Karl Hudson Interviews Luis Salazar Jurado on Semantic SEO and Structured Data

It's brilliant being a developer. He's able to see a lot there as in fix. You look like a technical guy that very detail oriented and you like to solve puzzle and understand how things work and that's something that with the developer mindset you kind of have to adopt anyway. That's it. We try to to always try to implement the same approach and eventually it's not going to work. Obviously, all of these URLs are about to change and you're talking like 10,000 URLs, but if organic is like over 60% of the traffic, then you have to be very very careful, extra careful. [Music] Hi guys, it's Carl from Sir True and today I'm joined by Lewis. I'm going to call them because we've just had a conversation on this and he told me that if I don't continue um calling him the way that I would normally call him. I'll end up just messing it up midway. So, we're calling him Lewis from now on for the rest of the podcast. Lewis, how you doing? I'm pretty happy to be here. Thank you for having me. Excellent. So Lewis is like a guy we've met, we probably met about three, four year ago. It's been a good few years now, hasn't it? We met in July 2023 in Birmingham. There we go. Good memory, you see. Good memory, too. What's that? Two years then. Two years ago, give or take. And he's probably one of the most advanced SEOs that I know. Obviously on par with the likes of Cory and stuff. You are what I would say is one of Cory's tops top students, right? And obviously technical SEO developer come SEO, right? So you were a developer first and then you've sort of went into the SEO world and started mopping up. That's right. Doing a lot of technical SEO, fixing servers, fixing obviously because he has the background which I think is brilliant being a developer. He's able to see a lot of eras in fix. What what programming language was it originally? A PHP. Excellent. So although he's not as advanced as a Python developer, that's why I'm going to say obviously Cory is a bit of a Python nut and he Yes. But yeah, because he's got that mindset. I think one of the key fundamentals for becoming a successful SEO is the ability to programmatically fix things with a developer's mindset. I think and I think that's sometimes what's missing when you have like wild cards in the mix like DY and stuff like that. It often causes more chaos than it does um I don't know structure let's say and he often doesn't understand those elements either. So that then causes even more chaos. Um but yeah so why don't you get started? Tell us a bit bit about yourself, your background and what made you get into SEO. Well, as you mentioned, I was a developer in the past and around 7 to eight years in the field. I was working in a digital in a marketing digital agency and a colleague of mine at that time said, "Hey, why don't you test or why don't you experience SEO because you look like a technical guy that very detail oriented and you like to solve puzzle and understand how things works." And I started like kind of engagement with the SEO just solving issues that uh you know the marketing agency sent to me and from that moment I started to you know get more and more engaged with the SEO firstly technical I mean the the first the first two three years was purely technical SEO technical SEO audits and then I realized that uh everything is uh connected like link building code web performance optimization and everything else after I would say six to seven years in SEO three years three years ago I switched to semantic SEO basically a random video on YouTube on the right side bar said something like semantic SEO training something like that and the first video Karai published on his YouTube channel when he announced this semantic course that was the first moment I knew about semantic SEO and since then what I've done is you know going to events meeting people like you and meeting other guys in the field and learn a lot and let's say now I combine the technical background developer background with semantic SEO that as you mentioned he I think it was the key to understand more easily a bit more easily the Christ framework for because at the end of the day it's nothing more than a relational database or a graph representation of information. That's the the simplest definition I I would say about the semantic SEO and that's something that with the developer mindset you kind of have to adopt anyway. Yes. I mean it's like concepts such as subject predicate object or entity attribute value is nothing more that the relationship between databases. Yeah. nothing more. Yeah. So, basically connecting the dots essentially is Yes. Help it understand. So, I'll be the I'll be the guy who tries to make it in layman terms for if people are watching and they're not as technical and if I get confused, God help them all. That's all I'm going to say. Um, yeah. So, when it comes to So, you've went into I assume at that point were you employed employed by whatever or were you always doing the freelance bit? No, I I was a developer for around eight years. Yeah. And then I switched to SEO. My first experience was working in a SEO consultancy business for a year. The only thing I did was technical SEO audits and understand the let's say the semantic SEO and the link building and how that three let's say three pillars link building, semantic SEO, technical SEO and onpage SEO at that time was the three pillars in order to grow the traffic. When I realized that I could make more money as a freelancer, that was nine years ago. I said, "What am I doing here?" And then I I switched to freelancer. And since then, I don't think I'm going back to ever because I guess it allowed you a bit more flexibility in where you work. You can try I know you do travel quite a bit like like myself, you know, it gives you that flexibility to see the world a little bit as well, right? Yeah. And on top of that, that open your mind to new ideas. Um, it's funny. I'm reading now a book uh one of the chapters I read yesterday. It was uh try to avoid tunnel vision because as humans when we like what we learn and it works. That's it. We try to to always try to implement the same approach and eventually it's not going to work and you need to switch. You have to get rid of your old ideas from time to time. the methodology I've adopted with SEO. To be fair, I never ever I don't, if I'm being personally honest, I don't even think I'm an SEO. Like, I know a bit and I can fit I like to say I'm a bit of a I can fiddle with a website and it seems to start to work. But a lot of the time it's it is guesswork. A new industry, new niche, you are guessing at the start and then you wait until the data comes in and then that data helps you build a bit more of a template. But it always varies. Like I've literally done it across multiple niches where I've thought this strategy will work brilliantly in this niche and then I've done it and it it does not work at all. Like not at all. And it's like oh wow that's changed. Yes. Now uh I don't remember the source of information but nowadays SEO I think it was corre when he said SEO nowadays is more like a business development profile because you have to combine product with technical SEO with PR campaigns with marketing with content with social media distribution. So nowadays I would I approach SEO more like a business development. Had a call earlier and it was a guy who you know he's going through the mill of SEO agencies anyway. He asked me for some advice and I'm like well I was like for a start I can see the SEO agencies done absolutely nothing and how long have you got these in play for? like he showed us his search console and search console was actually like way higher 16 months ago before this agency got in than where it is now. So, I'm like, "Right, you're in a negative ranking state. Your impressions are dwindling. Your clicks are dwindling. They've done nothing for 16 months because I can see this. You still got like it was like a WordPress site and he's still got all the standard like, you know, like categories where it's like archive and all of this. Like all of that still the same." I'm like, "You've been paying these decent money for 16 months. What What are they producing?" And he's just, "Oh, I have no idea. I don't know." And I'm like, But then I'm saying well look this is where as an SEO for his sake they should be coming in play they should be fixing technical errors to start with looking at the crawl server logs you know where the crawlers are hitting is there any blocks in the like blocking the crawlers it getting stuck on old historical stuff that isn't there. Um then from that they should be setting up tests. They should be testing you know adding elements to the page track tracking impressions in search console and clicks seeing what keywords are coming in. Like all of this data you should be every month being presented with oh these are all the tests we've run this month and here's the improvements or here's the you know the losses. There should always be losses as well. I'm like at the end of the day it's it's an overtime game that will compound. it's not, you know, you can't look at it just month one. And he's like, in about half an hour, he's like, "Mate, I I don't even I don't think I've ever had an SEO company explain this to us." And I'm like, "It's because most of them, unfortunately, just give a report." Like, they just give a report. And it's it's a sad thing in the industry. And it's what I was saying to him. It's a very sad thing to see. But the reality is it's it is a boring job. Like, we enjoy it. I assume you would say it as well, but if anyone was to actually watch what our dayto-day is, you would think by God that is boring. Yes. Maybe maybe because people don't spend the time to realize that nowadays it's so easy at least with SEO tools and you have a little bit of interest. It's so easy to check if something has been done in a website or not and how long or what was the last time that something changed in a website or things like that. And with search console and Google analytics and all these tool you can at a glance realize like hey the this site has been obsolete. I mean nothing have changed here for a long time. I mean you can also test it in let's say most sites I I would say a considerable percentage of site are WordPress and just checking the version of the WordPress you can go back in time and say I mean this sites hasn't been taken care of for months or years. So what's interesting with the WordPress angle, so the same guy is about to go through a site migration and let's see your opinion. I'm not going to say my opinion, but let's see what your opinion on this. So he's currently URL structure is the standard in WordPress, you know, and it's like product category slashcategory name slash product or whatever. And then his actual products are just forward slash product name. That make sense? So he's wanting though to have it so it's forward slashcategory forward slash product name the nicer way of having it right not having to have all that extra product type and category crap in but at the same time he's getting a website redesigned so the whole theme will change and at the same time obviously all of these URLs are about to change and you're talking like 10,000 URLs now obviously what would you advise from that point What would your advice be to him at that stage? Is the same domain or a different domain? Exact same domain. Yeah, exact same domain. Okay. If it is the same, if not, what I would do is first understand uh if uh what I call the or it is called the 8020 rule. if 20% of the pages get 80% of the traffic, yes or no. Or to understand what pages gather most of the traffic because if you're going to change the the whole site, it's going to be more costly that if you change only a section or let's say the pages that do not get the traffic. So if you don't change the core and improve the and keep the same URLs, it's likely more uh profitable long term because it's it could be like kind of starting from zero or almost zero. Besides that, I would I would try to understand if the brand powerful enough or not. what percentage of run terms in terms of crawl rate just obviously because I only have search I don't have server logs to actually see like where it gets about 5,000 visits a day from the crawlers it looks like give or take okay and now I would like to understand the brand brand terms versus non-brand terms crap terms are not very good at all not really a big brand okay we have around 10,000 URLs no brand terms around 5,000 clicks a day. URL structure is not the best. Also another KPI would be uh average price per sale and uh amount of sales per month. also a bad seller product in order to understand if they have like a one seller product that sell like 80% or generate like a huge percentage of the revenue because and when I have all that information I would decide because brand signals nowadays very is very powerful and if you mess up with some URLs but you have the power of the brand I will I will start from scratch but if you don't have run terms and you have to start from scratch unless you generate the perfect topical map with the perfect URL it's going to hurt everything perfect and then on on top of that when you do the changes and all the redirects you need extra link building in order to give extra power unless you have the budget you have the knowledge about the all the details about the redirects and on top of that you have the topical perfect topical mount. I wouldn't do it. It's what it's essentially what I was alluding to, but they're doing it like they are doing it. And I was like, if I'm being 100% honest, if you I wouldn't The problem I I see this time and time again is people do too much at once. Yes. try to I always try to explain Google Google gets a template of your website gets a it understands you know based off a template of the site and then you're about to fully change that template and not just fully change it you're going to change the entire theme architecture because he's changing themes and everything you're going to then also change all of the URLs so you basically just just destroy the template and then you're going to trigger loads of errors because there's going to be loads the 301's hopefully if the company that's there do does it right which based off what I've seen I don't think is doable and I was like if I'm being honest I I was I was like I would do a stepby-step strategy I would implement the URL changes now that's the same theme as there slowly introduce the URL changes so it starts to understand the new structure at the same time as increasing better links and stuff like this and then once the URL structure and the crawl rate stabilizes again I would then look to swap the theme. I would add on top is like I would do it section by section and another metric I would include is what's the percentage of organic traffic in the overall traffic of the site because if it is a huge percentage you have to be even more careful. If it is like let's say uh less than 25% like now they I started working in March with a brand where organic traffic is the fourth source of traffic they have like direct traffic uh pay-per-click email marketing social got social social is even stronger than organic so and that's no problem but if organic is like over 60% of the traffic then you have to be very very careful extra careful this is what I was saying I was like the amount of times I've looked at websites and like a good friend of mine he probably won't mind the same but he owns like an e-commerce store and he went from um Magento that's what he was on which we all know is it is a brilliant e-commerce engine but it is messy and clunky and heavy and he was ranking for all sorts. I think about 2 million organic traffic. Like big big site. Swapped to Shopify. Oh dear. Over 350,000 44s. Yes. That he had to start trying to fix but obviously couldn't fix them quick enough. And his website literally just like boom tanked and and slowly starting to recover now. But very very and the problem is because it tanked so bad, he then doesn't have the budget to spend on link building. I'm like, it's a two-pronged sword now because if you're doing a site migration, you should also have a decent budget aside to plow loads of links in because you want, you know, justification for Google to revisit that website and reindex it all essentially is what you're asking because it has to understand and create a new template and just people don't seem to understand this. that it's it's crazy because that's what I also consider and you've probably seen it yourself, Louisis, when you check server logs and let's say you found a cluster in the server logs that just isn't getting much love, doesn't seem to get much impression, stuff like this, and it's a little cluster down here. You build a couple of decent links to that cluster and all of a sudden, boom, it just like skyrockets that cluster, right? And that's what I think people don't put the two and two with link build like they don't do the connection of you're actually helping feed the crawler helping you know it understand the template that's what I like to say that's what link building essentially is for yes and as you mentioned it's very you have to be very careful especially with big e-commerce if you are just switching the framework but you're not changing the domain on top of that you changing the whole technology in the whole in the whole site is like very costly for for Google because um I mean imagine like using this situations in this occasion the analogy of you are moving from one house to the other where you know you have to not only move from one place to the other but let's say you have to communicate with all your friends your new address and you have to move everything all the furniture and everything and putting in order in the new house. So imagine that if Google have to do that with probably thousand or hundreds of thousands of sites every single day. Yeah. From a even though technical but logistic or logistically way of think is like very costly because you have to rec crawl every single URL again. Yeah. which is cost for them and if there are probably a few thousands competitors in that niche why should why should Google worry about your site when there are other 10 site that can offer the same product or service well that's exactly what I was saying I was like look you know in this example as well like he has like category pages that they they don't rank amazing might be page two or page three they've never had a title tag change they've never they don't even have a paragraph of text on like nothing like it is just a shop for and I'm like mate you need to start developing some content on these categories you need to be doing internal linking product pages should internally link back up to the top level category there should be a you know a breadcrumb navigation there should be some I'm like there's so much to change on this site I just I was like personally I would just do it step by step improve the site as you can um as much as you can obviously install some type of I don't know do you use any tracking tools like SEO testing or um SEO crawl have you heard of that one you know where can um you can essentially you might use big query for it but I I don't go into big query but you can essentially annotate onto the SER and say I've edited this page you know I've put whatever changes and then it will track the impressions and clicks and what other keywords might come in for that specific page and then it'll email you the test data of like oh you know this test seem to have a positive impact after seven days or 14 days whatever you set so it's a pretty cool way of doing it because you can always be constantly testing and trying to improve and I'm like this is what you should be doing all of this and he was just yeah it was just funny funny let's continue down that rabbit hole So this website is in a negative ranking state, but it's I would say in more neutral stroke negative. It's not like hemorrhaging like we've seen, you know, sites that have been hit by HCU and stuff where they've Yeah, it's almost like flatline. This one still gets decent decent, but it's not really in a positive one. What would your first steps of attack be? What I would try to understand is in the pages that already gathering traffic, I always start with the question what am I missing? But what is missing which is I go to firstly the meth mental SEO checklist and I try to assess if there are opportunities to keep growing or keep gathering more traffic in that landing pages. Today I did a publication on LinkedIn where we have been working for two years in a landing page. Yeah, I've seen that. I've seen that in the insurance business and everything we've done is to fine-tune what we like 10. It was 10 clicks a day. Now it's hundreds. Yes. So what we did was to fine-tune the page and we have all the what we do is we track all the minor changes we do in every single page. So we can assess in 3 6 9 months if the changes we did were successful or not. Let's say what I would do coming back to your question is first assess the pages that have potential identify those SEO optimization that are missing and try to get more a little bit more boost from those pages and then filter the sites like category pages, product pages and so on so forth. I would say do we have potential here? Yes or no. And then I would start with the most powerful in order to get the boost of traffic. And then I would if the migration is inevitable, what I would do, hey, let's save this back or this batch of pages because these ones are going to be the most likely are going to keep gathering the traffic unless demote or kill the pages that do not have that much potential and focus their resources in what it's more likely to work. Yeah. and do not spend time, resources, content and anything else in the pages that are very unlikely. And then balance the the budget and the resources and hopefully get good results. So let's say I'm playing devil's advocate here because I would be interested to hear your answer though. Let's say it's a dog training website. Mhm. Dog dog products, you know, toys, whatever, chew sticks, stuff like that. And let's say we found chew sticks as the best cluster. So we're now evaluating this chew sticks section might be a collection chew sticks and then obviously the individual products. So we see that the impressions are pretty high, clicks are all right and you're now thinking right I need to expand on this. Do you you've noticed that there's maybe some keywords that are missing. Do you add those keywords to the top level collection page or do you try to create subpages within the cluster first to test to see if that improves it? Yes. uh we're now now going to approach this question such as an statistical linguistic and distributional semantic issue which is again I start with the question what am I missing which means in the content what am I missing first let's say is a category product I'm going to evaluate the entity itself which is the product and I'm going to say okay we have the template what attribute am I missing in the entity and Then I would optimize that. I would optimize the whole section. Not creating more content, but again optimizing what I already have. Like maybe I'm more frequently shape, color, like all of these attributes. Yes. I mean it's like do we cover the price? I mean do do you remember the presentation Kai did in Chennai last year when he approached the product and he said it was kind of okay we have the product and do we cover the price do we cover the colors do we cover the description do we cover uh related products so you go step by step at the checklist and in that process I'm pretty sure you miss something I Do do I have all the information in the product? It doesn't matter if it is the feed from the provider whatever and I start thinking like okay what information is missing here that is is going to be useful both for the search engine and for the user and then uh from that I begin researching maybe it is a structured data in the e-commerce do they have the JSON D product is h the JSON D product validated. Do they have all the attributes about the product? Are they missing some attributes about the product? Price, size, color, dimensions, description, title, image, add all attribute of the image. I go to to the whole checklist. But I wouldn't I I I would only create more content if uh let's say I do kind of mining or little topical map for that section and I check okay what am I missing here either it could be possible that we're missing a very important article about that category product or something like that but that would be product afterare or something like that. Yeah. Oh, that would be my approach. Yeah, that makes sense. So, essentially, you wouldn't you'd almost avoid opening up new pages and see what you can improve on the existing one and how far that gets it essentially. And then do you get to a point where you think right like at what point do you tick off, right, I think this is as far as that's went. I'm now going to focus on this cluster. Do you typically have like an impression level where you would kind of go that's grown nicely. Let's now focus on this one because obviously you've only got a limited amount of time, right? So you need to try and get multiple plates spinning almost. Or do you at what point do you think that's complete if that makes sense? What my approach in that would be first I need to enhance the overall site like what technical optimization content optimization what I'm going to do that the benefit spread all over the whole site if not it doesn't make sense because the equation effort return is not going to be positive and from that moment I would focus on let's say two fronts. One the batch of the category optimization let's say and the the second batch would be like the enhance the site overall either by minimizing redirects internal redirects or for improving the crawl rate or I would also improve the web performance optimization. What can be done that we can reduce the the time that the page is loaded? This kind of stuff. Imagine that. Let's play kind of mathematical game. If a on average a web page is 2 megabytes and you reduce it uh 200 kilobytes, it is 10%. Imagine 10% less cost per page for Google to crawl your site. That's a lot. If you then compare that to competitors where you know your competitor might still be at 2meg and you've came up with a new website that is a one megabyte 50% less. You're giving it more reason to want to come back to yours. Yes. Yeah. So these little things compound over time. If you're lucky between code updates and code updates, you can get a boost. Yeah. So, um, also in the migration process, I and I remember this from Kora is try to do the migration in the season with the lowest traffic and uh try to uh let's say set the server in order to minimize also the crawl rate. Yeah. So minimize the crawl rate in the server side with in the in the season with the lowest traffic and then you do the switch and then you try to tick all the boxes possible because and then when you let's say change the settings in the server and more and more traffic comes crawl rate is going to increase and hopefully you interesting. So you would actually limit the crawl rate within what robots or text or HD access sort of stuff. Try and limit the the crawlers accessing the site or you Yeah, I didn't think of that to be fair. So you're almost intentionally restricting the crawlers to reduce the amount of errors that they're going to find in hope that you can fix them quick enough to then turn on the tap again. Yes. And also what I would do, what a thing I I'm seeing nowadays is Google even though you block robots with do not crowd this directory and you put no index tag robots in the code you still see it in server logs but the bug is going to keep coming and on top of that if you have user signals it's like user keep clicking in those profiles I mean it's like a is like a ringing bell for bots and keep coming even though you provide all the signals not to crawl those pages. Yeah. Yeah. No, I definitely see that as well. I've seen that quite a few times in um like when just checking through server logs and things like this. Right. So, you've answered that question. Okay. So, what about back? Right. So, we've now changed the site. It's in a positive ranking state. Give me some dos and don'ts because people I think this is the problem. people start getting greedy. Now we're citing a positive ranking state. I think we've even seen it on some examples of Cororey stuff where you know people have maybe used Cory for 12 months, got really a huge success rate and then they've went off on their own and then they realize that they're in a positive ranking state. So this website can rank for anything which in theory it does. But then obviously something happens at some point and it tips, right? Well, my approach was and still my approach is play the infinite game. It's pretty unusual. I mean when I compare and contrast my approach with other acos because my three oldest clients are 9 years, 8 years and 5 years and what I realized in that experience is from time to time the algorithms are going to demote you but long term I mean if you're not willing to invest at least a year uh I don't think that the SEO campaign is going to be successful because it's like you precondition yourself not to invest the time and the resources in order to make it successful. That would be my approach. I mean when you when you see a project growing positive ranking state everything's rosy this is when you have to pause. This is what I do. I mean I pause and I say okay how can we improve because the easy way out is more link building. or more content, you know, but that's not my approach. I mean, I do encourage everybody that see this that think about I always start with the same approach. What am I missing? What can I improve in this page? So, I was having a conversation yesterday with Constantine. I think you've met. Yes, I know. And he me and him often have debates. I think every day I pretty much speak to Constantine and um I was saying to him right on our site, you know, what should I start publishing next? And he was like exactly what you said. He went mate, stop worrying about publishing. Start start tweaking, start editing your existing stuff and improving. He's like, you don't like too many people get obsessed with publishing new content. And the reality is for every one new piece of content you publish, you should have done 20 edits to existing. He's like just as like chucking random numbers out there, but like that's the balance. He's like, you test obviously set up something like post hog or something that's tracking user signals. track the user signals via, you know, heat maps, stuff like that, and try to improve the user experience in general on every single click that's coming through. And guess what? In time, you'll be rewarded because Google should be seeing through Chrome data or whatever, they should be seeing that happen. And improving those user signals, he he says, is the biggest factor for rankings is user signals. Yes. I mean coming back to the publication I did I did today on LinkedIn let me I would say let me go back to summarize let's say over the last two years let's say we did like four to five changes nothing more but we spent a lot of time I spent a lot of time with the client trying to understand what they want to sell in the landing page itself the call to action the funnel itself and then we fine-tune a little bit of design, a little bit of copy, a little bit of micro semantics, which means either rewrite a specific section or include related terms or synonyms that we that we were missing. That that that's the exercise. I mean, I didn't didn't buy links. I didn't create it related content to that target. Editing that existing page and fine-tuning it. Yes. Were you using much AI to help fine-tune? Do you do much AI stuff or No, the approach I do is the one I share. I think it was I don't remember if I share it publicly. I share it in little groups in Chama and the masterminds. It's like every two three months I check search console and I go landing page by landing page. Of course, I track all the changes in all that land page and I go back. I do a kind of exercise in order to it's basically a filter to request to search console and identify new terms or terms that increase the impressions. And this is kind of the leading indicator in order to move forward. Mate, you definitely need to use SEO testing or something. I don't know if you've got it for, but this is what SEO testing does. So it like it spits out a nice email and says this page is gain these new keywords and these new keywords have x amount of impressions and then obviously you then notice that they're not present on the page. It says these keywords aren't present on the page. Implement them on the page and all of a sudden it's it's exactly what you've just said there essentially is like that's doing itself rather than you having to dig into it. Yes. But the I am in the process of automating this with right okay query or yes I have a kind of joint venture with a guy you know where we are connecting search console API with bigquery when I have enough data I'm going to have a chat with this guy in order to generate the kind of automation related to that because my let's say manual and traditional process I do with my sites implemented let's say semantic SEO improvement we want to automate basically both the identification of opportunities and then the recommendations about those opportunities and not not only landing page per landing page but at bulk that's our medium long-term goal so how much would you say obviously we've mentioned a lot so far about like ranking states and then it's all primarily on page. How much would you say you look at sort of links and link building? How much would you say that has a place or at what point would you think, you know what, I I've done a great job here. We just need some links. It might sound odd, but over my I would say nine years in SEO, I only was once forced to buy links because the client forced me to do it. But now I'm in the I gaming industry and the g the game is different. Now I have to buy links. But outside I gaming I do not buy links because firstly my filter is I only work with brands that have enough power. Yeah. Second, in the process, in order to assess if it's going to be a success or not, I do my research about how much they invest in their brand, in the PR and so on so forth, which is basically, you know, I 100% agree with you as well because I've got a friend who is absolutely crushing it. I think he launched his business over a year ago. He's just hit the million pound mark. Um, obviously absolutely crushing it and awesome guy. He originally started with just paid ads, spending all on brand doing social media, Google ads. Didn't look at SEO at all. six months in gets me to have a look at SEO and I'm like mate it's the easiest site to rank in the planet because it has all these brand signals like people coming direct in the browser loading the site um people searching for the brand just every brand signals popping off and it's like big search volumes and you just literally you you make a tweak and it's like like just small little micro tweaks like you maybe changed a sentence on the page and all of a sudden you've gained 10,000 impressions. It's ridiculous. Yes. I remember after what you mentioned, I remember two years ago in Chiang Mai, Craig Campbell shared with us a business. It was a brick and mortar business. And he said, "I bought the business. I remove everything only traffic for paper click." In fact, his presentation in Chia two years ago, he mentioned the project brand signals, traffic that convert, user signal and he did the you know the numbers and he said I don't need to invest in anything else he just need to he automate the I don't remember the framework but I do remember the fact that he mentioned that I don't do any other thing that payer click because it's traffic that convert and all the signals for Google are there should tell you what I also because we're talking about signals again and saving Google money. What I don't get is people who don't invest in the paid ad side, which is literally the only way Google makes money from you is by you spending on paid ads. Wouldn't it make sense, although they will never ever admit to it, wouldn't it make sense that Google might just prioritize a website that's willing to spend some money with them over a website that isn't? I cannot confirm or deny that, but but I would I would share another idea. I I I think I heard about the same event. Yes. Is it the Chai? In the Chimai one in Chai 20 Chi 2023 there was a guy who said I convinced the client for 6 months. He gave me the budget and he focus all the budget in perp generate the traffic and the user signal and the revenue. And when and he doesn't do any SEO. So when he ran out of budget he see again with the client and say hey we started with pay-per-click you have seen that this is profitable keeping more budget and then he started doing SEO for six or 9 months after that and like you mentioned also doing SEO after all those signal all these historical data is there it went up like crazy so it's one of The key things I always suggest to anyone I speak to, it's, you know, you should always be spending on paid ads because at the same time, an SEO can do an amazing job. You know, you could do an absolute stellar job and get the website ranking number one, but you don't know what keywords are going to convert and generate that return on investment for the the business. That's why they should be spending on paid ads because that will tell them very quickly where they're going to make money and that'll tell you very quickly what are the main terms we should be optimizing for and you build a strategy around that as opposed to going in blind using ahrefs pulling out the same data every other SEO on the planet has access to um and then trying to guess it. Why not just spend it? You know, it's it's not even that expensive nowadays. They give you free credit for I'm sure it's like spend $2,000, they give you $2,000. So, you get $4,000 of budget, which would tell you what to do month one. Yeah. It's like the pay-per-click data is the north start. You get that as the beginning and then you move from there. Exactly. And guess what? If you can make it work and you're making money, the business is happy because they're making money. So, keep that going. And then in the same time, start developing the SEO strategy which will help them make more money. And I assume this is what you've probably done with most of the your aged clients. They probably love you for life because they know you at the end of the day, they speak to you and you give them a return on investment. Whereas so many SEO agencies go in and they just don't have that mindset. It's like such we were talking about it earlier like a tunnel vision. Tunnel vision and it's just yeah it's detrimental. Traffic alone is not going to give you the traffic without conversion is not going to give you anything. No. Exactly. Exactly that. Um, so we've talked about positive, we've talked about, well, actually we haven't really went into that. So I'm going to move slightly off topic into So you're in the back of search console and you've went into crawl stats and you see file types and let's say GS is it's a WordPress site and GS is 90% of it. It shouldn't be, but let's say it is. What's your key steps there? What are you then looking at to help improve the because I know chorus is realistically want to get the HTML ratio as high as possible and let's say HTML is 9%. What are you then looking at? Are you going into obviously you'd click and see what scripts are loading? Are you then trying to reduce the amount of scripts down so it's all like loading from one script? Are you blocking scripts and robots.ext text or trying to stop stop it being crawled. What are you doing to see best bang for book and to change that ratio? Well, first I try to assess the overall state. Let's assume I have no data. Let's let's assume I just have search console data. Yeah. Um I can crawl the site but I do not have access to the server at all and I don't have communication with the developers or the CIS admin whatever. Yeah. What I'm trying to assess is how a bot crawl the page and render the page and after that I can go in the DOM I can use search console and trigger the page in order to make like a kind of analysis first of all is the page crawable and it can be rendered and if it cannot be rendered I try different ways in order to guess now because Google over time has been I mean we cannot not see now the cache version of the page. Google deactivate JavaScript in the browser. So over time you have to it's a fun game for me which is try to guess what's going on. So what I do now is I look at the code and I analyze the header in order to see if there is some JavaScript that is blocking everything. Yeah. because there might be some JavaScript that is blocking everything and if that JavaScript get a lot of time it could uh let's say put the crawler off something like that. You can use Chrome or Firefox load everything check the time check the timeline the waterfall feature the waterfall and then you can spot at a glance like hey there's like either a massive for too much JavaScript or this JavaScript it could be generating this issue and then you focus on that because the it could be so many things with that in mind then are you then potentially trying to remove them out the header, put them in the footer if possible, like deferring them maybe so they load after the DOM. Um, is that what you're trying to do? Or I also disable CSS. Yeah, disable CSS to see the structure of the page overall because maybe in the header they're loading uh massive stuff and then everything is displays to to the bottom. Then I compare and contrast with the crawling because maybe I don't know the content is is in the thousands of pixels below the play they should be and things like that because do you have much weight on the the kilobytes as well? So do you try to get your schema was it the first 40 kilobytes of the page load? Do you want ideally your title tag and your schema to be loaded within the first 40 kilobytes? Yes, I mean I always recommend to everybody I work with and pretty much I'm going to share what I do is make sure that the metadata tags are in the header as soon as possible and in the right order like title description and so on so forth. Secondly, always the structured data I put it on top, not on the bottom because if the if the crawl do not have budget, uh at least it's more likely they find the JavaScript quite quick in the header than in the footer. And you would you usually position that? So when I'm positioning that I usually do title, your meta information and then straight away it is schema like this the the uh schema script on that's typically how I do it or would you do you usually put that even above the title and meta description? So no and what I would do also is understand how the JavaScript are loaded in what order because maybe there are so many requests to JavaScript that the bot don't even have time to go to the metadatax because all these frameworks with so much JavaScript and is is a nightmare either I try to communicate with the developers and say Hey, could you put the metatags before all this massive JavaScript because it's going to delay the time for the bot to digest our data. No, that's good. It's really good to know. So, I think we'll move on to server logs now. So, I've got a few notes of questions because it's quite nice actually speaking to an SEO. Um, okay. Okay, so you've got some server logs and now it's amazing or I also refer to them as access logs because it's amazing how much people just don't know what a server log is. So obviously this is typically you get them from your web hosts. If you're using a shared web host, chances are they're restricting them to maybe one day rollovers. I know I believe Kinster is a threeday rollover. So you have to create a cron job to extract it and save it into a different folder because it's going to roll over every 3 days otherwise. Now what are you typically looking for? Well, A, what would you say is your minimum amount of time on a server log? And then B, what are you typically looking for? Firstly, each project is different. If the site has a lot of traffic, I limit it to a week because if not, the size of the file is going to be so big. Yeah, it's going to be huge. If the size is not that big, I try to get a month of data. Yeah. And then I assessed I filter by Google bot in smartphone and then I try to understand why the bot is spending so much time in a page that in theory has no value and try to identify either 40 or four internal 301's redirects all these kind of issue mind are you then so let's say you find a page that it's just a random page and for whatever reason. I've definitely seen this happen before. Google bot just is visiting this page. Are you then going, "Okay, let's have a look at this page. Let's put some internal links higher up on the page because obviously that then might encourage it to visit other sections." And then do you wait for a week and see if the other sections have then increased? If they have increased, obviously that must then be working. Or do you kind of just go, do we need this? Should I re through redirect this page to another page to encourage that? Well, I use that approach, but from time to time, even though in theory you do the right things, bots keeps crawling pages that it doesn't make sense at all. It's like uh but anyway I try to first focus on what's important which is fixing as much issues as possible and then with those kind of situation where you are trying to guess why the bots are wasting so much time in those pages. If the page it doesn't make sense to keep it there either because it's not a money page, then I consider a redirect because at least you can transfer the value to somebody else. That makes sense. And are you looking at like I know in Screaming Frog it's CO2 emissions, right? That's trying to tell you, you know, how much it's actually costing the server to, you know, store and render. Are you kind of looking at that as well and almost the weight of it when it comes to the crawl and then trying to identify ones that are maybe a little bit too heavy and trying to improve that or you not really looking that granular at it? Now I try again going back to what we mentioned about blocking. You block something in robots GST and the bot keep coming at the end of the day they they're going to do whatever they are programmed to do. What I what I'm trying is to first minimize cost, second fix issues, and third transfer value. And with these three approaches, I would try to I I try to let's say get the most out of it because sometimes you don't even understand what they're doing. Yeah. So when you So let's say you're now we'll go back to ranking states because I quite enjoy talking ranking states because it's just interesting to see how people get out of different different ones. So let's say you're in a ne So this is the interesting part. You're in a negative ranking state, but if you're going through page by page, query by query and you're seeing that you know one page actually isn't in a negative ranking state and it's in either a neutral or a positive. What essentially you were saying prior is that's when you would focus more on improving that cluster and then with the hope that it's going to change the site the negative ones in general with time. Is that kind of the hope? Yes. I use it the analogy of the anchor because a negative section let's say a category of a set of URLs in your site if they are in a negative ranking state they're going to they're going to be an anchor to the whole site. Yeah. And eventually you have to make decisions because it's like one one category is gathering more and more traffic and one category is losing traffic. And then you have to decide if it makes sense to spend more time and resources in order to try to get rid of the or reverse the negative ranking stage. But you have to do something either radical because if not you're not going to I mean you have a limited time to reverse that because you know eventually it's going to die and either redesign the topical map or create a new section that from a scratch in order to start gathering more impression more traffic and eventually kill directly that section or improving step by step the whole section and waiting until the next court update. That's the overall approach. And have you seen or it's actually quite interesting you mentioned earlier how like being involved in certain websites for so many years. I actually had a same debate with a friend who he's kind of exited SEO but he's got e-commerce stores that do good money and it's all paid ads. He says throughout the years of having these e-commerce stores, like you're talking like maybe seven years, it's actually surprising to see it just comes and goes. He says like he sees organic traffic goes high, organic traffic goes low, organic traffic goes high, organic traffic goes low. And he says he sees that time and time again. And he hasn't actually touched the website. It's just with the paid ads. Is that something that you're seeing quite quite a bit as well? I mean from time to time uh there are pages or entire section that you lose traffic because it makes sense because Google could not rank one specific site in first position all the time or forever because it's always is such as life you know but if you if longterm you take care of the site or you take care of the brand or uh eventually You have all the ups and downs, but the long term you're going to win. For instance, a month in February, I started working with a company in Berlin, and these guys have been like over three years doing nothing in SEO, but the brand is very powerful. Is between 10,000 and 15,000 clicks a day, only brand terms. Wow. The guy that now is the product manager four years ago was my colleague in SEO. So he switched from SEO to product and because he was so focused on product SEO was not implemented at all for over three years and of course the traffic suffer even though the the brand is very powerful. But the good thing is now doing very little we can go back uh track see hopefully big jumps as well with very little effort. Yes because they have the data year over year and 30 40% uh decrease is a lot of traffic for a site with you know this amount of traffic on a daily basis. Uh but again it's like the whole universe of no brand keywords is huge compared with brand terms. So the opportunities are amazing. Is there any sweet spot because I know you keep mentioning brand keywords and no brand keywords. So let's say for instance your brand keywords are getting loads of actually that would that wouldn't correlate well. I was going to say loads of impressions but not many clicks. Well, that wouldn't really work because if if you're getting loads of I I assume your brand keywords are going to be your highest click-through keywords because you should be number one and if your impressions are going up, chances are people are clicking. But is there any like minimal minimum amount that you tend to like to see on like a brand keyword basis? I'm going to share with you my oldest client. He's an e-commerce. they bought not an IMD like close enough to IMD but the main entity they want to rank for is not the IMD. So what what we've done over nine years is try to rank the main entity and then build content around related entities. And now what I'm doing is over the last three years I know the key entities I want to rank for and the attributes about those entities and then I connected the I would say four entities with rank for and then generate like a kind of attributes around that and keep expanding from that. So if you start from scratch and you don't have brand power, you can build step by step these connections among entities and then create a kind of brand around that because if search demand is there, you have like kind of uh protection. Yeah. And do you do you think that's what a lot of the EMDs benefit with when you've managed to acquire an EMD? It's because it's already got the search demand. Yes. Yeah. And if the I mean if you build kind of brand around those searches surrounded the EMT then you can grow and this is what I'm going to do in the eye gaming thing. So it's also what I did for many many years in I gaming EMD domains that's what was our bread and butter. I think we had at the time of me exiting we had about 45 domains sorry active websites. I I dare say about 25 maybe 30 of them were EMDs. So a lot of EMDs, a lot of experience of they're probably my favorite way of ranking if I'm being 100% honest. Yes. Just again because it's almost like it's kind of like when you're what we just touched on there. When you're going into search console and looking for branded search and AMD already has branded search because it's the AMD, right? But again, obviously, it's just making sure that that that it ranks or attributes it. And I guess that's why you're on about connecting the entities, the various attributes to the entity as well, help solidify your position of being the EMD. Yes. The the more historical data you have, the better. So that's why my approach is longterm because if you push uh too fast too really might breaks. Yeah. But if you create the signals over time this kind of more natural approach and algorithms give you a little bit more confidence let's say. Yeah. Yeah. because it because you cannot I mean think about I don't remember think about that each domain is like a kind of business only a few business since inception time since they they are born grow naturally very fast only a few only a handful of them the other ones either grow very slowly or die why in the internet is pretty much the same not all business can be successful fast growing business. So Google, let's say that Google have to assess which site should rank or not. Let's say in the link building process and now let's say relearning. I'm not going to generate 10 links in a week and then I'm going to be like three four months with no links. I have to create like kind of steady approach with content creation and so on so forth. The same logic is for any business online or any site. So with that in mind though, so I'm going to chuck a little curveball here. You could say the natural process of bigger business is usually they do digital PR and stuff like this. They have an influx of links and then that they're done for the next quarter. Then they do another batch because they have a next budget next quarter, next budget next quarter. And that actually sometimes in some industries you see that sort of growth is almost like you have to match that level of growth because obviously that's what they I assume Google makes a bit of a template on what it expects to see from that industry if that makes sense. Yes. I mean my second oldest clients 7 years the company was bought in 2021 and since then brand terms have been steadily decreasing. Yeah, guess what? The SEO has been decreasing as well. Now they're for certain terms for let's say like a batch of keywords. Now they realize hey competitors are doing PR campaigns, marketing campaigns, promotion and so and you the way we convince let's say the management was we went to Adwords and all the competitors are like this like like growing like crazy and we are like kind of steadily decreasing the brand power and brand searches. So, and then you do an open question like, hey, why Google should choose us instead of the other guys if nobody is looking for us anymore as much as in the past. So now hopefully we are creating new landing pages like new verticals and of course marketing campaigns and peer campaigns are going to be launched natural sort of search. No, it's interesting. It's mad how like basically I think when you start to get to a high level in SEO you almost become um a general marketer I would say more so than just an SEO like it's obviously your goal is SEO and ranking but you almost have to have an open so not tunnel vision an open eye for the other verticals that are going to come into play there and I think that I've seen that more and more to be fair most interviews with highlevel guys it's not always just about keywords and links it's you know user signals brand signals improving everything yes I I have the kind of my turnship here I use every day to work most of my time is communicating with clients explaining what I'm going to do and what to do it I would say over the last two years if I analyze my real time in the rankings and the analysis since is getting smaller and smaller and smaller and the communication process strategy why we do this we have to connect marketing with that is more of my day-to-day job yeah it's actually funny because like one of the consultancy roles that I have is or that I do whenever I go in with anyone on consultancy I literally I say to them up front if you want a glossy report and you want me to spend spend time doing a nice report for you. I'm not the guy. You go to this agency over here. They'll give you a lovely report. You'll not get any results, but you'll have a nice report. That'll that'll tick that box for you. If you want results, then leave me alone. I'll work on it and we can have a a half an hour call. I can show you search console impression data. You know, I can show you all the data what we've been doing and that'll be the results. It's not I don't like doing PDFs where it's, you know, a 60page PDF that I've manipulated to hell to try and justify the money. I'm like, "No, no, no, no, no. That's not how it works." I couldn't agree with you more. Uh, my system is before we start working, I communicate with them the way the communication between us is going to be. The documentation is going to be like this. The reporting is going to be like that. The reporting is nothing more than a video. Yeah. A two, three minutes video with Loom. Yeah. And that's it. And that's just talking through what you've done the maybe some consult data. Done. And the the good thing is you reduce the emails, you reduce friction. Yep. You have more time to think, you have to more time to analyze. the the customer is happier because to be honest they they don't what I learned that most reporting that people people don't see the report people don't see the the this neither the spreadsheet or the nothing they don't see that they just want the valitation let's say or the are we moving in the right direction yeah that's it um when they understand that is I mean the It's more fun also because it's like imagine generating a monthly report for all these people. I mean it would be like 10 15 hours wasted every every month. What exactly? So let's I'm aware I've just looked at the time there. We've been talking for quite a while, haven't we? Let's summarize that. I want to finish off then with obviously because you've got some clients there who you know that is a long time. What? Nine years was your maximum one. Nine years, seven years. Nine, seven and five. Five. So for anyone else who's doing SEO consultancy or um maybe even an agency owner, what would you say is your biggest tip to them to help keep that client on? Because I know most SEO agencies churn rate is probably six months. And that's me being lenient like maybe six months. It could be less. What would you say is the key that that's helped you succeed? What helped me out over all these years is try to prepare in advance for what they're going to need. So they do not request information for you. You deliver the information and the proposal before they even think about it. Let's say nowadays AI is say the flavor of the day. Every time they my client mention about the eye, what I do is to go back in time and say ask questions. Do you know when Chajipi was launched? Do you remember? Do you remember when Gemini was launched? Do you remember where all these times when they I forced themselves to search on Google and when they realized the time that Google promise AI and the product hasn't been fine-tuned or optimized yet how chd have evolved and perplexity and all this too and they realize it's been more than two years now guess what nothing much have changed So I'm not saying that I'm don't pay attention to what's going on. I don't say I'm done good doing research and test and so forth. What I'm saying is in two years nothing big happened. Maybe the next two years something happened but don't get mad. And from time to time, every 3 to 6 months, I just generate a kind of, hey guys, due to the new releases and where I do believe the industry is heading, I would recommend you to do this, this, and that because that and then they take it into consideration. Yeah, that's that's my approach. So almost like being the forefront of it, but you're being in control of the narrative. Yes, I mean I have to this week has been let's say I've learned a lot in the mindset and communication because I got one nugget from one idea in a WhatsApp group in the rank and rent cohort and that idea trigger a research that gave me an insight that I could share with all my clients and all these guys in the cohort and relate to other component which is nothing more than freshness in local SEO. I mean how if you keep the freshness uh Google business profile manager landing pages local signals so on so forth you are ahead of the game because most people most people is set it and forget it set it forget it and build links. So that would be another approach in the not only in the local SEO but in the holistic approach. No, that's all good advice. Very very good. And obviously we're nearly an hour and a half in. Jesus, this has went, hasn't it? This went by fast. So how can people get get hold of you, Lewis? They can find me on LinkedIn and they can type my name. Obviously, I'll put everything at the bottom of the the um video as well. Have the team grab it. I'm not very active on social media to be honest. Just a LinkedIn context doing the job that's why. Yeah. I I like the guy of the day said to me, "You are the kind of guy that always on the trenches. I like that place. I mean, this is where I've been all my life. To be honest, the day we met in Birmingham in July 2023, it was the first, no, it was my second time in an international SU event. Wow. The first time was in 2016 in Milan and I went there to meet Billisloski in person and from 2016 until 2023 I was playing the game and I forgot to mention I went to Burmic and also to Met Cory personally. Right. So, another thing I would advise is go to SEO events because guess what? Almost two years later, I met this guy. I'm doing a podcast with him. And who knows what could happen if you go to a SU event and you hang out with someone, meet someone. It could be like the beginning of a long-term relationship. I also think just for like in general expanding your horizons, you know, like you don't know what you don't know and like there's a lot of like knowledge out there and people especially if you're more of a one-man band or you might only have a small team, you know, you are limited by time and if you can share what you find working in cohost or you know various groups that with people you meet at conferences as well, it it just all helps it helps with your obviously test everything like don't just take for bait and what someone says because again it varies niche to niche but it at least helps you give some ideas or some um test ideas to run. Right. Absolutely. All right. Excellent. What I'll do, we're just approaching the 1 hour 30 so I'm going to terminate there. It's been great speak to you Lewis and then obviously I'm sure we'll meet again soon. Thank you very much for having me and I hope to see you