Olga-Andreas - Made with Clipchamp
===
[00:00:00] Olga: Hi, everyone. It's Olga Zarr from SEO Sly. This is SEO podcast by SEO Sly. Today I have a special guest. This is Andreas. Andreas, how are you doing? SEO Sly. SEO. Done. Right. I'm Olga Zarr, an SEO consultant in SEO since 2012. Don't forget to subscribe to learn SEO for free with me. Now let's get into the show.
[00:00:28] Andreas: I'm well, thank you very much, Olga. I'm very well, thank
[00:00:31] Olga: you. Yeah, yeah. I am very happy to have you here.
[00:00:35] Andreas: Good to see you since, uh, Barbados SEO. ,
[00:00:37] Olga: yeah. I think we met on Barbados, I think during the lunch. Like the second day it was, I don't remember. Yeah, it was the second day. Yeah, it was raining like crazy
[00:00:50] Andreas: there.
[00:00:50] Andreas: Oh yes, it was, wasn't it? I think everybody's laptops were getting
[00:00:54] Olga: soaked. It was cool. Are you going to Barbados again this
[00:00:58] Andreas: year? Uh, I, I, I know I'll go again. I don't know if it's this year though. So, uh, okay.
[00:01:06] Olga: Okay. Me too. I'm not sure as well. Okay. Okay. So, um, can you briefly introduce yourself?
[00:01:13] Olga: You are like an SEO veteran with so much experience. I am sure this is This is going to be a super, super interesting episode, but if you were to introduce yourself for someone who doesn't know you, can you do that in two sentences or three?
[00:01:28] Andreas: I'll try. Uh, yes.
[00:01:30] Andreas: Hello. My name's Andreas. I'm the, uh, founder, uh, SEO consultant of. RTOS, and I'm also the writer of this book here, uh, Data Driven SEO with Python. And, uh, yes, uh, just as you said, Olga, I've been working for over 20 years in this industry. So I've worked on, um, very small companies all the way to your mega enterprises like Amazon's and all across different industries.
[00:01:59] Andreas: So, uh, yeah, it's been a very interesting time.
[00:02:04] Olga: Yeah. Yeah. Like 20 years. Yeah. So what were the very, very beginnings then? Like 20 years ago, how did you land into SEO? Uh,
[00:02:14] Andreas: by pure accident.
[00:02:17] Olga: Like everyone, I think.
[00:02:18] Andreas: Yeah. I, so I was, um, I was actually working as a qualified accountant at the time. And, uh, and I met some friends.
[00:02:30] Andreas: Uh, we went out on, uh, for an evening, it was speed dating actually, and there was this guy there, who, I don't know, I may regret sharing the story, uh, so I met this guy, and he told me he was making money from his bedroom, and Money from your bedroom. That's very interesting. So he told me how he's got this website and it was earning, uh, money, uh, automatically, you know, you'd, um, you'd Google, uh, this keyword term.
[00:03:02] Andreas: I'm, I'm keeping it anonymous just to, uh, Out of respect to him, you'd google this, um, this, this, uh, the search term and he had three websites in the top 10. I thought this is really interesting. Then you would go to his website and you would upload your file. Uh, the website would, you leave your credit card details and and within minutes Your file has been repaired and uh, you know, this was automated income and gosh, that's really interesting I was like, how'd you get this traffic?
[00:03:36] Andreas: How'd you get to the top of google? And that's where my journey started really if i'm being honest I dreamt of making money in my sleep. Yeah.
[00:03:43] Olga: Yeah the same with me I also I came across an article Exactly about that making money online and it was like It was for me. It was like 12 maybe 15 years ago when I heard about it for the first time so adsense was like Could bring tons of money back then way more than it can now I think so Yeah, the same the same story.
[00:04:08] Olga: So so what so you created your own website there?
[00:04:12] Andreas: Yeah, I Yes, well, I I created my own website just to learn how to build a website and learn about seo I probably like most people at the time I read a lot of Webmaster World, I started experimenting with what I read because obviously not everything you read is true.
[00:04:33] Andreas: I think that's true today as it is of yesterday. Yeah. So I just started testing what I learned, I started going to PubCon in London. And, uh, I just, nobody wants to talk to me when I first came. I couldn't blame them if I'm being honest, cause I, I didn't know anything. So I didn't bring anything to the table in terms of SEO.
[00:04:56] Andreas: And then I just said, right, okay, well, I started trying to work out who knew what they were talking about. And quite, quite honestly, I, I just started writing checks to say that, um, I clearly know nothing, but. If you for a few hundred quid a day, could you train me which was a lot more money than 20 years ago Than it is today.
[00:05:20] Andreas: Uh, you know, could you just teach me what you know? So I did this, uh, you know approach several people and learned from different people and that I think that helped accelerate my learning because obviously, you know it cut Potentially five years of trying to work it out myself to actually, you know, getting all the, it wasn't just the fish they were giving me that the knowledge that worked today, but it was also the mindset and the thinking, uh, as to how SEO worked, how search engines worked, what the principles were.
[00:05:54] Andreas: So all of these things, I think nowadays we're much more luckier because you know, there's a lot of, uh, guides, uh, expert guides, um, expert courses that you pay for, but at least they are there. They weren't there 20, 23 years ago. Um, so it's, yeah, that's basically how it got going and, you know, experimentation, testing, uh, that, that's, that's really, that's really how I started my journey.
[00:06:25] Olga: , you had, you had different jobs, right? And Oh, yeah. For how many years you have been your own boss,
[00:06:29] Olga: yeah.
[00:06:30] Andreas: Yeah. So it's quite complicated because I was kind of on my own from the beginning. Okay. Okay. And, and, uh, we had this regular customer at my. Parents restaurant and he wanted me to teach him how to use the internet and he was okay It's not that he he was not very clever or anything he clearly was because he he got kicked out of school and He ended up with you know a successful business selling Uh, conservatories and double glazing and things like that, but he just didn't understand how computers or the internet work So so anyway, I would spend these friday afternoons Uh after my job finished Teaching him how the internet worked and I told him about seo and what I was doing and what I was learning And he seems a lot more interested in that than The internet so he said look I know you're quite well paid as, as an accountant, but quit your job, I'll match your salary and we'll figure out, you know, how we can turn this into a business.
[00:07:37] Andreas: So, effectively that was like my first SEO agency or consulting. Okay. So he sponsored me. Now the business folded after one year. So, uh, that's because, uh, and when I look back, it's because I didn't know how to sell. I was very passionate about it to accelerate my learning further in SEO, uh, to help them get to the top of Google for, uh, Competitive terms like double glazing.
[00:08:06] Andreas: I simply hired a company in India and saw what they were doing as well. And that also helped me. So I got a lot of information from many different sources, if you like, to see how SEO works and, and yeah, that, that also helped me along my journey. Okay.
[00:08:25] Olga: Okay. And can you tell me more about the experience you get, like the, the variety of projects?
[00:08:31] Olga: You mentioned some, so like what types of niches, what types of projects, like e commerce, affiliate, like what were the types of SEO you were doing, which you enjoyed the most and what is it that you focus on right now?
[00:08:47] Andreas: So I had a real variety a lot of my experience was b2b And as I became more confident, I started getting exposure to e commerce um e commerce websites and businesses And that's obviously a lot more technical because there's a lot more that could go wrong on an e commerce site Yeah, and then i've I've been doing things on jobs boards, uh, one of the top 10 jobs boards in the UK.
[00:09:15] Andreas: Uh, so I mean, yeah, there's, you know, my experience does span both e commerce and B2B. So, uh, I love e commerce, uh, in the sense that there's more technical issues to solve and. And obviously it gets more complicated with JavaScript and, and just the way the world is going with apps and things like that. So that's quite interesting.
[00:09:40] Andreas: And, and some of the things that I do, uh, is trying to model which pages are generating the revenue, but are not quite getting enough of the, uh, if you like page rank. And not giving enough visits from the search engine so um That that's that takes up quite a bit of my time on the commerce side of things
[00:10:05] Olga: Mm-Hmm. . And I heard you also had have some interest in Blackhead, uh oh yeah. Part of a CEO PBNs, like, can you tell more about this? So that was ,
[00:10:17] Andreas: that was much more, in the earlier days, you wouldn't be, uh, caught dead doing that on the, on the brand, uh, Uhhuh, unless it was casino related. But, um, yeah, I, so you know, in the early days, Phantom and Phantom surfer.
[00:10:33] Andreas: They had a product called, um, this is embarrassing now, but they had a product called the Phantom Master and what it did was it generated, um, what we call spider food pages, right? So, and they had enough, they had a databank of IP addresses. Um, you know, of the search engines like Google. So they knew their software knew when, um, when these search engines were visiting your spam sites and, and when it was, and if it was a search engine, it would just show, um, it would just show the sort of spider content.
[00:11:08] Andreas: But if it was a human visitor, you could redirect it through an affiliate link to a merchant site. Um, and it would rank pretty high in the search engine. So, uh, now. Uh, a lot of affiliates were using their products primarily for the purpose of making money off affiliates. But I was, and still am, a terrible affiliate, which is why I don't do it.
[00:11:31] Andreas: Uh, but it gave me invaluable lessons as to, you know, what the limits were in SEO. So that was really interesting for me because, uh, you know, I, I think it's one thing to do, no best practice, but I think if you can really push the limits. Uh, without doing any moral or ethical harm, then, uh, you know, it just increases your knowledge.
[00:11:53] Andreas: Like some of the things that I, I learned were how you could get an index, a new website index by linking to it for your XML sitemap. So that was quite funny. You know, you could even effectively link to sites for an XML sitemap of an established site. Yeah. Yeah. So it's just stuff like that.
[00:12:14] Olga: Uh huh. Okay.
[00:12:15] Olga: Okay. And now what is it mainly regarding seo that you do now? What what services you offer?
[00:12:24] Andreas: Uh, so just holistic consulting everything from the technical auditing side of things to um to content optimization to digital pr so helping Clients, uh, well creating that content for them creating the research You That will help their content add unique value on the net and then promoting it via outreach to to get it into Uh the top five of the serps and then helping it learn backlinks automatically or autonomously
[00:12:58] Olga: And do you have a team or is it just
[00:13:01] Andreas: you?
[00:13:02] Andreas: I do have a team, but I'm really the front, front of house. So people generally speak to me, but I'm hoping that will change in the next two to three years where I'll, I'll have senior consultants working alongside me.
[00:13:18] Olga: Okay, what is data driven SEO? Tell me from start to finish. What does it mean?
[00:13:25] Andreas: Yeah, so data driven SEO is really just applying data science to to the SEO problem or challenge.
[00:13:34] Andreas: So, you know how we've got these tools like your crawlers, your SEM rush, uh, your Ahrefs. I don't want to leave any major tool provider, otherwise it might get upset, but it's basically analyzing all of these. All of this SEO data and looking for patterns, um, that are verifiable and repeatable so that when you make the recommendations, they're reliable and they're actually likely to deliver an improvement in your traffic.
[00:14:05] Andreas: So, uh, it's just really using statistics. And, uh, I mean, when you think about it, a lot of these tools, they just give you loads of data, right? But the data science is that extra, that, that, that next step, if you like, as to how do we analyze that data ultra efficiently so that we can, Do better SEO. So that's what I don't know.
[00:14:31] Andreas: I don't have taken the very long way around of explaining what data driven SEO is, but hopefully that explains what it
[00:14:38] Olga: is. Yeah, but to a very beginner. Okay, so I have. data from Ahrefs, Temrush, GSC. Give me a specific example. Like what can I do? How can I do it? What can I learn?
[00:14:53] Andreas: Yeah. So using data driven SEO, for example, you might want to work out which bet, which backlinks were the best, you know, now it's not, you know, best practice, if you will.
[00:15:06] Andreas: Might say, well, the best links are the ones with the highest domain authority, for example. But often I've seen cases, for example, uh, using machine learning where, uh, I found that in the furniture space, the best backlinks were those that got the most shares from Pinterest. And it actually showed that, um, the machine learning model showed that if the Pinterest, if the link from Pinterest didn't get.
[00:15:37] Andreas: More than, say, 300 chairs. Then it wouldn't be as powerful. So it's just stuff like that. I think, um, I read in your newsletter, actually, that, uh, you were talking about Quora. And that's the kind of stuff that data science does. Except Quora have managed to package it in, in the software.
[00:16:00] Olga: So do you use Quora as well?
[00:16:02] Andreas: I don't use Cora. I've got my own, uh, I've got my own systems that I've built, uh, and started building, um, in 2015 when I started getting into machine learning. Um, so I don't use Cora, um, but I would imagine Cora is a good system because it would be based on the same foundations.
[00:16:24] Olga: Yeah. Yeah. I think so. I think so.
[00:16:27] Olga: So data driven SEO with Python. So Do I have to know Python? Do I have to learn how to code with Python to be able to be a good SEO who also can make good decisions driven on data?
[00:16:46] Andreas: Strictly speaking, no, because I started learning data science using a language called R, which is literally the letter R.
[00:16:56] Andreas: And it's very similar to Python. So you don't have to learn Python. However, I'd recommend that you learn Python because it's the language that everybody uses. And, uh, it's just much easier to, to share your, your code, if you will, so that other people can use it and understand it. Um, I mean, I've written a book on it, but it's probably more on the science side.
[00:17:21] Andreas: Uh, I would also say that it's also helpful that the more you can increase the knowledge in the area of. Uh, of statistics and mathematics, then your Python becomes more powerful and let me give you an analogy to explain what I mean. Um, if you pick two people, okay, they both, um, but one's read zero books, the other one has read a thousand books.
[00:17:50] Andreas: Who do you think has more to say in that language?
[00:17:54] Olga: I guess the second one.
[00:17:56] Andreas: Yeah, exactly and it's the same with python right because you can learn python and you can be equally fluent by the person Who knows more mathematics or computer science is going to have more to say in python than the person who hasn't learned mathematics so Uh, I think the stats side Is very important as well Uh, I'm not saying don't bother learning Python if you don't want to learn stats.
[00:18:23] Andreas: By all means, learn it because it will still give you the appreciation as to how a search engine might process a website or a webpage. But if you can learn the math side as well, then it gives you a deeper appreciation as to how a search engine might work in total. Uh, and you know, there's quite a few courses, uh, that are on, that are on the web now and in person that can teach you these things.
[00:18:48] Andreas: Can you recommend
[00:18:49] Olga: any?
[00:18:50] Andreas: I haven't been on any. Uh, I've already, I've just finished filming a course with O'Reilly Media on data science SEO. Um, but there are other really good courses. So without the self nomination, uh, there's Elias, Dabasses, I hope I said his name right. There's his course on data science SEO with Python.
[00:19:11] Andreas: Uh, again, I haven't tried it, but it seems like he's getting really good feedback and, and I trust that he's doing a great job. So that would probably be a really good starting point, uh, for, for people just starting out in, in the data journey for SEO.
[00:19:28] Olga: And can you give me like a real life example of an SEO project where you applied data science and what you learned and what the outcome was of that learning?
[00:19:43] Andreas: Yeah, sure. Um, it's quite a few to pick from, but I'll pick a recent one. Tell
[00:19:49] Olga: me more than one because
[00:19:50] this
[00:19:50] Andreas: is like, yeah. But there's quite a few. So I'll start with the first one then. So the first one is I was working. I'm still working with a, uh, an online retailer that sells, uh, fashion jewelry and, um, you know, with thousands of products, uh, we, we wanted to reshape the information architecture so that more of the revenue generating products.
[00:20:15] Andreas: Would rank higher and obviously be more visible So what I did was I looked at their internal and external page rank And then what I did was I compared the statistical distribution of that page rank With the distribution of their revenue. So what I was looking for was Was for high revenue pages that were Uh, pages that, that had low page rank, but very high revenue.
[00:20:48] Andreas: So what I was looking to do was to redress the balance of pay the ratio of revenue to page rank. And so that's an example. So from that, I was able to create a solution where, uh, we managed to throw more internal links to, to those higher revenue pages that were, that weren't getting many internal links.
[00:21:11] Andreas: And the result was? Uh, the result was, um, well. We beat their monthly revenue target by 76%.
[00:21:21] Olga: Oh, nice. Nice. Very happy. Yeah. I can imagine that. Okay. Okay. And, uh, any
[00:21:32] Andreas: other example? Yeah. Another example is that, um, you know, I worked with, uh, another retailer, separate one, but again, in fashion, uh, they got downgraded.
[00:21:46] Andreas: Okay. Um, they, they simply lost a lot of rankings, uh, in, uh, following an update, uh, last, later last year. And so what I did was I took a load of data. Uh, before and after just to find out, um, you know, where they lost the traffic. And, and then what I did was I took a load of data on the competitors, uh, and then slammed it into a machine learning model to find out what could explain the differences in ranking according to offsite and on page features.
[00:22:27] Andreas: So, uh, and that gave us a lot of useful insights. And we managed to plus once we knew what that was, we, you know, there's hundreds of rank factors, but what the machine learning model did was it allowed us to look into where we should focus our efforts, you know, and what we could. Uh be confident would work because you know best practice seo there's there's hundreds of things you could do And we we managed to recover 80 of their traffic following Yeah, so they were very happy with that.
[00:23:01] Olga: And what were the things that you should focus on? What what the machine learning model tell
[00:23:08] Andreas: you? Uh, well in this particular case, there were some really odd things like not branding your title tags
[00:23:16] Olga: You Okay,
[00:23:17] Andreas: so that that was one. Uh, the other one was thin, thin content. They were putting out loads of PLPs and PDPs, but not putting any content on there.
[00:23:28] Andreas: So that was something that was picked up as well.
[00:23:32] Olga: Okay. Okay. And talking about machine learning, machine learning models. So do you like create your own? Like, how do you do that? Yeah. So what
[00:23:41] Andreas: I do is, um, I, I don't write them from scratch. I'm still using libraries like, um, you know, uh, XGB boost. So these are, you know, open source models, but in terms of coding it and deploying it, yes, I have to write that from scratch.
[00:24:01] Olga: Nice, okay Okay. Okay. Tell tell me like more details more. Yeah This is like geeky stuff.
[00:24:09] Andreas: Yeah. Well, the the funny thing is is that the models itself are um, uh, That that's that's the sexy bit and that's the last five percent But there's a 90 the 90 percent if you like that that is the hard work and that is getting the data using APIs to get on site data.
[00:24:34] Andreas: So, um, you know, in getting SERPs data as well. So for example, a lot of, um, a lot of, uh, APIs, a lot of rank tracking tools like rank ranger, for example, uh, they offer an API and you can collect, um, you can collect SERP data. Um, So, so you can see who's ranking and what the ranking URLs are, and then you can use those to go and get the data on those URLs so you, you can crawl them.
[00:25:02] Andreas: And again, it's, if you, if you imagine a giant spreadsheet, you know, you've, you've got the URL, you've got the keyword, the ranking position, and then you've got like, say, word count, title tag. Uh, all these different things, the core vital score. So we're just trying to get as much data as possible on each of those URLs, just to see, think of it as hypotheses that we can test much like Cora probably does already.
[00:25:31] Andreas: Yeah, yeah, yeah, yeah, exactly. That's all it is. But I'm just doing it by hand using a self coded machine learning
[00:25:38] Olga: model. Yeah, yeah, I think so because Cora correlates, I don't know how many, like, I think more than 100, 000 factors, kind of, measures correlates. And, and yeah, this is, this is like, this has Done so much heavily heavy lifting and I don't have to be such a geek Such a pro when it comes to data analysis as you are and I can still Get the leverage from for example from this tool 100 yeah, and uh Tell me more about your book.
[00:26:17] Olga: What is it like? Is it like an intro? Is it for more intermediate SEO analysts? Like, do you have other books? Like, tell me more about it. Yeah.
[00:26:27] Andreas: So, uh, so this book, um, I'd say you'd be expected to be an SEO expert because if you don't know SEO, it's not going to be very interesting for you. But it does, you don't even need to be a Python coder really to, because there's code in the book that you can Literally used to start building your own scripts to to start doing your own machine learning your own You know, it covers all of the different work streams of seo from keyword research to algorithm recovery uh content recommendations page rank, all of those sorts of things.
[00:27:08] Andreas: So, um, that's the way I structured the book because I want it to be a useful manual that you as an SEO could dip in and out of, you know, uh, you know, I want to scale up or automate, or, uh, I want a more statistical approach to keyword research, for example, uh, But the chapter on keyword research will show you how to get data from Google search console how to process it how to Estimate the maximum number of impressions for example How to do forecasts, you know, if you want to do forecasts for your reporting So it's got all all bunch of goodies in there that you can do SEOs will want to know.
[00:27:47] Andreas: So, uh, you know, I've had some very good reviews. Somebody said that it's criminal how much I'm giving away. So, you know, yeah.
[00:27:57] Olga: Because I haven't, I haven't read your book yet, but definitely on my list now. Well, hopefully if
[00:28:02] Andreas: we meet in Barbados again, I'll uh, if you want me to I'll sign it
[00:28:08] Olga: Okay. Yeah, sure, of course Maybe
[00:28:11] Andreas: we could do a podcast there
[00:28:13] Olga: Yeah, yeah, like with with with palms and the ocean in the backgrounds.
[00:28:18] Olga: Yeah, hopefully no rain
[00:28:20] Andreas: 100 yes
[00:28:22] Olga: Yeah, yeah, perfect Okay, so you have a lot of powerful data about SEO, how search engines work. So with all that knowledge, like what's, can you tell me more about what's, what has been going on with Google in your opinion for the last, I don't know, half a year, maybe even, even longer since like the helpful content update has, um, has been there.
[00:28:48] Olga: Like what are your observations, field observations?
[00:28:52] Andreas: Well, I think what's coming quite, uh, obvious to me, if you like, is the, is the, um, is the emphasis on proprietary data, it's definitely becoming more data led. I mean, we're seeing the data set schema appear more in, in the search engine results. I think the other thing is, is that the, if you, if you really want to succeed in.
[00:29:19] Andreas: In, um, you know, search going forward. And today is that you need to bring something unique to the table. So if you could bring proprietary data to the table, because as I'm sure, as you know, there's a lot of like, you know, must, must know trends about target keyword X, you know, it's been in the web. We now got chat GPT, uh, you know, we have a chat GPT enabled webmaster web.
[00:29:46] Andreas: So, you know, back in the day it was automated content spinning and then it became human spinning.
[00:29:55] Olga: And now we have, yeah,
[00:29:57] Andreas: yeah, we have chat GPT. So it's on steroids now. Uh, I don't totally buy that. Um, you know, I, I don't think we should be asking ourselves. Uh, how can Google tell the difference between, um, content that's been chat GPT, uh, generated versus human written?
[00:30:19] Andreas: I think the question should be, um, you know, what, what is it that, what is it that separates content from delivering the right user signals? To those not delivering the right user signals. So I think that's the other aspect so proprietary content But also use the signals and I think the key question for us is, you know Uh, and maybe these two are linked, you know What is it that we can do as seos to make sure that when it comes to machine learning from a google perspective?
[00:30:53] Andreas: You know, they're going to do stuff like uh Clustering, okay. Clustering of content that has the right user signals versus those that has the wrong user signals. So what are the things that we can do to make sure that our content lands in, in the smaller cluster of the correct user signals?
[00:31:12] Olga: Yeah. Yeah.
[00:31:13] Olga: Because like at first when chat GPT became popular, like people were talking about, um, that Google will, how Google will know if my content is AI generated, then we have all those AI detectors, people trying to check whether content was Created by humans or AI, of course, those AI detectors do not really get it right.
[00:31:36] Olga: And even if they did, Google removed even that, uh, that sentence from their documentation where they stated that AI content, they don't want AI content. They, they basically changed that. They, that they care about contents created for people, not, not by people, something like that. Anyway, they, they kind of admitted that they don't care if it is AI content.
[00:32:00] Olga: As long as the content really does what it's supposed to do and even if we Let's even assume that google is after ai content and google Actually wants to check whether your content is ai generated how much resources right? It would take google to even be able to detect and
[00:32:19] Andreas: exactly. Yeah, like Exactly.
[00:32:22] Andreas: It probably takes five percent of the total resource required Uh, i. e. it's 95 percent more efficient to just see Is the content delivering what users want? Yes or no, you know, and it's far simpler So i'm not surprised that they retracted this and it's probably because they improved their own internal systems on being able to Focus on the right things rather than the wrong things
[00:32:45] Olga: Yeah, yeah, totally, totally.
[00:32:47] Olga: Okay. Okay. So tell me now a more personal question, like what does your normal SEO day look like? Like, do you, do you work like 16 hours a day or maybe three hours a day?
[00:33:03] Andreas: Uh, well, I'm getting busier, so I do, I have had to pull the odd one or two AMs. Uh, there was a couple of weeks. I've had a few 2am's, but, um, yeah, uh, there was one Friday, which where, where I worked until 4.
[00:33:21] Andreas: 30am, so, uh, luckily that's a one off, but generally speaking, you know, I, I do long days, but I try to break them up with small breaks here and there. So that, that's, that's, uh, and I generally spend the first part of my day is, you know, once, once you get all your sort of exercise out of the way, I, I look at across all of the clients, um, I've built systems to, Check very quickly what the performance of the clients are doing And that's quite good because obviously it keeps my finger on the pulse I think it's very difficult to be tuned in unless you're doing that So, uh that that's one of the things I do and then obviously I have to work through my list Not just delivering on client projects, but also running a business as well
[00:34:15] Olga: So you mostly now have the client work, of course, running business, or do you also like build your own websites?
[00:34:22] Olga: You mentioned you're not really into affiliate SEO, but do you also like have your own sites?
[00:34:29] Andreas: Yes, I have a few sites. I would call them more hobby and testing sites. Uh huh. Uh huh. So, uh, I am interested, I'm not going to say what they are, but I, uh, I, I'm very much, for example, I use machine learning to, to, uh, uh, for picking my stocks for buying and selling.
[00:34:50] Andreas: So I really enjoyed that. It seems to be reasonably profitable so far, touch wood. Um, I'm also, uh, into cooking, uh, I don't really post on that, but, uh, I've got, so my parents own restaurants, so, you know, whilst you can take me out of the restaurant, you can't take the restaurant out of me. So, uh, I've got over a thousand cookbooks and I love cooking at the weekend.
[00:35:17] Olga: But are they, like, your cookbooks, or?
[00:35:21] Andreas: Oh, no, no, these are cookbooks that I've
[00:35:23] Olga: bought. Oh, I thought that maybe you wrote them
[00:35:26] Andreas: or something. Oh, no, no, no, no. I tell you what, though, I would really love to write a cookbook. Uh huh. For
[00:35:31] Olga: sure. Do you have a cooking website?
[00:35:34] Andreas: Uh, no, no, I don't. But, um, you know, I, I've, uh, I've written a few recipes on Google Docs and stuff.
[00:35:41] Andreas: So, uh, you know, if you ever want a cooking tip on how to make the perfect steak, uh, I could definitely advise you.
[00:35:49] Olga: Oh, yeah. Yeah. I used to have, I think I still have a cooking website, but it is, it has been like running on its own for six years without me touching it. But I
[00:36:04] Olga: don't, I don't even know if I, I can log into that without like trying to change the, user and password in database because like I have abandoned that site, but it's still, but it still has been bringing some, some money from ads, but like totally on autopilot. Okay. Yeah. Yeah. Yeah. I wish I could create more such sites.
[00:36:28] Olga: If only, if only I had a little bit more time. So yeah. So talking about AI, like what's your take on that? How do you use AI, chat, GPT, Claude, Bart, Gemini, like. Is it like part of your SEO process? It
[00:36:48] Andreas: is, yeah. Like, for example, if I'm tasked with having to optimize, say, hundreds of webpage titles, I definitely use AI to generate the draft, uh, you know, with some inputs and that can be very useful.
[00:37:05] Andreas: But I do find But there still needs to be some human moderation. But still, you know, um, I think it's much better than staring at a blank spreadsheet and then, you know, having to write them from scratch. If you can edit something, you know, it can be, A, more motivating, and B, a quicker process.
[00:37:28] Olga: Yeah, yeah, totally.
[00:37:29] Olga: And I think I heard it, I don't know, this week that Google created this tool. I think it is a Chrome extension that, that is kind of your, your writing assistant, that whenever you are on a website typing into a form, writing an email, it can, Kind of help you. And I think this is an exaggeration because like, sometimes it takes more time to create something with the right prompt as if you were just to type the, like, I don't know, two sentences of, of a message to submit to, I don't know, to, to a website you, you want to ask about.
[00:38:07] Olga: So
[00:38:08] Andreas: like, It kind of reminds you of how things were Like in the early internet days, right? Yeah, we had dial up and then we had very Awful search engines before google came along and then, you know You just know things are going to get much better maybe not at the time but um, yeah Yeah, I can imagine five years from now, you know We yeah, it's almost like we get to the point where we just think what we want and then the ai just does it, you know
[00:38:37] Olga: Yeah, so five years from now.
[00:38:38] Olga: Do you think we seos will be out of our jobs? What what are your predictions for us?
[00:38:45] Andreas: I think it will be different. Um, I i'm i'm a real You Uh, I don't think SEOs are going to be like the horses were when the age of cars came along. So I, you know, we're, we're, um, I think we're best deployed when, you know, we're humans.
[00:39:05] Andreas: Right. So we're, we're our, our best. Skills and gifts are in creating and responding. So I don't think SES will be out of a job. I just think the job will be very different. You know, I think it will be okay. Well, how do the AI has said this, what is the solution that we should create? And, um, you know, there's already AI programs and there have been AI programs that can compose music, but can they compose music as good as.
[00:39:36] Andreas: Whatever your faith, you know, Taylor Swift, she seems to be the. The very fashionable, uh, popular, uh, musician these days, you know, will they ever be able to do as close to Taylor Smith? Probably not. But the point is, is that, uh, they won't be able to do SEO. I mean, they will, but they won't be able to do it as well as an intelligent SEO.
[00:39:55] Andreas: Yeah. So I, I, I think, I think it's, it just, in fact, if anything, it could increase the price of SEO because people will want SEO done well. Um, it just, it probably just increases the, the range of people that can benefit from SEO, but of course. Uh, it will just mean that you know If if you rely totally on ai the seo won't be as good but your pages will still be optimized Will you get traffic probably not but at least you'll be on the web.
[00:40:29] Olga: Yeah. Yeah, totally totally and Can you recall some weird seo situation you have been in recently or not recently something like weird totally weird
[00:40:40] Andreas: Uh, well this this this is quite weird. How about A friend of mine wanted to become, uh, Britain's best looking chiropractor. Okay. So what I did was, I optimized his image and his body.
[00:40:58] Andreas: Uh, Batlink profile, so that when you typed in Britain's best looking chiropractor, or even best looking chiropractor, his image came up in Google. Oh! Okay! It's, it's weird, but, uh, that's the answer I have for your question. I'm sure there's more weird examples, but that comes to
[00:41:17] Olga: mind. Yeah, yeah, that's a good one.
[00:41:20] Olga: Okay, so what advice would you give to people who are just starting out in SEO?
[00:41:28] Andreas: I think, uh, the advice I would give, and it hasn't changed in 15 to 20 years is, is that, uh, I say 15 because obviously I wasn't an expert 20 years ago. So, uh, I would say you, you should. Build your own website, right? Because and you should optimize it because if you don't know How to build your own website.
[00:41:53] Andreas: I'm not saying you have to become a master developer, but at least if you you know It's pretty easy to build, you know Install a wordpress for example and then start putting content and optimizing it and one of the things I I used to do Is I used to optimize For brat for for completely Uh made up keywords just to start very easy and I think it's really important for seos to do this because if you're an seo consultant or even if you work in the company and you're you're going to make Recommendations to get company website changed, you know, there's there's there's quite a bit out of state there There's your company's livelihood or your client's livelihood.
[00:42:33] Andreas: So if you're giving advice on something that you don't know, that you have no direct experience of, then that doesn't seem ethically correct. You know, can you imagine taking driving lessons from someone who's never passed their driving test or hasn't even driven a car? Wouldn't that be weird? I guess . So, so why would we take SEO advice from people who have never built and optimized the website?
[00:43:00] Andreas: Even if you've had your website built for you, why would you take advice from someone who doesn't optimize websites? Uh, yeah, firsthand, you know, it doesn't make sense. So that would be my advice. And obviously, you know, the, the more you can learn, the better. I mean, uh, you know, I'm not a website developer, but I still learned how to build a website.
[00:43:20] Andreas: I still learned. How to read and write HTML. I learned how to write content, you know, okay. I've written a book, but I'm not a content writer. You know, I don't know. Uh, I don't have the tools of English literature of storytelling. You know, I don't know about. I know what tone of voice is and what the use of it is, but I wouldn't know how to put a tone of voice together.
[00:43:44] Andreas: Um, so I think it's really about learning as many different things as possible. You know, I've had to learn digital PR from scratch 20 years ago. You know, we started with reciprocal links and triangulation links, and then, you know, PBNs. Blogger outreach then it transitioned to the blogger outreach, but you know I had to learn that from scratch because there were no kind of writing SEO courses on on these sorts of things, you know, so I think the more you can learn about the different aspects I even did Paid search marketing, uh, once upon a time.
[00:44:26] Andreas: So I think that that was really useful because you get to learn more about the sort of higher intent buyer ready, um, keywords, uh, you know, and it gives you an idea of, you know, it's just the more different ways you can look at SEO. I think that the more perspectives you gain and the more value you can bring to the table when it comes to advising on SEO.
[00:44:50] Olga: Yeah. Yeah. Perfect, perfect advice. Okay, so Andreas, where can people find you? What's the best place to connect with you?
[00:45:00] Andreas: Uh, I'm on LinkedIn. I'm on XI am, uh, my website is roff.io. Uh, LinkedIn. I'm reasonably active and, uh, I sometimes write for, uh, on crawl or search journal. Um, so I, you can't really miss me, uh, um, on the web.
[00:45:23] Olga: Okay. Okay. And so YouTube, do you have YouTube
[00:45:26] Andreas: or not? I do have YouTube. Um, I'm slightly embarrassed that I don't know how well maintained that is, but I'm definitely on YouTube.
[00:45:33] Olga: Okay. Okay. Perfect. Okay. I need
[00:45:36] Andreas: to improve.
[00:45:37] Olga: Okay. Everyone has to, has to improve me too. Okay. Andreas, thank you so much for being my guest.
[00:45:46] Olga: Yeah, it was, it was a pleasure and a ton of things. I feel very well motivated now to get more into data analysis. Definitely. And I recommend everyone to reading the book Data Driven SEO. So yeah, thanks everyone and thank you. Bye. Bye.
[00:46:07] Andreas: Thank you. Cheers