their brand?
Just to wrap up and then we'll probably end up end the podcast part of the show. If you can stay on for another 5 10 minutes, we can look at
Some of the stuff that people commented and see if the if you want to respond to that any of it. But just

Conclusion and wrap-up

1:10:33 To I think in some and this is only my opinion David my perspective is that you know Neil's got every right to say whatever he wants but
The biggest contradiction I had that is Neil does another podcast with a with another co-host called Eric Siu the marketing school podcast and it's all about using AI and using AI content. The whole I cannot I cannot watch a full episode because I can't stand Eric. It's you know
That's he doesn't gel with me. Eric I can't so I
Whether he's right or wrong it doesn't matter. [clears throat]
He just he just I just can't do it. I can't watch a whole episode because of Eric. He's all he's he's an individual. This is just my opinion. That's all over the place. I mean, literally all over the place. But their whole podcast is about AI and using AI agents and you and letting those AI agents produce all the content,
Right?

1:11:48 So, that's a bit that's a bit of a contradiction, isn't it? Right.
Like his podcast. His podcast was AI written. And I think you've got to ask yourself, what is what is Neil selling you? What are people when people tell you eat is important or this is important? What are they selling you? Maybe they're selling you for your like. Maybe they're selling you for your view time. Maybe what you're doing is you're inflating their expertise in something they've made up so that they're an expert in it. And you've got to ask yourself, why are you participating? Why are you a willing participant saying like, "Wow, this is great information." if it doesn't make sense and it's not true. Right. I think we
Well, I would summize that's why you've kept your agency the layout of your agency the way it is because you actually are very passionate about SEO. You like doing SEO. If you were head of enormous agency, you would no longer be doing SEO. You' just be managing people, wouldn't you? Right.
And you don't want to and you don't want to manage people. You want to do SEO? Is that correct?
Propeller head. Yeah, I'm a propeller head. Yeah, absolutely.
You want to do SEO?
I want to be an engineer. Yeah.

1:13:01 Right. I think it's been a fabulous show. Hopefully you agree, David.
I do. Thanks, Jonathan.
What's the best way for people to find out more about your enormous knowledge around SEO and your common sense? David,
Search for king of AEO, David Quaid. Don't click on James Dy's pages. Just scroll past them. I actually want to do an experiment. Do a search for King of AO and skip past James Dooley. See if we can use Pogo sticking to push him out the way. That'll be a fun experiment for the next few days. But otherwise, just come and find me on LinkedIn or YouTube.
Yeah. And like I say, we do this podcast every Friday at around 12:00 noon. We were slightly early this week only by about five six minutes but it's been a fabulous and if you could join us we also try and answer questions or comments that are left during the show and you will find that on David's YouTube channel. So there will be some bonus content. So it's a good reason to go to his channel and sign up on his channel. We will be back next week with another insightful show. We see you soon.

Sponsor message

1:14:16 This podcast episode is brought to you by Lifter LMS, [music] the leading learning management system solution for WordPress. If you or your client are creating any kind of online course, training based membership website, or any type of e-learning project, Lifter LMS is the most secure, stable, wellsupported solution on the market. Go to lifterlms.com and save 20% at checkout with coupon code podcast 20. That's podcast 20. Enjoy the rest of your show.

Audience SEO Q&A

1:14:57 So, let's have a look at some of these comments. Dave, I don't know if you've been watching any of them.
Yes. I think this is a question from Alan Osborne. So I Yeah, I think we can skip that one. I think he's just thinking about it.
Yeah, I think skip that one.
Yeah. So, I think he's saying you know, it's the elephant in the room. So, yeah, let's dig into this one. Google says use eat, but it's not a ranking factor, right? So let's break this into two cuz Google says it's not a ranking factor is true. Google doesn't say use eat. Google says consider eat and that it considers eat for y. So this annoys [snorts] me because eat is meaningless without y but we use it all the time. So Google doesn't say use eat, right? Google doesn't say you have to be an expert at all ever. Google says some people consider eat for YMYL topics. If Eid is used, is it a way of making content give slight trust signals? Yeah, it is. It could also raise questions, right? It could also make content look untrustworthy. EAT could be Chase's logo, right? Companies like Microsoft and Chase don't put author bios on all of their content, right? So, in many cases, their logo is ET. In some cases, the Microsoft logo is the antithesis of ET, right? Like if you go to Germany to any of the government organizations where they're trying to throw Microsoft out, a Microsoft logo to them would spell the opposite of eat something they do not think they should trust does not have expertise.

1:16:42 The evil empire,
The evil empire. So is it just simply rewriting a content a different stage of tier? It's not about content. You can't sprinkle actually Google it. You can't sprinkle Eid into content. Eat is about whether or not people trust your website and an entity as a whole to buy from it. And for some sites where Google thinks they pose a threat to society and it's not anything and everything to do with money or your life or anything that could be money or your life. Like someone was trying to say to me that if you hire a lawyer you could miss out on a million dollar judgment therefore it's YMYL. I just thought it was a classic case of But where it's YMYL Google may deploy quality engineers to actually investigate your site and it may require certain things but Google's not saying that you can use E to make content more trustworthy. It's not at all what they're saying
And there's so much confusion around that and they're relishing in it. And the problem I see with that all of this has created is the number of people who come back and say like, "Oh, I've tested my content and has an eat score of 99% but I'm not ranking in Google." And it's because everyone has scrubbed authority out of the system. And I can understand that Google doesn't want people to obsess about backlinks. But I did a video called where I I go through the Google SEO starter guide. The SEO starter guide would answer all of these questions. It would have debunked Neil Patel in 10 seconds. I made it in 15 minutes. The Google starter guide answers all of these questions. No SEO wants to read it. It's like they're allergic to it. I read it every day.

1:18:20 I don't read it every day. I use it every day to show people that like the answers already there.
Yeah. So like cosine similarity which is a way of so basically
What does this mean? Sorry, sorry interrupt. What does cosine similarity mean?
So it's basically everything Google does has to be broken into mathematics. Everything a computer processes has to be broken into math. Same mathematics, right? And so, cosine similarity is a way of joining phrases and words together and saying, is this similar? That's that's
All right. That's beyond my paid pay scale, but there we go.
I drunk the inbound Kool-Aid. Eat is taught as the glue between SEO and AEO. I've had Neil Patel success signals. I thought it was Eid, but now I'm seeing a slight different. Thank you. I'm I'm really glad. Thanks Alan for dropping that note. Yeah. Eid is not in SEO and AEO. Again, if you look at GEO, they've built GEO on all of the confir confirmation bias signals that EAT was, right? So, they're going to brand marketers and telling them that LLMs love ET. LLMs love brand. They love I that your brand messaging is on point.

1:19:33 We've had people on Edward's podcast that have said LLMs go out and seek out brands and validate and verify brands and even [laughter] watch those. I just think they're totally bonkers. Like I mean, you know,
Basically, you're looking at people who refuse to use critical thinking. They've just they're like, "I was taught this at college about branding and now Google's been the big evil and it's never rewarded me for building a great brand and now LLM is my hero and it's come to rescue me." Well, it's a kind of it's incestuous because the way especially if you're using it through Marble and you're talking to ChatGPT or any of them and you're having a conversation, I do it slightly, but I resist doing it because I know the consequence of to myself and I've just forgotten the word what's that wordization where you're
Yeah. Do you pronounce it better than me? Cuz I butchered that. How do you pronounce that?
I butchered too. Anthropomorph anthromorphalization. Right.
That's it. There's an enormous amount of that going on, isn't there? Would you agree?

1:20:46 Always has been. Yeah, absolutely. And with LLMs, it almost feels natural, right? I've I've heard of people speaking of Claude like it's a real person.
And I find that creepy.
Yeah. And I never talk to Claude and go like, "But we've already discussed this." So I just the minute I find myself typing something like that, I just I would just stop and walk away. But yeah, it's I I think that again, we almost have a predisposition to want to ascribe those things to Claude, right? That
And I think a lot of people that you've that you've just mentioned, I think there's a lot of that going on.
Definitely is. Yeah,
There's a lot of it. So I laugh of it but I I think they're not aware of it that they are transfer transference you know psychological well-known occurrence where you transfer feelings or notions to somebody else. [snorts] I think that's actually going on quite a bit myself, but what time
A person that I had a discussion with about Google and quality engines that I think you might appreciate on Reddit said, "Imagine if Google really knew everything. Then why is there so much disinformation and conflicting information about SEO?" Right? That's something that they should know. So basically if a PageRanks and said this is how to do SEO then it's almost like Google's had to approve it and that in itself would be a paradox right because Google said to not to not want people to know how SEO works.

1:22:29 Yeah. Well there we go. You lost me there a little bit but there we go. I think it's a good time for us to wrap up. It's been a great show. We will be back next week with another great topic, a great SE session with David that will educate you. We'll see you soon, folks. Bye-bye.