# Schema Is Useless for SEO (and Probably for AI Too)

Channel: Edward Sturm
Video: https://www.youtube.com/watch?v=DuXibA5y2Ng
Duration: 49 min
Language: English
Words: 9194
Transcript page: https://viewrankai.com/tools/youtube-transcript/DuXibA5y2Ng

---

[0:00] We are talking about schema and if you need it to rank or to show up in LLMs and I am joined by two of my favorite people in search engine optimization, Mr. David Quaid and Mr. Jay Kundley, two people that I love and really respect and uh this is going to be a very fun discussion. I don't know who should I start with on just like the basic question. Do you need schema to rank and show up in large language models? I think David's the better person to start

[0:28] with on that one. It's It's an absolute no. It's It's not even like 1% for me really. Um I Sorry. Go ahead. I I I didn't want to cut in. I for for for so many reasons, right? And and I think because of my um you know part of my role in the wider SEO community is is helping establish what works in SEO and and what are myths and how search engines work. 10 years ago 12 years ago I published an article on Kora that got that is now 1.2 million 1.5 million views and how Google works. And I've had a couple of Googlers actually

[1:05] say wow this is this is insanely close. That was an infrastructure level uh because I was working in an infrastructure company. But the whole idea that that search engines can build a relevance first system and then chuck it all out to focus on scheme pages with schema makes no sense and LLMs don't have those indices. So it just it's it's a it's a nonsensical non-starter. And then when you look at the the two biggest data pools, Mark William Cooks and SEMrush, it gets even better, right? So Mark Williams Cook is like there's no way LLM's process schema

[1:42] and SEMR who should be doing better. They should be following ARFs. I think ARFs do a really good job at presenting data, doing research. Seamrush are like how many paragraphs do you have? What's the paragraph ranking correlation statistics, right? These rubbish reports, pseudo science reports that they produce. One of the reports was uh substantiating schema and saying that out of pages that ranked x x percentage of of pages had schema but these were schema like article schema right uh this is just a correlation um of pages that ranked for faqs and had faq schema it's not what made it rank and I think you know I can't see

[2:23] anything that supports schema helps you rank in LLM because I can't see it helping I can't see how it makes it relevant or better. Um, it's not a trust, right? Like I I think some of a lot of the myths that I read online there there are a couple of angles. There are a couple of companies who have sort of like um started up as like schema SEO companies or or AEO companies. It's very common in AEO myths and they're like, well, schema helps Google understand the web. It it it does to a certain degree, like for flights, hotels, um it could help with click-through rate if you've got a your

[3:02] your review schema rating, but you still have to rank, right? You still have to get into the top five. You have to be visible for that to happen. If it's going to be hotels and you want to feature in a hotel search, you're going to need hotel schema. But for everything else like blog post, standard content, SAS, it's just not going to make any

[3:21] difference. I so I don't think it does anything for Google uh a Google search standard. Um I don't know LLM to be honest. Um I would say that I wouldn't be surprised if LLMs do use it to to aggregate information. And the reason why is because I mean Edward when I was on your show last time I was explaining we were both talking about how LMS are basically SEO in 2010 and like everything is just super uh

[3:55] outdated and old in terms of practices. So it's like really easy well at least in some capacities to kind of do like the old black hat SEO tactics to show up. Um and uh so I I was actually just I just had a um like an hour and a half long conversation with David Hunter from Local Falcon um because we are gearing up to do another massive study um on this. So we did the the Google business profile geotagging XF data study showed geotagging did absolutely nothing. And then I brought David Hunter uh onto our podcast to to kind of uh talk more about Local Falcon and and his business as

[4:43] well because it was used in the study. And I had mentioned the next thing I wanted to do was tackle schema. And so he brought up that he doesn't he also doesn't think that it does anything with um with with Google search at all, but he is a strong believer that uh schema affects LLMs, at least in its current state. Um I honestly don't care whether it does or not. Um but we're going to find out. So, uh, actually yesterday I just had one-on-one with Eva, who, um, was kind of my assistant in the whole the initial study and we just started mapping out the whole

[5:21] methodology and, uh, and everything. We're going to start, we're going to track and, um, kind of monitor to see whether schema affects one, Google, but two, LLMs, and we're going to track uh, the the big five uh, LLMs. So, um, I'm I'm excited for that to to get rolling. I'd be interested in the methodology behind that. Um, because if you look at how LLMs work, they don't have a copy of the worldwide web, right? So, they need

[5:54] a search engine to facilitate that. This method of marketing is so effective, I had to make sure it wasn't against Google's rules before I kept doing it. It's a form of SEO I call compact keywords. Whereas most SEO focuses on putting up articles to answer questions how, what, when, compact keywords focuses on putting up dozens of pages that sell to searchers who are actually looking to buy. These pages rank on Google and convert so much better than normal that when I discovered this years ago, I couldn't believe this was allowed. It's less work, too. The average compact keywords page is only 415 words. Compact Keywords is a 13-hour deep course on getting

[6:34] sales with SEO. A customer recently said, "Each lesson is dense with information. You're giving years worth of experience boiled down into 15 to 30 minute lessons with no filler or fluff. I feel like I'm gaining a new superpower. Compact Keywords is about setting up an SEO funnel that brings you sales for years and years and years. It works with AI. It's less work than traditional SEO and it makes way more money. You can get it now at

[7:02] compactkeywords.com. Back to the podcast. If you look at how LLMs work, if you look at um how the training works, they develop heristics, that's not the same as sort of like having a copy of the whole worldwide web with them. And so for a lot of data, they've got to go to a search engine. And in order to get schema, they would have to have some sort of processor read the schema in the HTML and deliver that differently in during the crawling process, which would further slow it down. What they seem to do is just grab the document, grab the HTML text and feed it straight. So, for

[7:34] example, they're not good at uh client side rendering content, right? Because they can't wait. They don't have a Chromium bot like Google does. The second thing is that LLMs are machines at managing content at bulk. Right? You can take a statement and say, I turn this into a 500page thesis and it'll turn it into a 500page thesis. And then you can say, you know what, I want to run a Google ad and give this way as an ebook, condense it down to 35 characters, and it'll condense it down to 35 characters, right? And if you give it a page of information, it can turn that into tabulated information in

[8:11] seconds, faster than um search, sorry, Google Sheets or Excel can, right? Um there's just no benefit for schema in in general purpose. I also don't see um I I can see instances where there's specific searches set up [clears throat] like it wants to get movie times or flight data. that schema can make sense, but I just don't see the benefit. I I think it's up there with LLMs or text. It's up there with most SEO myths where [clears throat] people are being told this is the secret that SEO experts don't want you to have, right? Um, and I've engaged so much with people on Reddit who they're entire subs on on

[8:52] Reddit that are um set up to platform geo disinformation, right? And there are people going, "Oh, here's my LLM's text uh builder." And so, how did you how did you come across that? How did you design the architecture given there's not and you start, you know, interrogating and and applying critical thought and they're like, "Oh, it helps LLM understand your website." Again, playing into the myth that LLMs crawl the whole

[9:15] web and build their own index, right? Which is just nonsense. And you would need to do that in order for schema to help, right? You can't at the last minute look at all of the uh results of a query fan out and then exclude pages for having schema or not. It just doesn't make any sense. That's so that's like my problem with schema and all of these myths is like

[9:36] why how does it add value to the system? Um and the push back I always get is well what harm does it do? Um why not just do it? It makes sense to do it. And it's like, well, I I don't like magic in software systems, right? Because that's when people start waiting. They go, oh, I added schema to my page. And you know, on most of the SEO subreddits every day, you'll get three to five people coming in. I have the fastest website. I've put JSON schema everywhere. I have organization schema, which is pointless because you're not an entity. I have personal schema, pointless because

[10:10] you're not an entity. Um, I have product schema, pointless cuz Google doesn't care. and um their website's not ranking and their their website won't index just their homepage. And it's like because they're buying into things that they can they can reach out and touch. Whereas things like developing topical authority, working out how to get back links, how to promote your website online. And I know that these are old parts of SEO. I know that these go back to like 1998 when I was first doing SEO in in in at when I was at DAL building up my own sites. Um, it might be old, but that's just unfortunately how it

[10:46] works. I don't think we can change SEO uh because we think it's outdated. Um, which is something I get all the time. You know, if I show a video, people are like, "Oh, that video is 10 years old. So what? It's how it works." Are there any types of of basic schema that all websites need or no? I don't think unless you're a known entity or in a specific niche, I think for e-commerce, it can make a lot of

[11:12] sense. um if you're in e-commerce and um agentic browsing is going to become more important I can see benefit to having some schema there but I I don't know enough about it I don't have enough practical experience like product like product schema um but like but like if you're if yeah if you're a SAS if you're a SAS how about that no absolutely none I don't think it creates any the the second thing that I've noticed about people who talk about schema is they're like, "Oh, it it Google understands helps Google understand your content." It really

[11:48] doesn't. Google doesn't understand it. We know that LLMs aren't really research tools, right? From again, the king of SEO fun experiment that I did, which some people still take seriously, but um what that does is it it teaches that they don't they're not capable of going back and and and re recalling the web to find out if that's true or not. And and

[12:08] who knows, maybe one day they will be. So I don't think if adding schema for 99.9% of cases helps. And I also I'm one of those people who I've worked on projects for the last 5 years. I've started four projects in the last 60 days. I have some websites with two or three thousand SER features, right? Like we just decided we wanted to go up after search features. We decided that in some cyber security areas we wanted to educate people and we've built six or seven different um FAQ types, right? We showed this when we were looking at the uh PAA um case study that we did. Um

[12:46] I never use FAQ schema and everyone says it's mandatory and I think part of it comes from web dev saying look this is a great way to structure data. This is how I would do it as a code. If I was a coder and I wanted to scrape content from a page schema would be much easier than scraping text. But if you're an LLM scraping the text would actually be much easier because it's it's native to how you work. Um, and I think also there's a slight bit of like texty creeping in where they're like, "Yeah, your average guy can't do schema." And honestly, your average guy can. You can go to an LLM

[13:18] and it'll build a schema for you. It's not rocket science, right? And so it doesn't actually mean that if Google sees schema on a page, oh, you know, I can trust this. You schema is as corruptible as HTML, right? How about local? Um, again, I I I've tested uh schema on local and and and and done nothing. Um, I found very little apart from physical location and reviews to be good on local. And that's just playing with a couple of local pages. Um, but unless you're a known entity, I can't see it helping in local. Um, I think I think SEO is is made up of complicated

[14:04] sounding topics like topical authority, page rank, dampening effect. It's boring. It's been there forever, but those are the things you have to master right now. Yeah, but you said uh 99% of the time schema doesn't impact anything. What's the 1%. So, I think things like um where where there's a specific search like a flight search, right? And Google needs to grab things like it could be very difficult if you start writing like okay the next the next plane out in New York is at 4:15 and it's flying to SNN and it'll land and you start having different time formats right like if you mix 24-hour clock and 12-hour clocks in the same

[14:42] text that's going to be very difficult for even a human to extrapolate data from. And so I think schema makes a lot of sense there. So, if you're working on airport SEO, it's it's pretty important. But I I'm pretty sure that most of the airlines already export an XML feed anyway. Well, apparently in the knowledge panel, too, because I think in one of your latest videos on here, uh you you you weren't sure if the topical authority of your own name versus the the schema markup on your own personal brand um impacted your

[15:12] knowledge panel. And thanks so much for bringing that up. That's actually perfect. Uh cuz I think since the number of searches for my name dropped, my knowledge panel disappeared and the schema didn't disappear. So I think um unless we can get searches for my name up, which Okay, so everyone just search for me, please. You should Well, you should delete the

[15:33] schema. That's actually true. So what I'll do is I'll delete the schema and then we can come on and we can like beg everyone to search for me. Um and then we can see if that works. But yeah, that that's a good question. Um, there's a couple things I wanted to talk about too regarding SC regarding schema. Um, especially I mean it's nice that you don't have a lot to say about lo like local business schema or local organization schema because that's what this this whole thing is going to be about and we're going to really see what's going on uh in terms of ranking or in terms of um just popping up

[16:06] because the other thing is just is rich snippets is um the other argument that people can bring up is okay does the uh the imply the implication is that well it might not help ranking but the because you have rich snippets, it can help clickthrough rates which can then um help you know your rank that way. Um I'm I'm personally of the belief that a lot of a lot of the schema that people are really promoting um it's going to add more complexities to kind of the health and safety of your website and and two main things is like review schema like Google updated guidelines in

[16:47] 2019 about review schema. So, if the business owns the review, you're not allowed to mark that up. Um, technically you can. Um, their actual words are saying we're just not going to show we're not going to show the star rating in the rich snippet. Um, but it is part of their guidelines. And they have like the the biggest red flag, the biggest red flag on the especially the review snippet guidelines right at the top of the page. Warning. If your site violates one or more of these guidelines, then Google may take manual action against it. Once you've remedied the problem, you can submit your site for

[17:20] reconsideration. And so, like, if you're just throwing schema on the website for the sake of throwing schema, especially review snippets, and you're and you're a review schema, and your website or the the company owns that review, that's that's against the guidelines. Um, it's it's strictly pointing it out. The other thing with like FAQ schema, too, in 2023, I did a little bit of research before this. Can you tell? Um, but in 2023, Google updated to only have FAQ page rich results limited to government and health websites. So, if you're like not within those those industries, then the FAQ schema is basically pointless in terms of rich

[18:06] snippets. So, like there's a lot of people just like tout and those are the main two, right? There's it's like FAQ and it's like reviews and like those are like the main ones. Obviously there are different schema types like like article type and like there's organizations like the ones that I plan on testing have a document for this um we have um has certification so in the landscaping and lawn care industry like uh pest applicator license things like that area served location branch like like branch um founder owner foundate um the other ones that I really like is the nakes code and the ISIC um code. So to

[18:48] actually indicating uh in the schema what the actual industry is. There's another one in there called like uh I don't know if it's like organization type or something like that but it's uh there isn't one for landscaping and lawn care. I think the closest one is like professional services. So if we have if if we test the nakes and the ISIC then um we can uh hopefully that'll be some indicator of uh you know the actual industry of someone searching landscaper near me but um um no I mean you were curious about the methodology too so uh

[19:22] do we want to break that down? Well I want to so I think you brought up an amazing so many amazing points. I just want to go back to one which I think is really interesting. Right. So, uh, let's go back to reviews for a quick second because a lot of people have had reviews removed. Uh, Barry Schwarz updated there a few weeks ago saying that Google was removing tens of thousand reviews. I wonder how much of that is like Google getting um local business owners to tell report on each other, right? Because there's obviously there's a lot of competition, but what you're saying there, and I completely

[19:52] missed those two points. Um, the first one is the um that you can't show review criteria if you own the review collection, right? So, do you see people potentially using that as a negative SEO vector to report thy neighbor um for bad reviews for for misusing reviews? I don't think people are reporting like Google I think I think it's more geared towards Google business profiles. I could be wrong on this because I'm I'm hyperfocused on the local industry or the local

[20:18] That's what I mean in local. Yeah. Yeah. I don't think it has to do with schema markup on the website. I think it's more to do with like fraudulent fivestar reviews on Google business profiles. Um, the the problem is is like if Google sees if if Google can recognize that you're a local service business and they see a review on your website and you have schema marked up on that, the first thing they're going to do according to

[20:40] their guidelines is just not show that. They're just going to disregard that completely. Um, what scares me is that big red banner across the guidelines that says you can get a manual action if you violate the guidelines. So, um, technically you could make an argument to Google and say like, well, you're just saying you're not going to show it. You're not saying I can't do it. But

[20:59] why do it at all then? That's a good point. Um, it's it's like one of those like really heavy-handed like police yourselves, but if you don't, we're just going to wipe you out. Yeah. So, we will not be testing the review one. Right. Right. This is going to be on our own clients, so um I'm going to pass on that one. I mean, I'm happy. I I love setting up domain names for testing. So, um maybe maybe we can set up a a fake local business in Florida. Um see if we can get the snippet to show up

[21:28] in search console. Yeah. Yeah. Yeah. Yeah. That's exactly up my street. I'd love to test that. Yeah. Uh but I think so. Jake, you're doing schema for your local clients? No, in the test we are. Yeah. In the test you are. But yeah, but but yeah, for your local clients, you're not doing schema. And what are your rankings like for your local clients without

[21:49] schema? Uh I'll say this, we have a 95% retention rate month over month. So rankings, I don't really care. Like as long as our clients are getting leads and they're profitable, we're good. Now, um but I will talk about rankings. Um I mean rankings are good. It's just a weird question honestly. Like no no offense to you, but people are like you know you know how's how are your like all the sites No, no, the whole point the whole point to this this episode is like, do you need schema to rank and do you need schema to show in LLM? And so the question is like,

[22:20] okay, you're not using schema. So for people wondering if you need schema to rank. Yeah. This goes back to like ever since you you brought the termin the term up, the first time I ever heard it was paro SEO. Ever since you brought that up, I've been saying that like crazy. I made that. I made that up. Did you? Okay, cool. Cuz I was tell I was talking to somebody about it and I was like, I don't know if you made the

[22:38] term up or not, but it's good. Made it up. Yeah. Thank [laughter] you. I appreciate that. I've been using that a lot and in the schema thing like if anything it just falls in the last 20% and even that to David's point falls in the last 1%. Um so uh no I mean they're all ranking uh fine I guess you know uh we're we're doing well without it and you know really it's it's focusing on can we create additional pages with H1s and page titles that match the queries that we want to rank for. Um, and that goes back to the project pages I was talking

[23:09] about in your last episode. Um, so yeah, they're doing fine. In fact, I think the only schema any of our sites actually have is breadcrumb and that's it. And that's just because it's autogenerated from I don't know if it's autogenerated from the CMS or if it's autogenerated from Yoast, the SEO plugin. That's that's part of the uh the methodology is figuring out where it's being generated from and then removing

[23:33] it. We use breadcrumb schema as well because I think it looks good in Google results. But um we've never had a problem with ranking. I think that's why I care so little about schema. If I thought it could work, I would use it. And I these these debates I get into in in Reddit. You you can tell that the person has limited knowledge and experience because they quickly turn into like attacks, right? So it's like people quickly turn to ad homins, right? Like that's when

[24:02] you know someone's losing an argument. And it's like, look, we're here to debate whether something in SEO works. If you don't like the other person's point of view, that's fine. But attacking the other person doesn't really help. And I feel like you're doing that out of a position of weakness. But um when you look at if I was if I was to look at the projects that we work on, we work on helping companies that are like typically pre- series A up to pre- series C series C investments. Turning companies into 100,

[24:29] 200, $300 million companies and bigger. We aren't going to hang around if we can't deliver results, right? The they've got SDR teams that are paid for. I think another misconception about VC companies is that they have billions and billions of dollars to spend on advertising and marketing. They don't. VCs want to see the lowest cost of acquisition possible. They don't want to see a heavy expensive product that takes a lot of money to market. They want to see ease. They want to see people going to your website, ringing up, and

[24:58] spending $100,000 in seconds. Right? That's what that's their goal with their with their clients with sorry with their with their portfolio. If we don't deliver those leads, we're out. So, we have to deliver on all kinds of things. We can't take risks with buying back links and we the PR that we do doesn't help with backlinks as much as people think. So it's it's it's it's really one of the toughest forms of SEO because we're up against Microsoft, Google, Amazon, Palo Alto, Citrix if you know all of these big tech companies that have you know DA99 sites and things like that. And we've got to eek away at

[25:34] little bits and I think that's what's made me a good SEO to have that sort of like point of view and experience to talk about because I've I've done that before and if believe me if schema helped I would deploy it straight away. It's it's a it's so easy to deploy, but I can't rely on magic and fingers crossed mentality. I've got to deliver

[25:55] or I'm out. It's just that simple. Can we talk about that uh Semrush article you sent me? Yeah, absolutely. So, the David, you sent me a Semrush article and the title of it is uh how do technical SEO factors impact AI search? And in there it talked a lot about uh how I mean there's a whole section on schema. I mean there's multiple sections on schema on there. And it basically just made a a big old correlation about how sites with uh schema tend to show up

[26:29] more in LLMs. And my initial reaction to this was like this is just and I I have a problem with all of these articles that always start with like you know hey we analyzed 300,000 websites and it's like well I mean you aggregated data and then you came to a conclusion based off of a

[26:48] correlation. Absolutely. And so um yeah so I wanted to get your initial thoughts on that because I think you sent that to me that came in hot. Yeah. I I and and I I dislike so much for doing these as well. Right. Like it's sort of like we're looking at the at the pages that made it. So, and I think we talked about this before privately is like uh why don't we start counting commas and see how many commas correlate with people who rank first as well. It's it just gets pointless. Like you just said earlier that FAQ schema is really only um applicable for government

[27:18] and healthcare, right? But if you look at the types of questions that people ask and the types of content that's surfaced and also the easiest type of content for SEOs to surface is typically FAQ content and depending on whether they add the schema themselves or their web team add the schema because they think that's the right thing to do which is a good enough reason. Um you're going to see FAQ schema depending on the types of data that they analyze. You're going to see FAQ schema present there. Um, I have I think my team develops around 150 FAQs a month. And we, you know, it's it's a great way to get in under the

[27:57] skin of these big brands. And so we, if we never use schema and we're always getting there, then it can't be a prerequisite, right? It can't be mandatory. And I think the exact same as you do, the these the survey by by SEMrush is just another first pass the post analysis, right? It's just going to you're going to bring in that schema that they're you're going to see that in

[28:20] those pages that are ranking. Well, the other interesting thing too is that one of the schemas that they brought up was um FAQ and review snippet which in most cases aren't even shown in search results. So like I don't even know what to extent they're being ignored. Now like I I this comes up this comes up in the argument about like um uh like like exact match domains too which is like there's typically a a correlation versus a causation in terms of in terms of rank. That's a whole another topic too. But the um with Semrush's article here I think that's 100% correlation. Like if people are

[28:57] adding schema to the site chances are they're probably doing other SEO things too. And so you can't just say that because schema that is up that means rank is up too. And that's that's the whole point of what our test is supposed to solve. I think with the review schema as well is it can like metad descriptions. It can only help improve your click-through rate if you're visible. So how did you get to be visible um

[29:23] if it's if it's not the schema, right? So you can't just be have schema and be on page nine if nobody's getting to page nine to see it. So it it might help in the top three, but you still need something else to get you there, which means it wasn't the schema that got you there. So it's it it's just a correlation thing. I mean, we might as well count like what what t-shirt was their SEO wearing and if most their SEOs were wearing branded merch, then maybe you should wear branded merch when you're doing your SEO. Uh because that could be a huge

[29:54] correlating factor, right? Can we read this? Can we read this comment on your YouTube video? Yeah, I I'll I'll read it. Okay. So So somebody said, "This is why we're actually having this podcast in the first place." And it's because I get comments like this all the time cuz like I make I make lots of episodes about how you don't need schema. You don't really need to worry about it. It's like the

[30:18] last thing that you need to worry about. So somebody said, "I'm not going to say this person's name. I don't want to throw hate, but the person said, "I would love to come on your podcast and debate this. Show me one site ranking without schema. False stuff." You said expensive. It's not expensive to set up.

[30:33] I can do a large site in about 6 hours. One, I just want to say you got to maintain it. And and two, um, lots of agencies will charge a lot of money for schema even if it is not expensive to set up. All right. The person said, "Event snippets. event snippets worked great for food service at the time they worked. Well, oh, you're admitting they don't [laughter] like the he I mean he you know he com I

[31:02] don't know if it's a he could be a she. person commented on a a schema on a schema video I said where lots of schema types were getting deprecated like so yeah and uh staying up to date with what works and what has stopped working has always been the protocol in SEO which I kind of disagree with I think like a lot of like what works is what's always worked um I mean like you know both both of you what the both the discussions that I constantly have with with you David and then Jake when you came on the show and we talked for two and a half

[31:36] hours a lot of it was it was the SEO fundamentals. In fact, someone sent me it was Rand Fishkin saying uh like tw it was saying like 15 years ago basically talking about entity SEO before the term entity SEO even exist existed. The same stuff that that geo people now are saying you need to do. Also the whole geo thing like it's so funny because these geo these like geo people will say LLM are so smart AI is so smart agents are so smart they can do everything and you need to put in schema or else they won't understand your site then how are they understanding sites

[32:16] without it right because they are they're they're brilliant at it um also and you know I really wonder if the people who say you need schema like if they actually rank themselves because like when you rank yourself and you see you you just see you don't need there are a couple of people that I talked to on X and I don't want to name any names but um you know some person was throwing a lot of like questions about you know some a lot of the activity on Reddit for example in the different SEO subreddit saying that why are people talking about things that are 10 15 years old and then saying you know

[32:50] tools like BERT have come along and changed SEO and I'm like they've changed things for the user right so we're talking at semantic keywords. Um, does Google understand different variations of keywords? And I actually was on a little mini podcast with Dave um Faro who was on with you recently. It was a great excellent show and we he has a tool um automation tool in using Claude that will take two different queries, run them and see what the overlap rate between results are so that you can tell does Google treat a keyword semantically or not. Now, yes, if you're a very high authority site, you could probably just

[33:26] use one instance of the keyword and still rank, but is there to help the user understand if like DR means domain rating or Dominican Republic, right? It looks at words around it. That's not necessarily helping the SEO out. That doesn't mean that as an SEO you can just put in one version of the word and you may or may not rank because Google's using LM. And I think this is happening a lot as well that just because Google changes something or does something for the user doesn't mean that the fundamentals in SEO have changed. And um you the reason you're going to see old SEO in places like Reddit is because

[34:02] that's what works. And I I said the same about what I loved about Charles Flo, Jackie Chow um and um Julian Goldie is that they don't have any debate about SEO. They've been doing the same SEO for the last as long as they've been doing SEO and it hasn't changed and they're minting it. So if they're not worried about the schema debate and the LLM's text debate, why are other people worrying about it? And I think that's actually the crux of the problem. I think the people that are worrying about it are still looking for this magic

[34:35] secret. Right. For what it's worth, I I went I went you said um I think Edward you said you know are are they even ranking themselves? And um I went to their website and I typed in exactly what their H1 is on their homepage. And I know this isn't the way to like truly test ranking, but like they showed up at the bottom of page three and and they're a local based agency in their area too. So like I like there's even a geog like a geographic tag to put in that

[35:07] keyword and yeah bottom of a page three. So um I mean schema is not helping them. So, also to the point like staying up to date with what works and what has stopped working has always been the protocol in SEO. That's actually not true. That's a way more expensive and uh and cost time-c costly way to do search engine optimization. That's not what you want to do. You want to make pages that just stay ranking. That's what you want to do. And like if if you are constantly doing like playing with short-term tactics that are going to stop working, it's going to cost a lot of time and

[35:44] money. Yeah. Every every tactic that's that used to work for Google that's working for LLMs now will eventually stop working. Think like everybody everybody's kind of like hitting on those black hat I mean I just saw like keyword cloaking was was uh influencing ML uh LLMs. Um and this is a this is a while back. I can't remember the study that I saw that showed this, but I'm like, this is like old stuff. I'm like,

[36:08] eventually it's going to stop, right? They're going to they're going to finally go through the Google phases of being like, oh, this isn't like people are using LLMs less because the information we're providing isn't as good, so we have to get better at it. And then like what determines quality and and then it's full circle back to what people actually want. I mean are people so just where where I don't know where you saw that claim but I I see people making a lot of claims like oh we analyzed this and you know they didn't analyze anything there's no data to share um saying what one of the most

[36:39] common ones I see for geo is like um it's not like old school SEO where um keyword density matters anymore. Keyword stuffing doesn't matter anymore. And I'm like I can't remember the last time it did matter. It's not it's just not a good way to do SEO. So I I wonder how many people are making that up, saying it it works versus it doesn't. I think a lot of go and AEO is playing towards people who don't want to believe in SEO or don't or or think that SEO has changed or SEO must change uh versus like um you know if if the last debate I

[37:14] had was about freshness, right? And does freshness play a role? If if freshness played a role, we would constantly be editing um our ranking. So what if you look at the sort of like conveyor belt of SEO work once you take positions you start moving on to the other keywords and get them into position and then you start testing do you get leads and business from that if freshness was a factor then we would have to constantly revise the pages that are ranking and that's just something I've never had to do. So again, I think that the people that buy into these myths are still looking for unicorn ways out. Like, oh,

[37:48] there must be a magic switch or there must be something hidden in the code or just something easier than actually having to do SEO. And I think that's what separates the successful from the not successful. You know, that that quest for the unicorn in SEO. Um, it might work. LLM text might give me that advantage without thinking about like why or how or how come

[38:10] How come other sites aren't doing it? Yes, Google has an LLM.txt, but how how often do you search for the queries that you're trying to rank for and see who's above you and why they're above you? Do they have LLM text? Do they have schema? If they if they don't, then why bother? And if you can add scheme yourself, go for it. But like I said, we we're I'm seeing so many people coming to Reddit every day. I've got 99 to 100% scores on on on light speeded. I've got Jason schema coming out the wazoo. I've got this, that, and the other. Google's not indexing my pages. And and 90% of those

[38:44] people will come back and say, "You need a sitemap. That's why we did that sitemap podcast or I I was so adamant we do it so quickly." Right? Um and I think if you look at the number of people, nobody's come to it and said, "Yeah, you're right. I I took my site map down and my site fell." And I'm not

[39:01] saying take your site map down, right? I'm just I think I I talk to tons of agencies all the time and um ones that are are cordial and ones that are competitive and the ones that are competitive that are doing SEO I think like when you're trying to sell SEO to a business uh you have to get them to trust you that it'll work over time which can be a very hard thing to do. An easier way to do that, just from my observation, is to kind of obfiscate and make things more complicated than they seem. Um, because as soon as you start talking about more complicated and

[39:40] technical things, um, people just start to trust you more because it sounds like you know what you're talking about. Um, and so instead of like, so even if it does take a little bit while like a little while to rank, like three to six months or even 12 months, but you start talking about like, yeah, we need to add like all the schema. Uh, we need to um I'm blanking on the technical stuff like improve site speed or lm.ext, you know, like put all that up there. Yeah, the

[40:09] site maps and stuff. We got to optimize your robots.txt file. Uh, as soon as you start saying that, I think I think most business owners eyes kind of glaze over and they're just kind of like, "Oh, okay. Well, I'll let you do what you need to do. You sound smart." And then um the minute an agency uh or an SEO like myself comes in and says that is just not going to help you do anything um it's it's easier for the the other agency to kind of discredit us because we're not saying the things that

[40:45] sound technical. I I agree with that 100%. Um, I know how one way to create trust is to be able to speak about a whole lot of complicated things and sound proficient and sound that you've got it under wraps. I get that. Um, I also understand the making claims and about ranking immediately and

[41:02] doing things in 30 days versus 3 months. One of the things that I do when I have a team that's like, oh, we're used to using this structure or we're used to using that, is I try to find an SEO project that disproves that as fast as possible. So, if I have a team that's like, "We need to have a rankmath score of 95 or we don't publish." I like, "Let's let's get a page to 30 and publish." And as long as it's got what it really needs, it'll work. And then that starts to help them cuz I I want people to think independently. I don't

[41:32] want people to be replaced by an AI robot in 6 months. I want them using AI robots to help them do their job. And that means they have to know how the levers work, right? Like I use that example when you drive your car down the road, you don't have the handbrake on and it in fifth gear and your hazards on and just have every single thing turned on, right? You turn on the things that you need to get where you're going as efficiently as possible. And I think that there are a lot of things that SEOs can do that are quick wins that you can

[41:59] get there within a week, like ranking for low difficulty keywords, uh low hanging fruit. And that's how I develop trust is go look I will I'm not saying I'm going to make you number one for your core keyword but here are things that we can do that will start maybe it's using YouTube maybe it's using um other shortcuts or uh maybe we can use um your Reddit um user account we can post something there and show how that ranks in Google really quickly for example uh or I'm all about like parasite SEO with the well I If it's your own user account, is it really

[42:35] parasitic SEO? Yeah, that's true. That's true. If it's your own user account, it's it's it's not that bad. I want to I want to wrap this episode with just tangible, straightforward. Um, do you need to use schema, which we kind of covered, but like what should you use schema for, if anything? What should people actually be

[42:52] thinking about schema for, if anything? I think again if if it if it's hotels um pricing information that Google has a specific search for. In other words, you do a search and the screen changes. It's not just your standard blue links, right? Definitely. If it's something like you've got a stock ETF system that Perplexity is using to show what senators are buying in terms of crypto, right, which is a really cool tool. Um then maybe you need um schema. I don't know. But I I I can't see many valid use cases. I think Jake's done his homework here today. He said that look, most of these use cases don't even apply. Um,

[43:29] but I'll hand over to Jake. Yeah. Um, [clears throat] and David, I'll email you my method. I'll email both of you the methodology too on the schema test. I don't want to say it too much because then people will start to criticize it and try to throw in their own their own test. But, um, when do I think schema is relevant? Honestly, I have no idea. Probably probably never. I I think I think my actual like philosophy on it is it's relevant if your UI is terrible um or your user experience is terrible and and Google and LMS have a difficult time understanding it and um if they do then

[44:07] you have a different problem. Um, if Google still uses it like and and solely relies on it for rich snippets, then that too. But I I'm also not convinced that that you need schema for rich snippets. Uh, I'm pretty sure Google's smart enough to to do it um or to to add those without it. But I also wanted to say too like like before um when we were talking about David, you mentioned this too like people kind of like they

[44:32] complicate things to sound smarter. There like there's two quotes that I really like. Um, one is, uh, Cody, my business partner, he he always says something to the effect of, and I had to write this down because I always I I'm I'm bad at quotes, but a fool complicates things to sound smarter. A smart person simplifies things to do the same thing. Um, and it's actually mirrored by an Albert Einstein quote, which is any intelligent fool can make things bigger and more complex. It takes a touch of genius and a lot of courage

[44:57] to move in the opposite direction. Oh, man. That's great. And that is so true. Yeah. I I like also I like what you said, David, which is go look at the SER, see if the SER changes, see if the search engine results page changes, and then think about, okay, if I add this in, is it actually going to make a huge difference in terms of click-through rate and conversions. And if you think about that, the answer still might be

[45:21] no, not not even worth it. We're tracking we're tracking um Google Search Console in this in this study. We're going to be tracking Google Search Console um uh I don't know what what do they call it? They call it like it's not they don't call it snippets, they call it like search experience or I can't remember if if your actual snippets are showing up. Um we'll track clickthrough rates and correlate that to rank. Um we'll track local um local rank and we'll track AI mentions and the positions that the businesses mentioned

[45:50] in the in the LLMs. I think that there are also things that people can do that are much more effective like using tables. A table is actually very similar to schema. Um it it's like a poor man scheme if you want. Uh it's a bad analogy, but you can tabulate data. And I think the the other thing about LLMs that I really want to encourage people to do is realize the prompt is not the query. So whenever I talk to people um and and they send me a DM or on X or Reddit and they say, "Look, help me out. I I I can't figure this out. I've added this and I'm not

[46:21] visible." The the query is not the prompt. In other words, what you're putting in and go watch the episode that I did with Edward about AI visibility. Look at the query fan out and look at how the prompt changes when you put the prompt when you put the query into Google, you'll see the different sites ranking. Once you figure that out, you get the visibility. Then you can try anything, right? So, for example, if you figure that out, you see how you're visible and you don't have schema, then you know yourself. If you add a table or if you add H2s and make them um anchor

[46:53] text that could improve your site links for example there there are so many things you can experiment always be experimenting there's no charge there's no cost to being experimenting that's why like use tables use H2s because you can do this without paying anyone you know almost all versions of WordPress support it right be um for the for the query fanout stuff we'll do uh for every location we're testing We're doing over uh 80 queries in the LLMs. Um well, the same query 80 times uh every week for five weeks. Uh and then uh um yeah, the test will cost us

[47:32] about $1,000 just to do. It's a big test. It's a big one. I look forward to sharing that. I look forward to the search engine search engine land article on it. I might uh I might go to search engine journal actually on this one. Oh, search. Yes. So, the last one I went to search engine land, but since Semrush since Rushrush posted this technical SEO thing, I thought it'd be funny to post it on

[47:57] a property they don't own. Uh, respect. Oh my gosh. All right. Well, this was a fun episode, guys. Thank you again for uh for being here. David, thank you. It's always a pleasure. Jake, this is your second time on and I hope we do many more. Likewise. Thanks, Ed. Thanks, Jake. Cool. Well, for everybody who watched this on YouTube, thank you so much for watching. For everybody who listened on Spotify or Apple Podcasts, thank you so much for listening. This is episode 926 of the Edward Show. 926 days in a row doing this podcast. And I will talk to

[48:31] you again tomorrow. Bye now.

---

About this transcript
Read from YouTube's own caption track and laid out by ViewRank AI (https://viewrankai.com).
ViewRank AI finds the videos already beating a creator's own average on Instagram, TikTok and YouTube Shorts, transcribes them from the audio itself in more than 60 languages, and turns what worked into new ideas and scripts.
Free transcript tools, no account needed: https://viewrankai.com/tools
How to read any video this way: https://viewrankai.com/llms.txt