Is the SEO Industry Dead At Last?

Paris Childress and Simon Young of Question Marketing debate whether the SEO industry is dead at last, arguing the hopium retainer playbook is finished while crawlable fundamentals remain. On episode 24 of The GEO Show they cover Google's goto redirect layer, ChatGPT query fan-out harvesting SERPs, OpenAI's Labrador search index, the collapse of keyword research and DA-link tactics for GEO, share of voice as a vanity metric, and why brand positioning in AI conversations is a different job than ranking for traffic.

Paris: All right, live again with Simon Young. What's going on, Simon?

Simon: A lot. So everybody's. It's quite exciting actually because there's a lot of people talking about what we talk about. Loads of people wanna get involved. I think I'm finding there's a lot of questions that are difficult to answer at the moment because it's, um, it's such a new space in terms of getting listed in whichever of the LLMs or all of, so yeah, talking to a lot of people. Got a few more potential new clients, which is good.

Simon: Um, so you and I both see the benefit of that. And.

Simon: yeah, it's, uh, it's moving quickly. But I've been I discovered I had about 300 videos on my YouTube channel that were currently unpublished 'cause I've I have been I have rebranded, right? So I unlisted them all. Yeah. Um, but, uh, Claude, in its wisdom, has said I should re-release all of that and, um, I have done. Yeah.

Paris: We, we spoke about that offline, but let's- let's revisit that for a sec. So you had about 300 videos that were just in some form of a draft or that you had never published for, um, from institution, and now you're gonna relaunch them under the Question Marketing brand?

Simon: Well, quite, yeah, quite. No, quite a lot of them had quite a lot of views. Um, Oh, okay highest, the highest view count on one of them is 750,000. you know, some of, some of them have got significant view counts. Um, I had delisted them just on the basis that, you know, they weren't my current brand. But. within, you know, I'I've cleared that up in all the descriptions.

Simon: Yeah. Okay. Yeah.

Paris: Interesting. Next, next step being to download the transcripts of all of it and, uh, add that to my knowledge base, right?

Simon: Yeah. One of the things that I, uh, I read in the, the GEO Show daily podcast that I'm doing now is that there's a new, a new tool that can parse out from podcast, and I guess podcasts then videos, audio and video basically, they can, they can take that and produce a transcript that's, that's crawlable and, uh, and feed that to the, to the AI engines for crawling.

Paris: Yeah. Yeah. It's gonna It's.

Simon: Well, the other interesting thing is whether you can get that picked up, right? Yeah one of the ex- one of the experiments that I am running is on, is on my own site obviously. So question.marketing being a brand new site as of July, um, I obviously need to get that ranked, cited, and currently I'm struggling with Google picking up a load of the content as perhaps AI generated, which- a chunk of it is clearly.

Simon: But I have added a lot of proof and rec- what is, what I'm calling receipts Yeah 2009 onwards, uh, where, you know, I've been talking about a lot of, a lot of this stuff for a long time. So Claude was talking to me about publishing unique content, all of my video evidence that I've got, hence diving back into all the old YouTube content to pull out evidence that I've been talking about answer engines for a long time.

Paris: Yeah. The receipts, as you call it, or the information gain. Um, but it's Yeah, I think it's a, it's a good strategy to try to republish all those videos. One of the things that I do now, uh, I have this repurposing workflow, it keeps getting a little bit better every cycle, and I'm using Grokbot mostly to do this, but I'm publishing a, an episode page Well, Grokbot publishes an episode page on the website, and it includes the full transcript.

Simon: And I don'And I I've, I've, I'I've been talking to a couple of new clients about recording audio on a Oh, well, they're, they're trying to do it on a daily basis, and- one of the challenges when I ran a YouTube agency was getting people to record content, because obviously, clearly, people don't necessarily wanna go on video.

Simon: Um, the surprising, um, finding from saying to people, "Look, please just record your thoughts," and this idea came from you, from the GEO show, where you're, you're getting a feed of news articles that you.

Simon: wanna comment on. I gave that- to another client, said, you know, "Get yourself a feed of all the industry news that's coming through, record your thoughts on it on a daily basis," and, um, that client is already doing that, and you know, it sounds, it sounds great. Um, I gave- him a couple of hints today to say, "Can we include a few more stats?" You know, let's talk about, uh, it's a recruitment business, so I said, you know, "Can we talk about, uh, salary ranges and where you feel, um,.

Simon: you know, that's moving.

Simon: in certain job titles, etc?" And they're, they're well into it. They're, they're off and running.

Paris: That's great. Yeah. Yeah, to me, that was a br- a minor breakthrough, because I thought I really wanna get the podcast started again, but I don't I need something that's gonna give me a daily script that I can work off of, and this is, this intelligence briefing is great. I actually recreated it, and, uh, it's even better now. And what I'm trying to do next is to add some visuals to it.

Paris: Someone made The first comment that I got from the podcast playlist in YouTube, and I think it's still the only Maybe it's, uh No, there's been a couple of other comments, but this was the first comment. It said, "Great work. Keep " Uh, it said something like, "Keep up the great work, but I would love to see some visuals." Because it's just me, it's just a talking head, and I'm just effectively, uh, kind of just staring into the screen, and it looks like I'm reading, 'cause I am reading a lot.

Paris: of it from the screen.

Paris: And, uh, I experimented with this feature in ChatGPT called Visualize. So if you just t- if you just go into ChatGPT, if you have any kind of data set or anything that if you wanna turn words or numbers into something more visual, you just type @ and then start typing visualize, and there's a special I guess it's a, like a command, and it visualizes the data. So it produced a whole deck.

Paris: It produced a 10-slide deck with visualizations of every story. And, and in fact, a lot of these stories have data.

Simon: Is this new with Astra then, or is this something that you've just discovered?

Paris: No, I don't even have Astra yet. Let me see if I can find it and share it with you 'cause it's here on my I have it in, uh, Keynote, and I'll try to share, uh, the screen here. And give you, uh, oh, no, this is not gonna be as easy as I thought. No. Str- struggling here with the screen sharing tool. Um, Google-.

Simon: No, I had a play with As- with Astra today to build, you know, one of the strategies that you were talking to me about outreach um, influencers. I think we sort of discovered that while we were on one of these lives, and then the podcast piece and the outreach to podcasts seems like a massive opportunity if we can get featured on the right ones.

Paris: Yeah. Did you check out that tool that, uh, was it that radar? That big radar that you said-.

Simon: I haven't honestly been on that one just yet, No um, might be interesting to plug that into what I built this morning with, um, you know, like the guest post or the guest feature on podcasts piece that I'll, I'll, I'll probably release that out to the public at some point, but I'm gonna use it for my own purposes to start with.

Paris: Yeah. Yeah. Great. Well, let's, let's pivot into the main topic because we, I think we're pretty provocative today. The, the title. Another, I mean everybody has been claiming the death of SEO for many years, I'd say at least 10, 10 years. What do you think, Simon? Is it now starting to look like SEO really is fading into the sunset here?

Simon: Well, I am, as you probably know, one of the people, and if people have followed me for years and years, I am one of those people who's bemoaned the SEO industry, having been part of it. only for the fact that I really have felt all the way through that we could have done with a professional qualification and a certification or a professional body.

Simon: And I think every time somebody tried to launch something like that, it just got called a scam or a, you know, top of funnel lead gen or something like that.

Simon: And it would've been nice had there been, you know, some sort of way to qualify who you were dealing with. Um, and a lot of people got sucked into dealing with companies where they're paying thousands a m- a month in terms of retainers, and Mm sold what I would refer to as hopium.

Simon: You know, they don't ever necessarily come to the top of search. The issue we have now, or not the issue, I think it's actually an, uh, an amazing time, we are now moving into cited, trusted, um, reference sites, which then come into.

Simon: AI, as we know. My side of that is I love the questions, you know, because in terms of marketing, it's getting somebody to ask the right question, and answering it then presumably leads to an inquiry. Simple as that. But SEO-wise, yes, absolutely you need all the fundamentals. You need technical SEO, you need the on page, you need the link building, and you need all of that, right? Mm in that respect, SEO is not dead.

Simon: You know, if you wanna call it search engine optimization, fine. Do search engines survive through the next couple of years? I don't know. Um, because we're already seeing a no-click marketing era where Mm are getting the answer from AI. So it could be a case that the term search engine disappears. Um, and then obviously SEO would disappear.

Simon: I think it's a time gen- genuinely that the whole industry gets a cleanup, and all of those people that do the work properly and put the effort in, get the rewards.

Simon: We have been in a world where SEO has been, you know, a bunch of tools that people use to make some lovely reports and maybe they build some links and they do some on page, Having run agencies, we all know that, well, maybe we don't all know, but we know that most agencies that are out there, if you take them and put them on a retainer, um, they don't do what I would call value, right?

Simon: And a, and a, and an- agency relationship between a client and somebody who's running their SEO is very difficult to quantify for the general client. You know, yes, you may well have- been recommended. Yes, they may have produced results in certain industries. The problem is you've had, you've had black hat, you've had all sorts of techniques to get people to the top of search. You've had people charging ridiculous fees to, you know, to people to get them into position one, two, or three, or whatever, right?

Simon: Or manipulating results for people. I- just think if we ask the question is SEO dead, no. The SEO techniques that are there will last the test of time because the fundamentals have to be put in place, and you Right use AI to a greater extent to do that now. Um, and you see people every single day on LinkedIn, there's another, you know, my Claude skill has got 20 different SEO pieces in it, and all these, you know, I've got this many agents running your SEO, and SEO is just run it.

Simon: with AI now.

Simon: I think we're finding that a lot of that is being ignored by the search engines. Um, Google's doing a lot in terms of this redirect, looks odd what they're now trying to do. Um, but remember, and if people are Yeah to this, remember, SEO is not just about Google, and I think this is where we'll find out, does Google win? that's probably the question. Is it a What, what's the, uh- engine anymore?

Paris: I want, I wanna s- stop you for a second on that comment about the redirect, 'cause I wanna dive deeper into that. Can you explain what happened there?

Simon: Google is now, Creating a redirect layer between the, between the sear- the actual search and then the result page, right? Yeah. So it's, it's odd. Um, obviously previously you would try and rank your site on the first page, preferably one, two, or three, get clicks- and those clicks would then come straight to your site.

Simon: It seems like a percentage, and I don't know what that is at the moment, but a good percentage of clicks are now going through a Google/goto redirect to then end up on the site.

Simon: . Which, what, why is that is another matter. I don't know. Um, it's breaking tracking, so it's not helping someone carry a pixel across into a site to get a conversion event, which- this is the other, this is a whole nother Don't get me started on SEO agencies not reporting their results in terms of, you know, actual results from the work that they've done. Um- but we'll You know, I don't know why Google are doing that at the moment.

Simon: It may be the, you know, maybe it will become clear that it is something to do with AI search.

Paris: Do you think that maybe they're trying to disrupt the query fanout process?

Simon: Could, could well be, yeah when AI is doing.

Paris: So most, I mean most of, uh, the prompts that you put in, um, the AI, uh, search engine will go out and do a real time search, and it'll use Google, Bing, and then maybe its other They're building their own search indexes now too, uh, ChatGPT, it was reported recently. But maybe they're, they're trying to disrupt this flow of, um, this synthesis process where the AI engine does a query fan out, and then it does I mean, it might do two or three dozen searches in, uh, in a matter of.

Paris: milliseconds, and then plow through.

Paris: the first three or four pages of results. But maybe it's, uh, it's trying to slow that process down somehow and disrupt it. What, so to effectively maybe break the ability for other, uh, engines to to harvest its content?

Paris: Yeah. Right. I mean, the same way, which is ironic because that's what Google did. Google, uh, did that to publishers. I mean, has done that to publishers for the last 20 years, you know, pulling out and summarizing their content in search results. But now someone is doing it to Google, which is mainly ChatGPT. So ChatGPT gets a prompt that goes out and does a real time search with a query fan out process, call it maybe 20, 30.

Paris: Uh, that prompt gets converted in, in real time to 20 or 30 search terms, and then go, it goes to Google with all those search terms and then tries to pull out and gobble up the first two or three or five pages of results across all those, and then synthesizes the answer. So it's basically, um, it's just basically pulling Google's results. It's doing to Google what Google has done to publishers, which is a really quick scraping and serving up of a synthesized, more compact, efficient answer.

Paris: But But it's- maybe Google says, "I don't wanna let them do that." Well, yeah, I suppose unless they're gonna, um Well, I would imagine a great deal of the job has already been Mm if they've, you know, the What was it I was reading? Is it Brave that has, or is it ChatGPT, has something in the background called Labrador?

Simon: Are you aware of that?

Paris: Yeah. I am, yeah. I reported on that, uh, on yesterday's podcast. Labrador is a project from OpenAI to build their own search index. So they're gonna-.

Simon: Right yeah, and then they won't have to rely on Google or Bing anymore. I think they rely a lot on Bing as well. They won't need to do that for real-time search, and this is where I think, uh, getting back to the is SEO dying, I think that it is. I think classic SEO is dying, and it's mainly because, um, first the consumer's preferences are changing, and people are not going through that old process anymore of typing in a search term with the intent of clicking through a bunch of links.

Simon: and doing their own research.

Simon: Now, people are shortcutting that and getting the answer straight from LLMs. And then the question is, well, doesn't, don't you still need to do, classic SEO in order to be ranked and indexed so that you can be discovered and synthesized into the answer in, when this real-time query fan-out process happens? Well, so far the answer has been yes. But now with ChatGPT, OpenAI building their own search index, their goal is to increasingly rely less and less on having to go to Google or Bing.

Simon: Yeah. And imagine a world maybe a year out from now where OpenAI's own index of the web is robust enough, maybe it's, it's as robust as Google's own index of the web, and they say, "We don't need to go to Google anymore, and we're done." So then do you need to do anything for Google?

Simon: If your goal is to get mentioned, recommended, cited by ChatGPT, if that's the main one you're optimizing for because they have all, most of the market share, and ChatGPT has its own index I'm sorry, just.

Paris: There's been construction banging for months- outside of my window and it still hasn't let up. Um, so if ChatGPT no longer has to go to Google for retrieval for real-time search to provide that real-time search-based answer, and retrieval happens through their own vertical stack, their own then do we need, as SEOs, um, do we need to bother with, if we care mostly about ChatGPT, um, AI search and ChatGPT, do we need to bother with Google, with trying to rank in Google anymore and trying to get our content indexed by.

Paris: Well, we need to get our content indexed by ChatGPT's new Labrador, which I've heard has even like five or six different vertical indices And maybe that's different. Um, maybe there's gonna be a way to submit content straight into that index without even having to publish it on the web. I mean, it almost begs the question, is publishing on the web, gonna remain the only way to, um, to share our proprietary knowledge with AI?

Paris: 'Cause right No, yeah it kind of is, but it doesn't seem like it's the most efficient way.

Simon: No, not, like I spent, I spent, um, probably half an hour this morning submitting all the question.marketing pages across Bing and others. And- for the fact that, you know, they And some of them were picked up immediately. So Google hasn't Google has indexed, um, sorry, I'll correct it. Has crawled all of the question.marketing pages, there're 150 something pages, but it hasn't indexed more than 10 yet. So-.

Simon: Yeah. for me, that feels like, um, it wants me to build the trust, it wants longevity, it wants third-party citation, it wants all of that, all the good stuff, right? But- my question then is if I get some, um, high domain authority references to certain articles, um, that will obviously speed up the process, but what we're then talking about is some people would call that traditional SEO. You know, you're then going Mm to get backlinks.

Paris: I Just promoting content it's not-.

Simon: Yeah. Yeah. It's not a I don't think it's necessarily you would refer to it as a backlink now because backlinks, as we talked about on one of the previous episodes, you can pay hundreds and hundreds if not thousands of dollars, um, to be featured on certain sites when And this is what we've been banging on about in terms of content for years, is earned content, like authoritative pieces of content that want to be shared are 100 times more valuable than going and buying a placement, right?

Simon: Yeah. So- you know, you think if you get the right piece of content on social or the right piece on the right article on, shared by the right person. So, um, you know, I was One of my articles in 2019 talking about answer-engine optimization, it's By any means the first mention. I think Jason Barnard mentioned it in 2018.

Simon: But for me to share an article, imagine if someone like him shared it or commented or got involved, the authori- some of the authority would pass into my article and then make it, if it's unique enough,.

Simon: able to be shared, you know, through a network of professionals and that's what's far more valuable. This is the piece that I don't think the traditional SEO agencies did or I think they probably understood they should be doing it, but it is a really difficult piece of work to go and Yeah, it's much harder than just going out- one, make that content and buying links. Yeah.

Paris: You know, Simon, it, uh, if we, if we try to break down, uh, traditional SEO, I I think about it in three, uh, three buckets. You have, you have your technical on-page SEO, then you have content, and then you have links. And th- these are the three You, you need good strategies for all three. So with technical SEO, your content has to be crawlable. It has to be Fundamentally discoverable and crawlable by bots.

Paris: And then beyond that, you need, you need to get into the index, which is something else. And then on page you need to have I mean, traditionally you should have the right keywords in your meta titles and H1s, H2s, and all that. There's keyword, uh, keyword placement, and I think that's gone now by the way. Um, but and then, and then it moves to content, and then content is, uh, presumably content for human readers.

Paris: Uh, and then you have links, and now links are being replaced by brand mentions. But if we break down each of those three, I could argue that all three of them are really crumbling, because I mean, with only one that may- maybe holds up, which is that your content still needs to be crawlable, it needs to be discoverable by bots, whether they be search bots or AI LLM bots, it doesn't matter.

Paris: If you have a JavaScript website, then your content is not, is not gonna get crawled, uh, by AI bots. You've gotta serve HTML. But ans-.

Simon: answer me this, answer me this Mm and this is another problem with the SEO industry. If you take the standard SEO approach, yes, they built a lovely piece of content, they've published it on a page, they've put in the keywords, they've done the technical stuff where it's structured correctly, and you've got H1s and you've got a paragraph and you've got all of that, right?. A lot of those people then stopped at that and didn't build in the schema.

Simon: So they didn't put in the organizational, the publisher, the authority, the structure in terms of questions and answers. They just couldn't be bothered or they didn't understand Google's actual language. And- One yes, Google is sat there going, "Well, 99% of the pages I'm looking at don't have schema, so I will do my b- best to read it and structure it," right?

Simon: But as soon as people now come forward and say, "Well, here is a properly structured piece of content," which was written, and like you always talk about, fills in an inf- information gap and is unique knowledge, right?. Well, the piece that actually includes the schema and the s- correct structure, it's like Google can now read it, I know we're talking Google again now, but.

Simon: you know, the search engine can pick it up and go, "That's exactly how that was meant to be shown to me." You know, and Mm I don't Do you think that that's part of what SEO should have been and isn't yet or-.

Paris: Yeah, I think that's, that is what SEO should have been, and most people never bothered to do the schema. I'm not yet convinced that the schema, well-structured schema, will lead to, uh, more citations in LLMs. And the reason is that there was a study, and I This was a few episodes ago i- in the GEO show.

Paris: But there was a study on a large, a large data set that showed, uh, unfortunately, that showed no correlation between pages with schema and those pages getting c- being cited in LLM responses.

Paris: And, uh, it makes me think that schema is a nice-to-have, but maybe it's not a must-have if you have well-structured HTML. Uh, I think that you can't have JavaScript, but you do need to have HTML. It should be well-structured. Ideally, it should be structured in the way that a user, a human user, is gonna ask a question ch- to ChatGPT. And so a lot of this is now FAQ that seems to be winning.

Paris: Um, but getting back to the basics of SEO, and this is why I think SEO is gonna It's gonna fade and fade and fade because one part of it is that people are gonna stop clinging to the hope that they're gonna get organic search traffic anymore. They're, they're gonna eventually just let that go. So you have the, the main benefit of SEO is gonna slow- slowly It is slowly dying. It's going away to zero-click.

Paris: The tactics, uh, p- people are gonna start to see that the tactics really are different, um, with the only tactic that remains overlapping between SEO and GEO being that you have to be w- you have to be crawlable. You have to have well-structured, crawlable HTML, and maybe schema.

Paris: That's still up for debate. But beyond that, um, content. You're no longer writing content. You're not writing your blog posts hoping that humans are gonna come and read them, because they're not. Th- just that's reality. So you're writing your pages now, maybe n- they're not blog posts anymore, but they're pages. You're writing those to be able to feed information gain to the models, and the only way that you can do it, which is to publish it on your website and also to pr- try to publish it on third-party sites.

Paris: And so the content game is totally different because you're not writing keyword Uh, content for human readers. And y- and by the way, you're not doing keyword research anymore. That, that silently has already died. There, there is no keyword research for GEO. And, and for SEO, it's absolutely central. So you're writing for bots instead of for humans, so the content game is different, and then that brings us to the link game.

Paris: The link game is up. I mean, people are not gonna pay $500 for a DA 80 because the AI bots don't, don't give a crap about a DA 80. They just wanna see that you're mentioned along, uh, along- by somebody trustworthy alongside, uh, other legitimate brands in your category, and that's gonna bring you credibility. So links are no longer the primary currency, 'cause links are, links are the currency for SEO. Links are not the currency for GEO.

Paris: And so I think, um, as ma- as much as people say there's a lot of overlap between SEO and GEO, I think there's not much, and I think it's going away very fast. And when people I mean, we're, we're, we're pretty much already, we're almost there now. And the last, you know, one of the last nails in the coffin, I know we had that coffin, uh, on our image today. One of the la- and sorry, it was not, it was a tombstone, but anyway.

Paris: Yeah. One of the last nails in the coffin is gonna be when ChatGPT rolls out its own index and they say, "Thank you, Google and Bing, but we don't even need you anymore for our real-time search. We've got it. We've got our own." And then-.

Simon: Well, maybe that'll coincide with their IPO, right? Yeah. And, and we know that they've got enough infrastructure to store the, an index and keep it up to date. I mean, they're building out hundreds of billions of dollars of data centers, so they've got the raw, uh, Well, they would've, they would.

Simon: You, we can't kid I think it's likely that they haven't already harvested the entirety of Google. Of course, they have.

Paris: Yeah, yeah. They must have. It's like all these databases that are out there like Apollo and Clay, all of these guys that are lead-gen-.

Simon: Yeah, mm-hmm databases, they just harvested the entirety of LinkedIn and Mm it in, and now they sell it, right? Mm yeah, they will have, they will have, uh, harvested everything they need, and now they're moving into- third-party citation and particularly, and this is one thing I don't think has been experimented with enough in SEO. Um, I sat with someone, poor, would've been seven, eight years ago, national brand, and we Mm talking about, um, using social media as, uh, an SEO tactic.

Simon: So in Mm of getting our audience to talk about our brand, right? In a n- in a good way. So Mm you've got review sites, great signals for AI. Um, you know, Trustpilot, all sorts. If they're, you know, G2 particularly in software. So if you are getting reviews and people talking about your brand, we know it works. But we don't necessarily see, um, content from brands, particularly in B2B, necessarily being shared enough in social.

Simon: So imagine a scenario where, and let's make one up, but a, an e-commerce brand is talking about a new product, and it's, they're hyping it, and it's coming, and it's something nobody's seen before. Their audience are gonna be all over it, and they're gonna be sharing links, sharing stories, sharing possible ways to use this product, etc. Surely, that's got to be a better signal- to AI and even to traditional SEO than anything.

Paris: You know, Simon, that I was, I was around back then too, and there was a very strong belief that social media sharing was kind of like the new links for SEO. So if you put out a social media post and you got a bunch of shares, uh, well, sharing, likes, any kind of this was when this was even pre, this was pre-Instagram. It was just Facebook and Twitter back then, if you remember.

Paris: So you would put out a blog post, and then you'd put out a social media post promoting that blog post and try to get as many likes, Facebook likes, and, uh, Twitter likes and r- and shares as possible. And people would manipulate the hell out of that because there was a very strong belief that th- this was an SEO ranking signal.

Paris: And then Google came out later, much later, uh, and said, "Guys, no," uh, they I think it was either Matt Cutts or it was the, his p- um, the guy who took over Gary Illyes.

Paris: is it maybe? Um, but they came, Google came out and said, "No guys, we're not These social shares and social signals are not part of our ranking algorithm at all. It's not part of SEO." And it just kind of put a, put a pin and just deflated that whole thing. Well, yeah, you say But maybe there was misgivings.

Simon: but I've I've always argued the fact that social, proper social UGC, as in user generated content around brands, 100% Mm in terms of SEO. And the proof is Mm when, for example, videos are made and Mm come So in You can't rank at the top of Google day one, right? You can, you can do as much If you've got a website and it's even Well, even if it's established, you put up a piece of content, you will not come to the top.

Simon: You have to do the work, you have to take the time, you have to get all the, you have to get all the links, you have to do it all, and it still might not get to the top of Google. Whereas, obviously, Google owns YouTube, and Google will serve video to the top of YouTube in 20 minutes. Brand-new video. You could have a brand-new Mm brand-new website, and you put up a video and it will go, it will be served to people to watch, right?

Simon: Because the signal they want is it engaging? Is it valuable? Right? Now, if you apply that same principle to what we're talking about, new piece of content, is it engaging and is it valuable? And that's the only answer you need really, is does the content satisfy the answer or does somebody find it valuable, right?. And then do multi-people, multiple people find it valuable? And then it becomes the cited piece, right?

Simon: Because it will If, if it's not true or it's, you know, made up claims or whatever, it will eventually, it will just disappear out of the index 'cause people will just ignore it or scroll past it, or obviously time on site, you Mm time on article and scrollability and all those sort of things matter hugely in SEO 'cause the more Yeah stay on the page, the better. The longer they were on there, did they interact with elements of it?

Simon: Did they then go on to buy things, etc, right? Yeah. All of that mattered massively and still does in SEO. But when we move into AI, the problem you've got now is if you are cited People aren't just clicking.

Simon: They're not clicking a link to give a signal to say, "Well, that worked." And if OpenAI move to a model where they lead you through, "This is a piece of content. This is why it's valuable. Yes, you're engaging with it. We can see what you're doing with it. Um, we'll republish that to someone else because that You found it " You know, l- like, if you go right down into the model of, well, let's say Simon is 54 years old, right?

Simon: And he's looking for a set of golf clubs. I'm in market for a product, and I start doing my research traditionally on Google, I would be put into an in-market audience, and I would get served ads, and I would get directed to other products like that, right? But if you take it over into the AI universe, they're, they'll know much more about me probably than I ever know about myself or can remember.

Simon: And they'll go, "Right, Simon's 54. Simon likes golf. Simon has looked at these certain brands. He " And they'll know why I looked at that one and not this one, right? That one was shiny and that one wasn't. I don't know. They'll, they'll know all sorts of stuff that I won't. But when the next Simon appears, i.e., somebody who is very similar to me, and they'll start categorizing search into, well, they already have, right?

Simon: But into very specific demographics to feed another Simon a similar result because Simon found it useful. Do you see where I'm going with that? we end up to AI can do a lot more there that Google, traditional Google can't.

Paris: Yeah. Yeah, the memory, the memory's getting more and more powerful and, um, soon you're gonna be able to draw on so much. I think the promise of real personalized search results never I don't think it was ever really delivered, but it was promised for many years and talked about. Google is so personalized and they used to say, "It's hard to even try to r- rank on Google because everybody gets different results because it's all personalized." But actually, I don't think that really happens, but I think it is definitely happening now.

Paris: There is hyper-personalization. And that I That also, on a side note, that means that the current primary KPI, which is share of voice, I think it's, it's very misleading and I think it's a complete vanity metric. Because these tools, um, especially the tools that are, that are just doing one try, one run on the prompt and seeing what's there, it's, it's just such a crap shoot. And, uh, you're, you're not seeing any kind of personalization.

Paris: You don't know, uh You, you get different results. I even reported today in the, in the podcast that with the launch of ChatGPT 6 Astra, the citations are all different. So, so now what happens to your share of voice report when you used to have, I don't know, 20% share of voice when users were using 5.6 Sol, and then today they're using Astra, and its citations are all different. So, you know, now your share of voice drops.

Paris: Does, does it really matter? I mean, what matters is that you're, you're getting people to eventually come to you, uh, through the web or in person and say, "Uh, ChatGPT recommended you guys. That's why I'm here," or, "By the way, that's why I sent my agent to, to go and shop on my behalf and buy." Um, but that's, that's the only real way to measure properly. And, um, just to go back-.

Simon: But we must, Yeah. Go on. Go ahead.

Paris: No, please go ahead.

Simon: No, no, look, we, look, we have to absolutely end up in a universe where OpenAI, et cetera, are all running their own ads within the platform, because that is Google's model, right? Google's main income is from its ad revenue. But we'Yeah gonna end up, we have to end up in a, um, I am running open ads on OpenAI ads for clients, albeit that the UX of the ads interface is horrendous.

Simon: It's Yeah going back to 1995, right? But-.

Paris: Google Ads, Google AdWords version 1.0 in 1998 Oh, it, yeah, it's just about that bad.

Simon: Yeah. So we're, we're at that stage. Yeah. But it'll get better we must very quickly, we must very quickly get to the point where, and y- we go to, well, go to the Simon scenario. I want to target. I don't, I don't have to put keywords in. I don't have to guess by putting in phrase matches and broad match, all that sort of stuff.

Simon: I can go and give it a prompt and say, "I only wanna show this ad to 54-year-old guys who are in the UK who are looking for tailor-made golf clubs," right?

Simon: That, that's, and you could expand that prompt hugely into, you know, like imagine that a, ooh, an AI voice-driven ads platform within, uh, an LLM. Yeah. Oh, where you Where's the revenue? describe with your voice who you're trying to reach, and then your campaign goes live in one or two minutes.

Paris: Yes. Yeah. Oh, I'm sure that's gonna come. Well, yeah, we could go, we could go down the rabbit hole with voice. Um, but by the way, I read a statistic that Gemini, uh, Google Gemini, uh, they say that they now have, um, they have one billion monthly active users on the Gemini app, the mobile app, and 65% of the prompts, 65% are voice.

Simon: That's amazing. Wow. Isn't it? That's incredible I mean, of course, yeah, it's a mobile app, so you would, you would expect that mo- uh, it's a mobile app, so you would expect a higher degree of voice. But 65% is voice activated, which is-.

Paris: I don't, I don't know that I necessarily believe one billion users. It might be one billion Mm like, sessions. Um, Yeah, yeah. There's probably some overlap with the Google app there of something, where Gemini Yeah powering the Google app, and they say that this is a Gemini user. I don't know. That, I think they're double counting some of those users, because even ChatGPT has barely crossed a billion. I think that they report on weekly active users.

Paris: Yeah. But I think that they've They're, they were at 900 million for a long time. They couldn't, they couldn't quite get there. I think they might have just crossed the one billion. I'm sure now with all the buzz with, uh, with ChatGPT with Astra, ChatGPT 6, I'm sure they're gonna get there now. But, um, just I-.

Paris: I've been, I've been watching a few videos on s- uh, on Astra. Wow, it looks incredible. It, it blew the, it blew the doors off of the benchmarks. It looks like it's actually a significant step above Fable 5.1, which was released, you know, a f- few days before it. But I can't I don't have it. It's not available to me yet, but I can't wait to get my hands on it and try some stuff.

Paris: It looks like you're gonna be able to do serious video editing inside of Astra, and it might, it might just wipe out all of the-.

Simon: Was that was the link I sent you yesterday. I think it's gonna wipe out It, it could wipe out all the video editing tools that are out there quite easily. From what you showed me, um, that-.

Paris: Yeah, that video I sent you? Yeah, it took, it wasn't cheap, right? So it was, you know, to make Well, I say it's not cheap. You know, we used to, as an agency, we used to spend whole days in a studio with a bunch of actors and, you a scriptwriter and a director and cameraman and all sorts, right? And spend 5 to £7,000 in a day to make a piece of content.

Paris: And then- well, the one you sent me, that guy had spent, what? 90 bucks, right? In credits. I think, I think it was more like 30, 30 bucks in, uh- All right in tokens. But, but he was definitely under $100. And it was about, it was about, like, a two-hour workflow from start to finish.

Paris: Yeah. But that was But the prompt to- damn nice video. That was polished. Yeah. With, And- with int- with graphics, motion graphics. Not just stuffing in some B-roll here and there, but there was motion graphics and dynamic stuff happening on the screen. There was his avatar that looked very real.

Paris: I couldn't tell the difference, No, I couldn't that was coming from HeyGen I couldn't tell. Was it Yeah, HeyGen was what it had come out of, right? Yeah, yeah. Yep. And he'd ElevenLabs as well, and he'd used, that's how he used his own voice. Yeah. Well, yeah, you can import your ElevenLabs voice into HeyGen. I've done that before and, uh, it works great. But the last time I used HeyGen was about eight or nine months ago, and apparently it's gotten a lot better.

Simon: So is it, is it just a case of f- like, you're filming yourself and you give it, you give it a sample, and it then Yeah re-.

Paris: Yeah, you have to, uh- re-animate you? We just, you send it some footage, and you can even send it, uh, maybe one minute of footage of yourself talking. And you don't even need the voice, 'cause you do the voice training with ElevenLabs, and then you just bring your ElevenLabs voice into HeyGen. So you just need to, uh, you need to train it on your facial expressions. You should use your hands and do some stuff like this if that's how you naturally speak.

Paris: And then it just mimics It, it will just mimic your, uh, facial expressions, your body language. Uh, I guess your, uh, even just your body language as well, all of that, and it'll, it'll nail it. So-.

Simon: May- maybe then, uh, we all, we all end up as, you know, in the year 2500, we're all one These are one of the OG people from, you know, people collect us as- that's the original Paris. Yeah, right. Like an NFT version of you. Yeah, yeah, like, uh, an original baseball card or something. Yeah.

Paris: I think that there's gonna be a premium for real human-generated content when it gets so much It's gonna be so fast and easy to create this super Uh, content with human avatars too that look very real. But there's gonna be some watermark or some badge on some videos that say, "No, this is really me. I actually physically got in front of the camera and recorded this." And I think that there's gonna be a premium, and there may be, there may be an algorithmic preference too.

Paris: If, if I go into YouTube and say, "I wanna tell YouTube I wanna see human content. I don't want all this AI slop, even though it's entertaining as hell. Um, I wanna see real human beings." And I think that d- there's gonna be some kind of a filter for that, and there's gonna be a premium put on human-generated Not only human-generated video, but human-generated articles, words, I mean text, um, songs, music, where it's a human being singing and not a AI voice.

Paris: Um, human-generated lyrics. Ar- art, I think it's gonna be, We, we're gonna be, um, more precious, I think, because we are finite and AI is infinite.

Simon: Well, I- you've seen the film Max Headroom, haven't you?. Yeah. Classic from the '80s. So, uh, you know, those, the we're almost at that already, aren't we? Yeah, yeah.

Paris: So let's, let's bring it all back to the original debate, Simon. Uh, what are you saying about SEO? Dead, dead, dying, or, um, immortal, or something else?

Simon: Uh, dying and old school, as in, you know, doesn't It's needed, but I think the whole industry needs to grow up very quickly- and move forward with You know, it is not, you cannot call it SEO anymore. Yeah.

Paris: As much as I wanna disagree with you, Simon, just for the sake of making this all more fun for the listeners, I think I would totally agree with you that SEO is dying, and it is a, I think it's a very slow fade into the sunset. Um- it's not, it's not gonna be sudden.

Paris: And the fundamental reason is that the goals are very different. The goal with SEO is to get ranked, to get traffic, and then to convert that traffic in a way that you can measure and you can attribute your return on investment back.

Paris: to the, to those efforts. And GEO, the goal is entirely different. It's not performance marketing. It's brand marketing, and the goal is to get your brand visible. First, visible, but then also positioned the right way, uh, described the right way accurately in positive sentiment by AI. And in, and in other, you know, down the road it's gonna be also to, um, to allow AI to even go and shop or to interact with your brand on a, on behalf of a human user.

Paris: But because the goal is so very different, it's not to rank, it's not to get traffic, it's to position your brand in, uh, in an AI conversation. That's the fundamental, that's the fundamental reason why I think that, uh, SEO is dying, despite the fact that there are certain technical on-page best practices that overlap 100% between SEO and GEO, and that really has to do with, uh, well-structured, crawl-friendly content. But you're no, you're no longer writing the content for humans.

Paris: Uh, you're no longer chasing links, um, with high domain authority and do follow attributes. That's, that game is over. And, uh, and that's the majority of SEO really, uh, content and links, and, uh, that, that has changed massively. So I'm gonna say dying, uh, almost dead, and, um, we're gonna be in, uh, Brighton at BrightonSEO conference next month, and-.

Paris: I'm, I'm gonna be very interested to see how many of the vendors are still pushing, I don't know, link building packages or stuff like that. Um, how much is old school SEO gonna still be represented at Brighton SEO versus is everybody really gonna be talking about GEO now? 'Cause that's, that's where it's at.

Simon: Well, I think they, a lot of them are trying to call it AI SEO, aren't they? Yeah. I don't know. No. Um, that's where a lot of it's trying to go.. But I think fundamentally, like you say, what you're talking about in terms of GEO and the required discipline is completely different. I don't know why, personally, I don't know why, for example, PR agencies are not all over this.

Simon: They should you know, "Hallelujah. This is what we do." You know, "We should be taking your Yeah molding it, pub- publicizing it, and showing everyone why you're different." You know, you can't get, having worked with a lot of PR agencies- you cannot get PR unless you've got something interesting to say, right? Yeah. And, and that's what the web is now. Yeah.

Paris: Actually, we, uh, we spoke to a PR agency last week, and I think that GEO is the best thing that's ever happened to PR agencies. It's, it's giving them a whole new lease on life, Yeah now they have, they have a much larger mission than just, "Let's get you placed inside of, uh, inside of these media," but, "Let's get you placed inside of these media so that now, um, you b- you get famous in AI and ChatGPT, uh, talks about you to everybody." Um, but I, the PR agency, we spoke.

Paris: with a US-based.

Paris: PR agency. They're, they're struggling and they're looking now to make a real pivot and to change the whole discussion around AI search. And they're in a great position because their media relationships, uh, their Rolodexes are still there. I think they can still place if they, as long as they can source good content, they're still doing things the right way, the hard way, which is to go and interview the founder of the company and get the real story, get the real experiences, uh, get the real scoop, and then pitch that exclusively.

Paris: to different, uh, different media.

Paris: sites. And, uh, that's what we're, we're trying to s- to build into, at GEOforge, by the way, that kind of experience where, um, exclusive content can be produced and then, and then targeted and pitched to different, you know, media outlets. But the vi- the,.

Simon: the video g- the video guys should be all over it as well. You know, like there there was a world probably six months ago where all the video guys are like, "We're out of business." You know, because Mm anybody can go and make a video now. To go, only the best brands are now spending a lot of money on highly produced content.

Simon: And then, you know, okay, we're talking about the possibility that AI will now be able to make content, but I think the if the video guys, you know, use real content.

Simon: in that. And the hardest part is gonna be how does AI eventually move to disseminate human UGC versus AI-produced, even if it's high-quality AI, and then, you know, what is slop, right? And the slop Mm obviously get excluded 'cause it's just lazy. Um, Yeah what world do we end up in in terms of, you know, i- if it does feel like you've described a complete divergence between traditional volumetric content production without a lot of thought to think properly, make it unique and valuable, and now go out and share it, right?

Simon: the sharing piece is that way, you know, brand comes into it at that point. It'll be so interesting to see how it develops. Yeah.

Paris: All right, Simon. Well, we've gone a bit over our 30-minute mark again, but, uh, why don't we wrap it up? We both have agreed here that SEO is dying a slow but dignified death, I would say. Um. And, uh, certain fundamentals are still there, but the entire rules of the game have changed, the goals, the tactics, the strategies. Um, yes, the, it, there still is a component of producing content, but for a different audience, for, uh, a non-human audience.

Paris: And yes, there still is a path of promoting your content, but not through link exchanges, but the more legitimate, uh, human-to-human way. Um, but that said, it's, um, the strategies that would help you win on SEO are no longer the winning strategies. The, that playbook, if SEO is a playbook, then SEO really is dead because that playbook is dead. you can argue that, yeah, you still need to create content, you still need to share and connect with people, and have other people talk about you.

Paris: But the, uh, the actual playbook is entirely different Well, So any closing words, Simon?

Simon: aft- after this episode, I will, once I've shared this on my LinkedIn, uh, I'm going to tell my Claude to go and find all the people that have ever said SEO is dead, and say to them, "Here's a nice interesting article. Maybe you'd like to watch, contribute, and comment on it," and see if we get, Nice some interaction that way, which will be good. Um-.

Paris: That's a great idea doing I found, just as a closing thought, I found somebody who commented on one of my posts in 2019, and he said something along the lines of, "answer-engine optimization isn't a thing, and it won't be ever sold as an agency. Uh, what are you on about?". And I went back and commented on that seven years later and said, "This comment aged well.

Paris: Not." So, you Yeah in conclusion, the, the SEO guys, there's hundreds and hundreds of people that have always said SEO is dead.

Paris: I think you're right. The playbook has just been thrown out the window. Um, the SEO guys will argue, "Well, no, we're mo- we're evolving." Well, I don't think Mm quite understand where this is moving to. Um, by the way, I found it kind of funny. At the top of my LinkedIn feed, uh, the last time I checked, uh, a major influencer called Cyrus Shepard, I don't know if you've heard of him, he was formerly a former Moz um, he had, he has completed a major survey that, it used to.

Paris: be a Moz.

Paris: survey, I think he took it over, where they asked 131 SEOs, some of the most respected in the industry, what factors are the most important for ranking in Google in 2026? And, um, he just made a big post with all their faces. It got a lot of likes, and he said, "Yeah, big survey. Grateful for the time and generosity. Uh, collectively they contributed over 13,000 ranking factor data points and tons of deep insights." So Google ranking factors survey.

Paris: What is the relevance of that today, Simon, in 2026? I- is Does anybody wanna rank in Google? You just have to get ranked enough that you don't get, you don't get skipped by AI.

Simon: Don't get de-ranked, yeah. Don't, don't You don't nee- you don't need to be on page one. You just need to not get skipped by AI. Yeah. And when we go back to the very beginnings of the web, as in you've put a page up, can you direct traffic to it?. Not rank it in Google. Can it be found and can somebody get to it through, you know, a search and find your brand and discover you?

Simon: 'Cause they can'Well- stop people putting pages on the web Right. Yeah.

Paris: All right, Simon, why don't we wrap it up there? And been a lively debate. I hope, uh, hope the audience enjoyed it I'll And- tag Cyrus in and see if he watches Yeah, I think I will, I will do that. I don't, I don't wanna antagonize the SEO community here. It's a great community.

Paris: I'm, I've been a part of it for a long time, as have you. Uh, but I do think collectively we all need to see that, um, we should not be talking about how we should not be doing ranking factors surveys anymore in 2026. It doesn't matter. Or keyword research and stuff like that.

Paris: All right. Let's wrap it for today, Simon, and thanks for those who have listened all the way through, and we'll be back here Friday, same time, right, Simon?

Simon: We will. All right. Okay, everybody. Take care.

‍

Is the SEO Industry Dead At Last?

Key takeaways

Is the SEO industry dead, or is the playbook dead?

Why do Google redirects and ChatGPT fan-out change the game?

What still overlaps between SEO and GEO?

Why is share of voice a weak primary KPI?

Who wins if the SEO playbook dies?

Notable moments

Full transcript

See how AI answers describe your brand

platform

Resources

About us

use cases

compare us