213: Why Google & ChatGPT Are Ignoring Your "Dead" Content with Jono Alderson

Jono Alderson
===

[00:00:00] Happy 2026. What a start, right? So most of the internet is dead already. We just haven't realized it. Fully quite yet, and in this episode, the first episode in 2026, my friend Jonah Alderson, the technical SEO specialist, you need to listen, even if it hurts, your feelings will argue that 90% of websites are now zombies, so technically online, but functionally, they're dead to the AI agents and LLMs that now control a lot of the search.

[00:00:33] So if you're strategy. If your strategy in 2026 is to publish more blog posts or tweak your meta tags, you are optimizing for something that's not there anymore. Jono argues that the threshold, the era of dragging humans to your URL is officially over. So we discuss why AI models have developed this immune system that treats marketing fluff like a virus.

[00:00:56] Why things like LMS Dxt. Or a trap and you should be focusing on some other things and why. The only way to survive 2026 is something called Upstream engineering. Upstream engineering, Jono Dictionary, the terms he uses on his blog and, and, and when he talks about these things is a fascinating thing and just, just listen to the guy.

[00:01:18] So this conversation might scare you, but it's a necessary wake up call. So let's dive in.

[00:01:24] ​

[00:01:32] Sani: Jonah, welcome to the podcast. Uh, where is the threshold in 2026?

[00:01:39] Jono: Where is the threshold? What a great question, and thanks for having me. Um, uh, I think it erodes and disappears, so let's, let's take a step back. I think, um, all of, all of marketing and advertising throughout all of human history has been based on the singular concept of interrupting, disrupting, convincing, shouting large enough and convincingly enough. To drag humans across a threshold. And historically that was a physical thing, like come into my store off the street, I sell discount t-shirts. Um, and then very recently it became a digital thing where we go read my advert, uh, look at my LinkedIn nonsense, et cetera, et cetera. But this core concept is the same.

[00:02:18] Jono: It's all trying to get people to take that step across that threshold. And that's significant because before they do that, whilst they're out there in the the wild world, they are, um, impressionable. Um, open to manipulation from competitors. We'll see other people's adverts in messaging and aren't yours to control and influence. And that's scary. And, um, it, it makes it very, very hard to exert influence. And then the second they cross that threshold, do whatever you want, you can, uh, influence the messaging they receive. You can shape and craft the narrative. You can influence them to take actions. The whole discipline of CRO and experimentations based around this, right?

[00:02:55] Jono: We go, they're our hours. How do we squeeze the most value? Um, yeah. And that's been the defacto reality forever. And now, as of tomorrow ish, I think that's no longer the case. I

[00:03:05] Jono: think. We've entered a world where machines disintermediate or intermediate, um, and they make that decision and they don't allow us to access the audience.

[00:03:13] Jono: So the threshold's gone.

[00:03:14] Sani: You mentioned the, uh, CRO and the beginning of CRO for me and a few people I know, a lot of people I know was Lyft model, like getting familiar with the Lyft model. The only thing that matters outta the lift model is clarity right now. I mean, I'm not going to say nothing else absolutely doesn't matter, but those nudges you can have, like if you fabricated sense of urgency like you need to buy now or

[00:03:36] Jono: Yeah. Yeah, yeah.

[00:03:37] Sani: that is when the brow, the.

[00:03:40] Sani: Not the person when, when the entity browsing. Your website is not a human, it's an agent or any machine that does not matter at all. Like that, that that's going to be completely irrelevant. But clarity and then. And Not just on your website, but everywhere. And we'll talk about this 'cause you have a really good analogy for this as well.

[00:03:57] Sani: So, uh, the web we are living in, you called it surface less web. Um, what does that mean?

[00:04:05] Jono: So I, I think the web we've had so far has been very tied to publishing web pages on websites, on domains in neat little bundles. Designed to do this threshold thing right there is The Surface-less Webout there, which is search engines and social media. And there's in here, which is our website and our domain, and humans have. Broadly consumed the website through that modality. Since the beginning of the web, there's been experiments in different forms and different approaches, but generally speaking, we visit websites and add things to checkouts on URLs. Lovely, very cute. Very neat. Very much designed for addressable concepts and information that live at a place, and then you go.

[00:04:43] Jono: Okay. Now, increasingly the primary consumers of the internet might not be humans, but might be bots and agents acting on user's behalf or on their own agendas. Um, and then you start to ask, actually, is this modality the most sensible one? And is it even still relevant? And if I don't know, for example, chat, GPT is trained on the whole corpus of the web at some point. It doesn't really necessarily care or understand that my blog is on a subdomain and it's a separate entity or that that Christmas microsite I launched in 2013 on a separate URL is no longer relevant. Right? It's, it's not doing that human thing. We do a

[00:05:18] Jono: passing and welding or that necessarily, and then you start to ask some interesting questions, I think around. Is the smallest unit of consumption at URL and is the most sensible way to bundle URL's Websites. Maybe even not, maybe. Maybe my brand as a concept is an aggregation of statements on Sure. My website, but also Reddit and also that sub domain from 2013 and also probably, I don't know, the footage and a YouTube video of somebody driving past a billboard that I ran in 2020.

[00:05:47] Jono: Like all of these things in aggregate. Are my entity and how systems understand them. So this idea that we're kind of constrained to put text on a webpage, on a website and that being the, the format that we operate in, maybe not so relevant now. Who, who knows?

[00:06:02] Sani: Does that sound scary? Because it's almost impossible to track with the tools we 

[00:06:05] Jono: Yeah. Yeah, yeah. Yeah. Certainly, certainly the, the crutch and the lies that we've told ourselves for the last decade or so. It maybe be two decades of we can neatly. Um, attribute impression, share through to clicks, through to actions, through to revenue, and we can tie it all back again, blah, blah, blah. It's never been true

[00:06:21] Jono: and it's getting less true. And yeah, I find, I find myself, I find myself regretting younger mes um, um, dismissal of. Old school conventional advertising metrics and measurement, right? It turns out that column inches and salient studies and surveys might actually be the only things that we have. Um, and actually trying to interrogate panels of people and ask them what they think and feel might be far more effective, might be the only thing we. Um, when we don't have insight into how all of these systems and machines evaluate these in inputs, but that's always been the case. That's how people make decisions, right? We've just pretended it's not 'cause it's convenient

[00:07:05] Jono: and that billboard 

[00:07:05] Sani: was convenient that it was always about convenience and nothing but convenience. And Google was really, uh, at the right time, right place to push Google Analytics and, and like. Make us 

[00:07:15] Jono: Yeah. Yeah. Everything is channels and the whole world operates nevermind humans. Uh, yeah. 

[00:07:19] Sani: a guest on this podcast and the, the analogy was, or the question was, how come Dom Draper was better at branding than we are today with all the tools we 

[00:07:26] Jono: Yeah. Yeah. 'cause we made ourselves stupid and lazy. Definitely.

[00:07:30] Sani: wait for AI to make us even more stupid and lazy, because this was without ai. But yeah, this is a, this is a very interesting concept and a lot of people are talking about it. There's a lot of. I dunno how you feel about the, the term GEO expert, but there's a, there's a lot of GEO experts who never did any SEO in the past and they think they cracked it.

[00:07:50] Sani: I, I, I don't quite agree. I don't think anyone has, I don't think even, even the, the people building these LLMs really have cracked how they actually work in what, what they will say, but the shift from a convenient way to measure what's happening on your website and with your brand online. To whatever is next and to, to measuring how entities are perceived is, is going to be very painful for several years until we find a way and maybe ads, maybe ads inside LLMs are really, you know, those breadcrumbs that tell us this has a meaning that your, your position for when you pay and when you don't, you're nowhere.

[00:08:25] Sani: And, and maybe, maybe that's step. Uh, you also talk about, uh, it's your blog. And by the way, everyone go to John's blog and read every single blog post. The zombie web theory is another analogy that, that you have there. So basically you're saying that it's a concept where most websites are zombies. They're technically online, but functionally that this is harsh because agents cannot parse their heavy JavaScript and their bloated design, and you're preaching to the choir here.

[00:08:52] Sani: When you say that to me, because yes, of course React doesn't need to exist, but what's the way out of this? If, if there, if a brand has issues with slow websites, 

[00:09:04] Jono: Oh, yikes. 

[00:09:05] Sani: what is the way out? Do they just, important is the problem? Let's start there.

[00:09:10] Jono: That's a great question. So I think that the technical bits of it are quite down, like the, this is slow and this is bloated is almost a secondary problem. Like it's definitely a huge challenge at the moment that if you want to continue playing today's game of drag people across the threshold, convince them even with diminishing returns, you can still do that for a bit, but the systems, the LLMs, the agents, et cetera, are going to be far less tolerant. Of the kind of technical bloat and lack of quality control that Google has been tolerant of for the last decade or so. So everybody who's never quite got round to making their site fast enough or is still on React and is still building single page apps with 20 megabytes of blah, blah, blah, blah, blah, blah.

[00:09:49] Jono: Yeah, those types are gonna struggle because chat, GPT, OpenAI, Microsoft, so aren't incentivized to rebuild the same kind of systems that Google has built to force the way through that. So that's definitely a problem. I think the bigger existential challenge though is that many, many, many, many, many of these websites never truly provided distinct value. Like they, you think about, like, I use this example a lot, but you think about every dentist on the planet, they all have the same website and it's the same 12 pages and the same crappy blog with the same articles and, and

[00:10:22] Jono: top list, et cetera. None of those created any value. None of them said anything of interest or anything new.

[00:10:29] Jono: And if you are. An LLM or chat GPT today, you've already extracted all of that value. You already understand the pros and cons of teeth whitening versus whatever, blah, blah, blah. Everything those sites have ever set is now essentially worthless. Like it's just commodity. And if you were Google or OpenAI. What's your incentive to send traffic to those or to recommend

[00:10:51] Jono: them or to cite them when they're not solving useful problems? So all of the value's been extracted and what's left is just kind of a husk that continues to act as a brochure for humans that

[00:11:01] Jono: want to interrogate it directly and maybe see URLs on billboards. But it's not, it's not a source and it's not gonna, and businesses, this really scares me because I don't think many businesses have ever. Truly considered, what unique value does our website create and provide? Because businesses aren't designed to do that. They're designed to drag people across the threshold and

[00:11:23] Jono: sell to them. It's never been a function. So now we have so much of the web that just no longer has any value, certainly not as an input to the systems that will intermediate everything.

[00:11:33] Jono: So that's, that's a fun 

[00:11:34] Sani: marketing. Content marketing at, at its worst,

[00:11:36] Jono: Content marketing, like those two words don't go together.

[00:11:39] Sani: it doesn't make sense. It doesn't? Well, well when the purpose of content is, like you say, marketing and it's not even marketing, that's, that's just tricking Google into sending some love our way and 

[00:11:51] Jono: Yeah. It's a bag of tricks. Yeah, 

[00:11:52] Sani: It, it, it's horrible.

[00:11:54] Jono: I like the whole word content, right? So there's so little of it on the web. Like it should mean the thing that the thing is and contains the value like type, but it's just meant words on a page and yeah, it's not valuable anymore,

[00:12:06] Jono: it ever was, but. 

[00:12:06] Sani: whenever there's a, a boss or, or a CEO saying, we need more blog posts.

[00:12:11] Jono: Yeah. Really. Ugh.

[00:12:13] Sani: Do you, why do you need more blog post now? Next topic is, uh, next term, and we are going through terms that you coined. 

[00:12:19] Jono: yeah. We're just doing dictionary. Dictionary corner. this is great. 

[00:12:22] Sani: immune system, and I reread this blog post this morning. Uh, it's an idea that AI agents will objectively reject marketing as noise and all of the damage you do to your website or to your brand by having a, a horrible website or horrible experiences elsewhere.

[00:12:38] Sani: Will stay in the system because like, like immune system, it rebuilds and then it remembers what happened in the past. So tell me about, let's unpack, let's unpack machine immune system. Why does, why would an LLM treat, uh, sales copy as an allergen Almost,

[00:12:51] Jono: Yeah. I love this. It scares me 'cause it's the same flavor, the same problem. Like most websites are not designed to be good. Most websites are designed to say we have the best product, and it maybe turns out that they don't. Or that their

[00:13:04] Jono: product is the best, but it's twice the price of the rest or that their product is average.

[00:13:08] Jono: But their CEO, um, behaves in object objectionable ways, or they used to have the best product and they're like, most of marketing, I think can certainly on the web and content marketing is designed to be kind of distraction, right? If you, if you don't have the best product, you say we're the cheapest. If you have the most expensive product, you say we're the best quality.

[00:13:27] Jono: And all of this is storytelling and. And it's storytelling designed to influence and divert and convince humans and to find the right intersect in the best cases between um, people who aren't necessarily informed about all the options and are time scarce, et cetera, et cetera. And yeah, the best versions of this are about fit, but the reality is most of it is about manipulation. And then you go, okay, dear Chat, GPT, find me the best smartwatch. And it's. Smart enough to have done the legwork and the research and to evaluated everything everyone's ever said on Reddit and to have seen the billboards and to have understand babababababa blah.

[00:14:05] Jono: And it can cut through all of that. And it's incentivized to do a better job of understanding the actual quality and the actual fit and to start to learn to see through. All of the fuzzy marketing stuff designed to manipulate humans. You mentioned kind of false scarcity earlier. I think that's a really good example. All of the emotional manipulation, all of the tools and tricks that we've used as an industry to try and do that nudge to get people to cross the threshold, all of that is meaningless.

[00:14:35] Jono: I mean, two, two degrees. I think one is meaningless because these systems will see through it, but also if we consider. And to how these systems work, that they're vectorizing text, that they're flattening websites, that they are trying to evaluate the inherent value of the thing rather than the messaging about it. All of that just disappears anyway

[00:14:52] Jono: in that kind of chunking process. Right. They're not retaining, oh, that was a particularly nice layout or the way that you phrased this was particularly compelling. Certainly not in,

[00:14:59] Sani: the intro to that recipe took me back to Florence. When, when, when they're 20 years ago. I'll give you four words that started all of this bullshit and, and are responsible for, for most of it. Sell me this pen because that 

[00:15:11] Jono: Yeah, yeah, yeah. Yep. I hate it. 

[00:15:13] Sani: Which is Jordan fucking Belfort a criminal, by the way.

[00:15:17] Sani: He, he's the one who did it. And then you need to tell a story about how cool your product cheap or whatever. No, just show me, give me the, the, I dunno if you saw parks and recreation, but give me the, yeah, give me 

[00:15:28] Jono: yeah. Gimme a breakdown of its 

[00:15:29] Sani: Give me the best possible product and I need see if I want to buy it.

[00:15:33] Sani: And we are entering this era now. I liked it. So if that sexy, exciting storytelling content. Doesn't matter in the future. What does, does the boring thing, like terms, conditions, like all, all, all, do you need to focus on that or shipping policies? Is that 

[00:15:49] Jono: Yeah, I think so. I think everyone's equally done a bad job of

[00:15:52] Jono: this because it's boring and because it's work. Because having an e-commerce store and actually doing the work to describe. What is the fabric this thing's made out of and what's its dimensions and the other, and to do that at scale in a way that's meaningful and adds value.

[00:16:06] Jono: I think that's super important. If these systems are going out into the web and collecting all the data and all the evidence that they can find through all the time, and logging that and storing it and learning from it. Then failing to have clarity on Yeah. Things like what is my return policy and um, when was my company founded?

[00:16:25] Jono: And like all of this stuff that is just, should just be grounding for being reputable and trustworthy. Many organizations never get round to, because it's not sell me this pen, it's not tied to the, the end of the transactional process.

[00:16:36] Sani: Storytelling matters more. Uh, not, not only when the company was founded, but what do we stand for? Like, why do we exist? 

[00:16:42] Jono: so I love this. Yeah. 

[00:16:43] Sani: we a company? What, what do we want to change in the world and what kind of person will resonate with us as a brand?

[00:16:48] Jono: Yes, and there's a lot of pushback on the kind of brand purpose thing. I know there's been a huge amount of sort of, do millennials really what the, yada yada, uh, very interesting. But yeah, I think there is definitely a deeper point that being able to clearly and consistently articulate. What is your mission?

[00:17:04] Jono: What is your purpose? What problems do you solve? Like, which that is gonna be super important because again, these systems are gonna be acting on our behalf with a reasonable degree of understanding about what are my preferences, my past behaviors, and my expectations, et cetera. And they are going to try and align with that. Um, but then again, you have the same problem. Many organizations do a terrible job at this because they've never been forced to actually articulate what differentiates us from our competitors. And the reality is probably not a lot. And the,

[00:17:29] Jono: the vast majority of those dentists and insurance resellers and yada yada, other than maybe hyperlocal stuff aren't

[00:17:35] Sani: Hyperlocal is different. I think that, that you use Google Maps or whatever map tool to, to search for those. When you need a dentist, I, I'll type it in the map. Not, not Google, um, but all airlines will always be like matchmakers between the customers and the brands. And like you care about, uh, I dunno, environment.

[00:17:53] Sani: The, these brands care about environment as well. If you wanna buy something, buy from them, consider them first. I like that a lot because maybe that's how it should have been always and not the loudest, most, uh, convoluted storyteller wins. But hey, um, you also talked about machine punishing incoherence.

[00:18:10] Sani: So if your website says one thing and Reddi says another thing.

[00:18:13] Jono: hmm.

[00:18:14] Sani: This was never a thing. No one cared about this until recently. Uh, because your website, like you say, if I get you in a dark room, my website, I can convince you to buy anything or I believe I can convince you to buy anything. Now it's different.

[00:18:26] Sani: Now there's no entering the room almost, or just people looking into the room

[00:18:30] Jono: And people forget,

[00:18:31] Jono: right? They like, oh, that that, I dunno that that big shoe brand that had that scandal in 2015 is now a distant memory, but not for LLMs, right? That's still very relevant.

[00:18:41] Sani: Bill Cosby is not coming back because of L lom, is that what you're saying?

[00:18:45] Jono: Damn it. If there's one thing that chat, GBT Achieves,

[00:18:48] Jono: wow. Yeah. Yeah. I think, I think that through time thing is hugely underestimated for a bunch of reasons. One is, yeah, humans, humans forget very quickly, and when I'm trying to pick a brand to buy a set of shoes from. That incoherence of them saying, oh, we're great and we're legitimate and we're eco-friendly because, 'cause they were forced to be, because they reacted to some scandal that everyone's forgotten about.

[00:19:07] Jono: Mm. Awkward.

[00:19:08] Jono: Um, and then if it turns out that everyone on Reddit says, well actually they have this, um, terrible practice of I know encouraging mining, blah, blah, blah, blah, blah, I dunno, whatever. Um, but then that through time as well, I think is super interesting because all of this gets encoded and remembered, which makes the cost of. That incoherence and failure far higher. And if you think most, most businesses and startups kind of bumble along in their formative years and they make a few mistakes and the customer

[00:19:35] Jono: service isn't great and their returns policy, blah, blah, blah, all of those things, you, you, then, you then grow out of that.

[00:19:40] Jono: If you're successful and you get good, but now all of that haunts you, right? So the cost of kind of innovating badly is far higher.

[00:19:47] Sani: but. Will they be able to just pay and influence LM some other way?

[00:19:52] Jono: Yeah, maybe, and maybe that's paying to influence influences,

[00:19:55] Jono: right? If, if all of these things are fed and fueled by the collective corpus of everything that

[00:20:00] Jono: everyone says and doesn't say across the web, you can definitely use money to do that. And this scares me because that, that allows large and established and incumbent brands with deep pockets to kind of. Pave over the very problems that they should solve, and you can continue to be loud and just shift your budgets from, I don't know, writing blog posts to paying influencers on YouTube much in the same way that you just kinda outsource it to an agency and hope that it goes away. Yeah, you can still do that. And then we end up in a really weird place where the market is essentially the very large brands who can buy reputation and the very small niche brands who are artisanally useful, interesting, but can't target a big enough market to survive and nothing in the middle.

[00:20:42] Sani: But isn't that what we have now

[00:20:45] Jono: Yeah, yeah, yeah. You

[00:20:46] Jono: can, you can have the bakery or 

[00:20:47] Sani: There, there's something in the middle now, but, but 

[00:20:49] Jono: But, but no value, like the stuff in the middle is all just dross. Right. Interesting.

[00:20:53] Sani: That is so scary that, that, that I don't. I like a lot of things about the future and AI changing the fu, removing the hacks. Like this is, this is what I love about it the most because all those stupid hacks that worked don't work as well.

[00:21:08] Sani: And, and, and I love all of that, 

[00:21:10] Jono: That's kind of like why I still like SEO, despite its turbulence is in theory when it's done well and it's done

[00:21:15] Jono: right. It's about promoting value,

[00:21:18] Sani: it always works. It will work forever when it's done, right? Yes. I, uh. I like SEO more and more and more. I, I spent a few years in CRO and I think it just resonates better with me. I think more future-proof skill is SEO. If you're going to compare those two. And it's not even close. Uh, let's talk about the future.

[00:21:37] Sani: Let's talk about the infrastructure for the future. Let's talk about AI agents. And one thing I said about, uh, AI agents at Conversion Hotel. This is the year of AI agents because search interest for AI agents is high, not because. It's real, not because it's really 

[00:21:53] Jono: Give off mobile. 

[00:21:54] Sani: Well, I mean, it's coming.

[00:21:56] Sani: It's coming. At some point. People seem to care about it more than they care about it before. So AI agents, MCP, what will the plumbing of the internet 2026 and beyond look like?

[00:22:07] Jono: Love it. So I think the thing it won't look like, which we can kill dead in immediately, is this idea that we have like agent versions of websites and LLMs text files and kind of separate instructions

[00:22:18] Sani: doesn't work. No one ever said that works. That's wishful thinking at, at best 

[00:22:23] Jono: And I, I think it's, it's crazy for two reasons. One is none of the systems out there will ever be incentivized to trust

[00:22:29] Jono: it because why would I not just lie? and and for the same reason that Google won't trust me to have a separate version of my pages for them, like, and they're always gonna want to reconcile it against the

[00:22:41] Jono: original to make sure I'm not lying, which is three times the processing costs. It doesn't make sense. Also, it exposes all of my business critical information in a super easy way that competitors can just pass blah, blah, blah.

[00:22:51] Jono: None of that makes sense. So I don't think we end up in a world where we maintain like a simpler, two different versions of the system. And I even thinking further down the line, I think when you look at the journey Google's been on from scraping text to rendering and interacting with four pages in order to understand what is the experience this is trying to deliver, I think those system LMS will need to go on that same journey.

[00:23:13] Jono: It won't take them, it'll take them years, but they will want to understand. What does this page look like and feel like, rather than just what does the vectorized text on it? Um, for the same reasons s Google does. So I don't think we end up with a bifurcated web.

[00:23:25] Sani: Uh, another reason, another reason for that, people are bad at maintaining one version of their 

[00:23:30] Jono: Sure. Like, yeah. And I'm seeing this as a crutch where you can go maybe, maybe we don't need to make our website fast and accessible and actually callable. Yeah. I can see why that's tempting.

[00:23:39] Jono: Um, but no, it's not a

[00:23:41] Jono: thing. Definitely not a 

[00:23:41] Sani: that's not how it works. So the future is not that. The future LMS dot txt. Uh, and, and also when you ask any LLM, how should I optimize my website for agents, it always says LLMs T xd, that that's the first thing it says. And then you check, Hey, can you do some deep research to see if that actually works?

[00:23:59] Sani: Well about that. It's a proposal. Some people are talking. No, uh, it's just not a thing yet. Maybe it will be, but no one has even suggested or hinted that this 

[00:24:10] Jono: No, I don't think they will. And when you look at what's chatt PT Open AI's roadmap, realistically, they've got a two year runway before funding starts to become a really

[00:24:18] Jono: big problem. There's no way their priorities this. Yeah. Yeah. This isn't top of that todo

[00:24:23] Sani: No, absolutely not. They have the, the sex chat and all those things that are working on at the moment. So yes, uh, so it's not that Is MCP something that has a future, a real future replacing the websites even.

[00:24:38] Jono: Maybe, or something like it, right? The idea that as a website. Or as a platform, or as a system or a company, I ought to have a structured, consistent way of exposing what is the content I have, what are the services I offer, what systems do I integrate into and out of, and allowing, making it easy for other systems to do that.

[00:24:58] Jono: Yeah, I think this is very much a flavor, the web that Tim owners leave first imagined, albeit in a very different format, right? It's not schema and markup and HTML, it's agents traversing. F functionality endpoints. Yeah, I think that's super interesting. Um, I, I think there's some challenges with MCP as a format.

[00:25:17] Jono: It's not quite there yet. There are some risks with it as well. Um, and I think there's a, a deeper underlying existential risk, which is the same one we've talked about, which is most websites don't have value.

[00:25:29] Sani: Yeah, 

[00:25:29] Jono: And if you, we get to the point where we go, okay, agents can use my MCP endpoint

[00:25:33] Jono: to query all my things.

[00:25:34] Jono: And you go, what are they? What are they gonna do?

[00:25:36] Jono: I think there's an interesting journey that many brands will need to go onto to think about how their websites become service systems. Like, so if I'm, I don't know, the dentist's website needs to do more than just talk about teeth. It needs to have a queryable calendar, it needs to have structured pricing data.

[00:25:52] Jono: It needs to, let me ask questions directly to a dentist, blah, blah, whatever, like, and then you start to go, okay, my website, the, your human interface layer, just colors and pictures and text. But there is a much more arguably important

[00:26:04] Sani: All those things are things that they, dentists leave off of their websites for whatever reason 

[00:26:08] Jono: Yep. Yep. 

[00:26:10] Sani: calendar maybe, but, uh, pricing, it's never there talking 

[00:26:13] Jono: Because they wanna drag them across the threshold first. Right. There's, there's still still the assumption of like, fill out the lead gen form and then we'll phone you back. That is so dead.

[00:26:20] Sani: We'll send you the PDF.

[00:26:22] Jono: Yeah. Yeah. Well, this kills PDFs, right? That's exciting.

[00:26:25] Jono: No more gated white paper. 

[00:26:27] Sani: completely 

[00:26:27] Jono: Yeah. Great. Great. Good. 

[00:26:29] Sani: a good point. Unless the agent is going to ingest and, and, and explain 

[00:26:32] Jono: I'm happy. 

[00:26:33] Sani: and then, then, then it doesn't make sense.

[00:26:35] Jono: That maybe that finally kills Salesforce and HubSpot as the kind of defacto CRMs of choice. 'cause they're all

[00:26:39] Jono: built on the assumption that I want to do lead gen white papers.

[00:26:43] Sani: That would be, uh, that would be an interesting future. I, I'm gonna say, I'm not gonna say it's better or worse, but that would be an interesting future and I'm leaning towards one of those two. Better or worse. Um, MCPI agree with you. This is nowhere near, even though Google has now jumped on it and enabled for its services like 10 days ago, uh, seven days ago, at this stage, Mt.

[00:27:03] Sani: P is what? 1-year-old? 

[00:27:04] Jono: Yeah. 

[00:27:05] Sani: Something crazy. HTTP when he was 1-year-old made less sense than MCP does at one. I, I'll just say that. So it has, it has more of a future than people realize because what we have right now absolutely made no sense. When it was started, it was just, this is a book, put it on a screen, and, and that's why we have websites.

[00:27:25] Jono: absolutely.

[00:27:26] Sani: If, if history plays out 10 more times, I guarantee you like two or three times we get into this HTT P page based internet kind of thing, and eight times we find a better way.

[00:27:36] Jono: Yeah, I agree completely. There's so

[00:27:38] Jono: many weird things about it, like the fact that it's inherently vertical rather than horizontal. The fact that single concepts live on one URL and that we don't scroll, sign, I, who knows? Like Yeah.

[00:27:48] Sani: Design of the homepage. That looks like a magazine, homepage, 

[00:27:51] Jono: Yep. I think a lot of that is Google's influence and

[00:27:53] Jono: like the imprinting of their version of the web and like the, that indirectly kilt flash, for example,

[00:27:59] Jono: which might have been a different Yeah, there's different histories, different Definitely have very different webs.

[00:28:04] Jono: I like that. 

[00:28:04] Sani: absolutely. So one, let's talk about, uh, security of those things because as everyone says, M ds in MCP stands for 

[00:28:12] Jono: yeah. Silent. Yep. Absolutely.

[00:28:14] Sani: we don't have it. So do you think, do you think who will be responsible for security with these systems?

[00:28:21] Jono: Yeah, no one

[00:28:22] Jono: right, for the same reasons that no one one's responsible for performance or accessibility,

[00:28:25] Jono: RCO. And I think certainly, so I, I play a lot in the WordPress ecosystem and there's MCP stuff coming into

[00:28:32] Jono: WordPress core and all the plugins are jumping on it. And then I think the number of websites I see where somebody has a plugin that allows them to do invoicing and they've forgotten to go into the config and said,

[00:28:43] Jono: please make the invoice pages private.

[00:28:45] Jono: And now you go, okay, not just.

[00:28:46] Jono: Are they currently discovered, but now they're gonna be actively promoted and you have systems saying, come look at my stuff. So yeah, people are not prepared for locking down their current stuff, nevermind their next generation.

[00:28:57] Sani: Oh, that's, that. That should, because if, when WordPress enables MCP by default, and I'm guessing they will at some point, maybe it's gonna be an opt-in for, for the first few months, years, but all those should have been private pages become not exposed, but offered.

[00:29:14] Jono: Yep. And it's not just that it's, um, nobody has ever really meaningfully configured should my RSS feeds be indexable? And why is this login, why does this empty page exist? Right? Nobody is doing basic housekeeping at a URL by URL level currently,

[00:29:31] Jono: and that causes crawling, indexing security problems at an insane scale.

[00:29:35] Jono: And now we go, okay, we're gonna flip that paradigm and start pushing all of this into the night. That's,

[00:29:39] Jono: that's huge. Gonna be fun. 

[00:29:41] Sani: be so bad. Uh, another next term, upstream engineering.

[00:29:47] Jono: Hmm.

[00:29:48] Sani: that because I'm a website optimizer, have been one since 2009. I've been building and optimizing. I've spent time in WordPress as well. Not as much as you have, but also websites is what I have always done. For you as well.

[00:30:01] Sani: It's mostly about web, I mean technical, S-E-O-C-R-O. It's, it's mostly about websites. So you talked about the optimization efforts, shifting from optimizing the page to optimizing, to graph and optimizing. Just moving it upstream basically. I mean that's, that's the best way to describe it, I guess. How did it work for someone who, who for not some, for thousands and thousands of people who are building and optimizing websites.

[00:30:29] Sani: What does this mean?

[00:30:31] Jono: Yeah, this is gonna be super difficult because it's gonna require people to step out of the

[00:30:35] Jono: boxes that have been drawn around. Like if your job is currently to optimize or market a website, you're gonna see diminishing returns imminently because you've got fewer traffic. 'cause that's, if we're going around, systems aren't recommending you, et cetera, et cetera.

[00:30:50] Jono: And it's at some point your role is gonna have to shift or somebody else's role to go. We need, if we want to continue to sell our widgets. We need to actively manage our reputation on, I don't know, Quora. And we need to actively consider which journalists we wine and dine in order to influence whether they write about us in the future. We need to go and tidy up that 2013 Christmas sub-domain that nobody's thought about in years. We need to sort blah, blah, blah, blah, blah. And this list is gonna be very, very long and it's gonna. Extend to the entire web through all of time and managing that and everything that everyone says, we, we think that we need to run billboards to

[00:31:31] Jono: increase salience. None of these are our jobs today. They're diff at at best. They're different people's jobs and different teams and different budgets are worse than nobody's jobs. I think starting to have those difficult conversations now is gonna be critical because as brands start to move on this, the landscape will change very, very quickly.

[00:31:49] Jono: So first move advantage will definitely be a

[00:31:50] Jono: thing. So yeah, being brave enough to say. Waiting to try and manipulate people once they've crossed the threshold is too late. We need to instead try and influence how the machine perceives us. And that's not, that's not prompt tracking, which is what everyone is doing at the moment.

[00:32:07] Jono: Like when somebody types this thing and

[00:32:08] Jono: it's chat, GPD, do we show up? Sure there is a place for that, but it's far smaller than

[00:32:13] Jono: I think 

[00:32:13] Sani: the most obvious thing to do today, and that's why people are going that way. And I'm not blaming them 

[00:32:19] Jono: Yeah, it's copy paste the current modality of rank tracking into a new thing. It doesn't really fit, but it's better than nothing.

[00:32:25] Jono: Um, yeah, I think, 

[00:32:26] Sani: Uh, I dunno if, if K-pop has demon hunters have reached your home household, but here it's a big thing. Just imagine instead of spending millions of dollars on Google ads for your soda brand, if you just were able to place that soda and, and the soda pop is your brand, 

[00:32:43] Jono: Yes. 

[00:32:44] Sani: you never need to spend any more money on 

[00:32:46] Jono: And now what we're talking about is advertising and marketing,

[00:32:48] Jono: Which is what we should have been talking about all along. 

[00:32:50] Sani: been talking all along. 

[00:32:51] Jono: Um, and, and, and it's still digital, right? You're still trying to do it in that

[00:32:55] Jono: ecosystem in order to affect how things

[00:32:57] Jono: consumer perceive and the blah, blah, blah, blah, blah, blah.

[00:32:58] Jono: But yes, but, uh, we've, we've constrained ourselves so much. I get very irritated that the SEO ecosystem for this, because we have spent 20 years convincing the C-suite that the world of SEO is to tweak meta

[00:33:09] Jono: tags and get links and publish blog posts, and that's, we've defined that, right? We, we could just, as equally have said, our job is to influence how people perceive the brand.

[00:33:19] Sani: Uh, absolutely. Uh, I don't know. It's, uh, it's going to be a very painful change for, for maybe if you, if you're working with a website, talking to the audience now, block your website into browser two days a week. And force yourself To 

[00:33:34] Jono: do other stuff. I love that. 

[00:33:36] Sani: So maybe that's a start. Maybe that's, that's like a, a therapy session.

[00:33:39] Sani: We're not looking at our website on Fridays. We are doing other things. No one is allowed to touch it.

[00:33:44] Jono: And it's, it's not just the kind of reactive reputation

[00:33:47] Jono: management that we've looked at, like what are people saying about US West? But it's also, who's not talking about

[00:33:52] Jono: us? And I don't know, we don't sell in Germany, but that forum where somebody asks something about one of our competitors is now a weight against us, so maybe we should have a reputation, blah, blah.

[00:34:02] Jono: But like, you've gotta be on all of it and nobody is.

[00:34:05] Sani: Uh, it'll happen. And like you said, first mover advantage is going to be massive. Massive. 

[00:34:11] Jono: And brands will fail. Brands that don't move at all will just fail 'cause they won't be surfaced.

[00:34:15] Sani: That's sad, but it's true. Okay, let's talk about, uh, timelines. I love to talk about timelines. I love to talk about sliding door moments. I think we're, this is a sliding door moment for the internet and, and things will.

[00:34:26] Sani: There are many ways this can play out. I wanna, I want to ask you about the best and worst possible scenario for where we are, maybe to five years in the future.

[00:34:38] Jono: Yikes.

[00:34:38] Jono: Um, so I think the worst, the worst case is that kind of unpleasant bifurcation of the market where only unsustainable niche businesses or giants survive. Um,

[00:34:52] Sani: Hmm.

[00:34:53] Jono: and I think that's, that's not unrealistic 'cause it extrapolates quite neatly from where we are today. I'm nervous about. Also, nobody seems to be able to answer how we incentivize publishing new value that all of these systems are based on and trained on the stuff that was written yesterday and in the process, they have completely eroded anybody's incentive to do original research or journalism or to go out and blah, blah, blah.

[00:35:22] Jono: All of these things. So not only do we end up with this hole in the market, but. The ingestion of new content and data and information stops and nobody is incentivized or able to go out. I'm gonna start a newspaper. I want to go and find out what's actually happening. So there's a real, there's a real challenge with democracy and capitalism that sits away in that, which is scary. Um, and that could happen. Imminently. Right? That's already happening. The, the collapse of journalism is well documented, and a lot of it is

[00:35:51] Jono: because these models are changing. No one has an answer. Google don't have an answer. Microsoft don't have an answer. OpenAI don't, and everyone's kind of waiting for somebody to solve this.

[00:35:59] Jono: And it's, it's not ads in chat EPT, it's not a micro transactions, it's not bribing publishers to keep, none of these are good solutions, so that's

[00:36:07] Sani: I think independent research work and independent journalism does have a future. If you look at people, and of course this is a big name, but people like Ed Tron, who has built a huge following and he's a, he's an investigative journalist, that that's what he does. He does pretty well, but you need to have a following that you deserve, not just some legacy publishing business behind you.

[00:36:26] Sani: So correction 

[00:36:29] Jono: yeah, 

[00:36:29] Sani: should call this. Mm.

[00:36:31] Jono: yeah. I agree that, yeah, and that's, that's, and there's definitely a space for that kind of independently funded, crowdsourced

[00:36:36] Jono: kind of space, and that works with art and media as well. Um, but I wonder if that's enough. And I wonder how you get to that scale of following

[00:36:44] Jono: in this new web and that I I, that could all collapse as imminently as like the next few years.

[00:36:50] Jono: There's, there's a cascade effect where publishers fail and Google's inventory is now stale

[00:36:57] Jono: and Chatt starts recommending bad things and the internet just goes awful. And there's no, there's no way out that, that's scary.

[00:37:03] Sani: That is the worst possible timeline. And I agree. This is horrible because it, it's, it's collapsed. The models will also collapse. The future models of LLMs will be worse and worse and worse. And we are getting, we're looking at Wally 700 years from now and like, I mean, just in 10 years basically, what is the best possible timeline from.

[00:37:22] Jono: The best version, I think is we accidentally get a far better web and far better business

[00:37:28] Jono: where. 

[00:37:29] Sani: like that accidentally. Uh, if, if, if the best possible timeline is an accident, 

[00:37:33] Jono: It's an accident. Oops. Yeah, no, you, you, you mentioned this, you said, um, what did you say? I've already forgotten. Oh, you talked about like essentially tricks and

[00:37:41] Jono: hacks no longer working. I think that's the good timeline

[00:37:43] Jono: that we actually, we are forced. To invest in value and meaning and differentiation because doing anything less won't won't work.

[00:37:54] Jono: Right? So suddenly all of these marketers who are writing blog posts might be forced to instead say, we should take all that research and we should improve our call center customer service script. We should look at the quality of the cardboard we use in our packaging. We should consider our eco policies, blah, blah.

[00:38:11] Jono: And there's a, a vast realignment of resource and focus and activity focused on saying, how do we be the best brand or the best product, or the best solution for whatever problem, for whatever audience. And people do the hard work of that thinking in that doing instead and the world becomes a far better

[00:38:27] Sani: The problem with that is it's too much work.

[00:38:30] Jono: Yeah. It's, it's hugely difficult, but, but

[00:38:31] Jono: now and, 

[00:38:32] Sani: been like that always. I know, but now from the position of comfort where people just, you know, think I can act Chachi PT could do it for me in five minutes.

[00:38:40] Jono: Yep.

[00:38:41] Sani: They think people think it can, but that is a lot of work. Also, you said accident. This is an accidental thing that might happen to us, and then everything is fine.

[00:38:52] Sani: I think we're going towards the worst possible timeline based on everything we talked about today, and I think we're not. Uh, we are running towards the worst possible timeline and, and then things will correct. I don't know how, but the internet is about to get significantly worse. This is my opinion. I'm not putting words in your mouth significantly worse than it is today.

[00:39:10] Sani: Significantly less trustworthy. The quality of content I'm look at the lop, look at, look at SOA, and all that bullshit.

[00:39:17] Jono: Yep. Yep. Infinite amounts of just garbage.

[00:39:20] Sani: Yes. Uh, the signal to noise ratio is going to get. It's, it's already beyond terrible. It's going to get impossible to keep up with. And one of my resolutions for 2026, I'm not going to pay attention to any noise.

[00:39:34] Sani: Like I'm, I'm just doing my thing. I, I've picked my thing for 2026. I'm only working on that. I don't care about anything else. I don't want to hear about I. Cha. GPT, 5.3, 5.4. I don't C. It doesn't matter. That's the whole thing. None of this matters. I think this is all bullshit. They're pouring on us and expect us to, to react.

[00:39:52] Sani: No one cares about a new iPhone these days. Why would I care about 5.3 or 5.4 or I mean workers releases? Sure. For people involved or people working. Big deal. Yes. You're a core contributor. Absolutely. Respect to them.

[00:40:06] Jono: but no one's setting off

[00:40:07] Jono: fireworks in their 

[00:40:08] Sani: matter. It does. 

[00:40:09] Jono: Yeah. 

[00:40:09] Sani: Big. Big. Yeah. Big picture. It doesn't matter. 

[00:40:11] Jono: Yeah. 

[00:40:12] Sani: Final question for people in website optimization space, CRO, technical, SEO, uh, and any of that one piece of advice to stop them from becoming obsolete in 2026.

[00:40:25] Jono: I hate to say this 'cause it

[00:40:27] Jono: really goes against the grain of everything I've ever believed, but yeah. Um, you've, you've already said I go turn off the website or at

[00:40:33] Jono: least step away and go look at everything

[00:40:34] Jono: else and think about. Reputation and salience and column issues and surveying and talk to real humans. Um, yeah, do the actual market stuff, and that's a, is super alien and it's super difficult. And I, yeah, as a, as a tech SEO nerd, I find that super intimidating,

[00:40:53] Jono: but that's the thing that's gonna matter. Um, yeah. The role and impact of your, your words on your URLs definitely continues to diminish.

[00:41:01] Sani: There's life outside of the browser that, that, that's, that's the final 

[00:41:04] Jono: That's, that's the t-shirt. I love it.

[00:41:07] Sani: Jonah, thank you so much. What is the best way for people to connect with you?

[00:41:11] Jono: I think probably LinkedIn nowadays or just directly by my [email protected].

[00:41:16] Sani: Perfect. Uh, to everyone. Uh, we wish you a happy 2026. This will 

[00:41:20] Jono: Yeah. Good luck. 

[00:41:22] Sani: and, uh, please consider rating, reviewing and, and sharing this episode with a friend, and I'll talk to you next week.

[00:41:28] Jono: See you later. 

[00:41:28] ​