E147: Lazarina Stoy

Richard Hill [00:00:04]:

Hi there I'm Richard Hill the host of eCom@One welcome to episode 147. In this episode I'll speak with Lazrina Stoy, SEO and data science consultant. Lazrina is a genuine expert when it comes to data science and SEO using machine learning, idea discovery, data extraction, interpretation to gain a practical advantage when implementing strategies. In this episode, Luzino talks, what does data science mean in the SEO and e-commerce world? How can businesses use data science to improve their SEO strategy? What are the common mistakes most e-commerce companies make with SEO and how can you avoid them? How can companies use their data science to stay ahead? Hi there, bum bum bum. Hi there, I'm Richard Hill, the host of eCommerce 1. Welcome to episode 147. In this episode I speak with Lazrina Stoy, SEO and data science consultant. Lazrina is a genuine expert when it comes to data science and SEO using machine learning, idea discovery, data extraction and interpretation to gain a practical advantage when implementing strategies. In this episode Lazrina talks what does data science mean in an SEO and e-commerce world, how can businesses use data science to improve their SEO strategy, what are the common mistakes e-commerce companies make with SEO and how they can avoid them, and how companies use data science to stay ahead of the trends. That's really this day-to-day tool set and her predictions for the future of SEO and data science. An absolute cracking episode. And of course, so much more in this 1. If you enjoyed this episode, hit the subscribe or follow button wherever you are listening to this podcast. You're always the first to know when a new podcast is released. Now let's head over to this fantastic episode.

Lazarina Stoy [00:02:25]:

Yeah, absolutely. So I am Lazarina, I work in SEO, I specialize in data science. So I'm a consultant. I mostly work in agency side, and I've mostly specialized so far in my career in B2B companies, enterprise companies, big tech, SaaS, and a little bit of e-commerce here and there. I've worked with enterprise e-commerce stores and also small businesses. So how I got into data science, I did marketing as my university major. And then I thought, Oh my god, this field can be automated in like 10 years time. And 10 years is right now actually. And now we're seeing that automation is all the rage. So it was correct. And I did a very good thing to study a master's degree in computer science and machine learning and kind of specializing in natural language processing. So I suppose my passion for data science was more to do with automation, analysis of data and kind of helping marketers make their processes and their work a little bit more automated and streamlined and anything that can be helpful is kind of where I have specialized since when I have worked in agency side. So that's where my passion comes really, just motivation.

Richard Hill [00:03:55]:

It's great to hear somebody actually credit their degree and their master's degree doing the thing that they went to university to do, which is quite, it's quite uncomfortable. 100%.

Lazarina Stoy [00:04:08]:

Yes, I joke quite often when someone asks me, well, do you like what you're doing? And I'm like, well, yeah, it is what I studied. So I kind of chose this path. So that's good. I suppose it's quite nice to be able to come out of the other end and see that there is actually the application for what I did.

Richard Hill [00:04:28]:

Yeah, actually. So let's talk about that then. So when we talk about data science and SEO specifically, obviously as an e-com store owner marketeer, which is probably 95% of the people listening right now, you know, what does data science mean in an SEO world?

Lazarina Stoy [00:04:46]:

Yeah, So I have a few different examples and things that I've spoken about before. But in general, the way that I think people can think of data science and e-commerce is more about understanding your users better, understanding the data that you have better, and being able to incorporate this data into your SEO strategy. So at the moment, we don't really have the point in time that we can completely automate an SEO strategy or we can completely automate the process of idea discovery for content or for products. But we can get pretty close to that point, and you will still need a consultant, you'll still need someone to help you out. But the process can be a lot more productive is what I'm trying to say. So when I personally think of the opportunities that relate to data science and e-commerce stores, the first thing that I think about is just all the data that users provide. E-commerce stores that I've worked with don't really incorporate that into their SEO strategy very well. And what I mean by data that users provide, it could be a review that they leave for your product. How are you utilizing this data? Are you doing sentiment analysis on it? Are you doing entity extraction on it? Are you incorporating it into the process of content creation? Are you incorporating it into the process of new product discovery? All of those cool things can be potentially data science projects. And also social comments recently are something that I've been fascinated about how this can be incorporated into an SEO strategy. So if you have a brand that's an e-commerce brand that has a very good performance on TikTok or YouTube, how are we utilizing the comments that users are leaving on those platforms to extract new product ideas or to extract ideas for new content. And of course, that can be an SEO project, but it can also very easily be incorporated into the overall strategy of where that particular shop is going to go within the next few months. So I'm quite a big advocate of moving away from SEO and kind of becoming a little bit more involved into the business side of things, especially when you're working with e-commerce shops because at the end of the day, the site is part of the product and you need to have a good business direction in order for the SEO to be good and vice versa. So yeah, I think in general, if I have to kind of sum it up, I know I've given quite a bigger response. But I think this data science can be a very good way to help emphasize competitive advantages that the brand has and to kind of bridge the gap between the user and the brand, especially because it allows the brand to find patterns of what users want from it, new products, new content ideas, and in the longterm, especially for enterprise brands, it can be a very good way to kind of make them a little bit more competitive in the market. Oh,

Richard Hill [00:08:00]:

I love it. There's a lot of things there. So I'm just, I'm thinking which, where to go with it now, because I think, um, yeah, using the data and using information for all the different areas, whether that's reviews, user content, um, your own data, But maybe let's drill into sort of content strategy and data science and using data to help create a more structured content strategy. I think I know not just e-commerce, but in all sites, you know, we see time and time again a lot of challenges creating content strategy,

Richard Hill [00:08:35]:

a lot of data points, a lot of tools out there to use and get data from, whether that's search console or third party tools. Hi there, I just wanted to take a quick break to introduce our sponsor, Pricynk. Pricynk is a competitive price tracking and monitoring software that can dynamically change your products prices on all sales channels. They work with brands such as Samsung, Sony and Suzuki to increase their online revenue. If you run Google Shopping, which I know a lot of you guys do, this software is absolutely key to accelerating profit. 1 of the reasons I recommend Pricynk to my clients is because you can find out your competitors pricing and stock availability in 1 simple to understand dashboard, giving you a huge competitive advantage. If you have any questions about this software or you are ready to get started with a free 14 day trial, head to econ1.com forward slash contact and complete the inquiry form. And we will connect our listeners to the pricing team. Right. Let's head straight back to this fantastic episode.

Richard Hill [00:09:24]:

You know, and then, and then it's like, right. Okay. As an SEO or as a e-com business owner, marketeer, it's like, right. Where do we go with a content strategy? Now, what would you say about sort of content strategy and using data science specifically?

Lazarina Stoy [00:09:41]:

Yeah, so I use a few different approaches when it comes to just understanding user queries or keywords. So it could be in the process of how you're analyzing how you come up for users. So using Search Console data, But it could be just utilizing a third party tool to see where other opportunities are. So in the process of when you compile your keyword universe, regardless what sources you use, 1 very good thing to do, obviously we all know as SEOs, the importance of forming a content strategy around content clusters. So a very good technique to actually find out what the clusters are is to use Ngram analysis. So just understanding, you know, what are the different components of queries that are most frequently mentioned and then utilizing that to form your clusters. So it can also be a very good way to understand how a particular entity is mentioned. So if we're talking about shoes, are people wanting to compare in terms of using a best type keyword or are they comparing in terms of pros and cons and things like that. So what I'm trying to say is the first thing, I think at least as part of a good content strategy is just understanding the language that users are using, understanding the intent behind the keywords. So I've done quite a lot of different guides and dashboards and sheets, templates, and all of that stuff about identifying intent, regardless whether you're using something like SEMrush that provides this label or not. And you can kind of start classifying intent as well. I also always recommend for brands to have their own intent classification related to the products that they are serving on the site. So you have a little bit more complex categories for search intent. So it's not just informational and transactional, but it's more related to, is it related to a product that we're selling? Is it related to customer service? Is it related to anything else like shipping and so forth? So all of these things, I think, are amazing data science projects that sometimes can involve big data. And yeah, definitely most of them are beginner friendly and highly recommended

Richard Hill [00:12:11]:

for e-commerce brands. So as an e-com brand then, listening, we're going to look in our, let's say, I'm just gonna step through that a little bit. So we're an econ brand selling, let's go with barbecues. You're selling bar, I'm actually looking for a barbecue.

Lazarina Stoy [00:12:26]:

Go with something I have no idea about, sure. I'm kidding.

Richard Hill [00:12:30]:

Oh, let's try, Let's see how we get on. So I'm looking for a barbecue and I've been looking at all the different brands. Obviously there's brand, there's size, there's, you know, and then you've got gas powered barbecue, you've got a charcoal powered barbecue. Ultimately, you've got a store selling barbecues. You're looking your Google Search Console for queries that are already happening or are already there and then you're classifying them into different sort of areas of intent. Is that right? Yeah.

Lazarina Stoy [00:13:00]:

Yeah, yeah, absolutely. So I think you have given us a very good direction in terms of how we can approach it. I think taking the point of view of a user when you're doing your content strategy is super important because, you know, for the barbecue, you have so many different features that you want to make a choice between. So as an online shop, you have to explain what those different features are, maybe in the blog. So as part of your blog content strategy, But then when it comes to your product pages, you have another opportunity to actually provide more information about your own products for the same features. And then if you're really good, you're going to internally link between the blog and your product pages. So all of these different things. Of course, if you don't have good keyword research and a good understanding of your product, your business model, how you compare within the competitive landscape, your content strategy is not going to be as good, right? So yeah, absolutely, analyzing keywords, and then building up a content strategy that provides a little bit more information about the competitive landscape. Of course, enterprise brands might have some limitations to talk directly with direct competitor comparisons, but there is always a way to do a competitor comparison without actually mentioning competitors and so forth.

Richard Hill [00:14:36]:

So we would refer to that I think at our end like a topical map.

Lazarina Stoy [00:14:40]:

Yeah, absolutely.

Richard Hill [00:14:42]:

Yeah. Right. You've been told what to do guys.

Lazarina Stoy [00:14:49]:

Easy peasy. Easy

Richard Hill [00:14:51]:

peasy. Yeah. But I think, you know, that's a piece of work that quite often, you know, it gets very, it's done very quickly, you know, by mistake, you know, and, you know, a handful of keywords and half a dozen blog posts, and then you're wondering why you're not ranking or you're not, you know, you're not, you're not getting the volume. So it's obviously, I think what we're saying there is quite a deep dive. And it's quite a, potentially a broad piece of work that's got a lot of different areas, you know, it's not 2 or 3 blog posts on this and 1 or 2 bits of content on a category page. You know, it's quite a deep topical map, you know, deep deep research, which might have literally dozens, if not hundreds of pieces of content, if not thousands, depending on the, you know, the business and the, you know, the competitiveness, and then building that out. And then obviously you have different types of content. Some might be more of a technical piece, which might be more of a technical writer or more of a technical research, et cetera. Some might be more creative and have more personality to it and might need different types of writers and things. So mistakes, I think, you know, what sort of things do you see that are quite common when it comes to SEO mistakes, you know, and how can our listeners avoid them?

Lazarina Stoy [00:16:03]:

Yeah, specifically when it comes to e-commerce sites, 1 thing that is quite common is not having enough categories and tags. I think that is, you know, just circling back to what we said a second ago, just a lack of in depth research when it comes to the product when it comes to how users might discover you. So I think that's the first step actually breaking down the different ways, or the different parameters that users might use to make a choice about purchasing a product or viewing certain sections of products, not just filtering based on price and things like that. I also think that, especially with big brands, there is the mistake of not aligning the product names with the way that people search. So there is this really careful balance sometimes where you're trying to establish your own brand voice and how you want your products to be featured on the site. And then there's also the fact that you have to incorporate when it comes to product names and even just the titles of the product pages. You have to align that with how users search. So in the agencies I've worked with, we've seen examples of just, you know, missing out on a ton of opportunities just by not naming a particular product the way that people search. So I think this is very much more of a soft mistake as opposed to something hard in terms of the skills and how you can go about negotiating and kind of aligning with all the different stakeholders that might be involved in that decision. But at the same time, I think SEOs need to be involved in that decision because at the end of the day, we will be benchmarked on the lack of traffic coming in. So we have to bring the reason of why that is the case. Something else that I've seen, especially with big companies, is just the lack of investment in technical SEO and site fixes. I think it's quite common to think that if you are using a little bit more of an out of the box solution with the big companies like Shopify, that it will work perfectly, which we all know it's not the case. There's just a ton of basic things that could hinder your SEO, that these companies don't get right from the get-go. So I think just, you know, as an e-commerce company, just having that budget in place for fixing those mistakes is quite important. And just 1 last thing that I want to mention is that sometimes smaller companies and smaller shops have the advantage of being closer to their customer, which I think is something that big brands don't really utilize. Just the bigger you get, you kind of stop investing that much in analysis of data to see all of the different things that your customers are talking about. And that's not only on the site, but also on social media or forums and things like that. So I think data science is definitely something that can help with these types of projects because it will allow you to set up systems in place that do social mesh and monitoring, that do analysis of these meshes, and actually can give companies that competitive edge that smaller brands might have, just because they have a closer relationship with their customers. So, so, yeah, I think those are not that many, but quite

Richard Hill [00:19:55]:

significant ones. That's great. I think, yeah, that really resonates with the, you know, the people we speak to, you know, quite often there's some very much a lack of investment in the technical SEO side of things. And it's sort of getting that impact across what the potential impacts on revenues are or will be with certain changes. That relationship between SEOs and devs sometimes can be quite challenging 1. Um, yeah, platform has its own nuances, doesn't it? That, um, you know, something like. Usually not that much time. Usually maybe a, you know, half a day, a day of technical work with a developer, technical SEO can mop up quite a lot of challenges, a lot of crawl issues maybe or have not been indexed.

Lazarina Stoy [00:20:44]:

I was just going to say half a day a week for most of the sites would be life-changing if they could get that dedicated to to SEO but I feel like not that many companies have that half a day. So yeah just massive backlog sitting with a lot of a lot of different opportunities. I feel like it's the same for content. Well, at least some companies, especially coming from an agency point of view, there are companies that think that if you work with an agency, then that automatically takes care of a lot of these costs, which oftentimes, especially for enterprise agencies, that's not the case. You still have to have the time internally to implement, regardless if it's content, if it's technical, you still have to have the dev resource or even the resource to hire a dev agency. So the consultancy on its own is not enough because oftentimes you might end up with just folders and folders of amazing strategies and content and all of the stuff that you actually need to fix all the issues. But if they're not implemented, then we're just, you know, so common is it's so so when we look at Yeah, I mean, a typical SEO audit can say, right, you need to do these 30 things. But obviously,

Richard Hill [00:22:02]:

they need to be ordered in terms of commercial return, you know, and priorities and probably broken into right. These 3 or 4 things are massive priority. This is going to get you an extra 20% in revenue or 20% in search visibility, which could then lead to X amount of conversions, you know, sort of hitch it in, in a sort of commercial term, as opposed to here's these 300 things that they're doing, or, you know, don't say, right, here's 15, 000 alt tags to sort out. That's not really going to, that's not going to inspire anybody to do any work, is it? But whereas if there's, you know, the amount of site, we've seen quite a lot of sites, I mean, I'm sure we could share a lot of horror stories, but, you know, sites with hundreds and hundreds of blog posts that just aren't indexed and, you know, bogus lines of sort of no crawls and de-indexing of pages in robots.txt and all sorts of silly mistakes. But So there, and then you talked about the naming convention of the products or categories, subcategories. I think that's 1 as a marketer or the owner of a store, you can be a very much see it from your own perspective because you know the products. So then you're calling a category, something that you know of, but out there in terms of the people that are searching, you know, I see that a lot with branded products where a lot of branded products, we might say, you know, they're selling a Nike size 12 Air Jordan, but they don't, they might not put the word Nike on the site because everyone knows that there is a Nike, but if you've not got it on the page, you're not gonna get, you're not giving a strong enough signal. So just a simple thing that, But yeah, great. So you touched on it as your third 1, using the data science and using sort of automations and reporting, but what can our listeners do to sort of stay ahead of the trends by using data science to sort of give them the edge?

Lazarina Stoy [00:24:01]:

Yes, when I think about how you can stay ahead of trends, especially at e-commerce, there are a few different projects that I have seen surfacing in the SEO world that I would love to share because they're, in my eyes, are a perfect blend between data science and e-commerce, the e-commerce world and how it all happens. So when it comes to the Google Trends data that you have for holiday searches and different seasonality related searches. That is a great data source to incorporate into your holiday related calendars for a contest strategy. And also the keyword data that you have from SEMrush, this is a great resource to analyze for the trends that happened in the previous year in order to be better prepared for this year's season, if that makes sense. So Just incorporating this is something that I often see as a missed opportunity. You know, if you're not doing that, if you're not analyzing how the previous holiday season went, you know, you are missing out on potentially capitalizing on these searches this season. So that's 1 way I think that companies can improve their competitive advantage by using data science. Another thing that I've been quite vocal about is that we don't quite utilize Google's natural language API as much as we should as SEOs. And it is created for us, you know, or at least marketers with kind of that data science, you know, mindset. So I think that, you know, doing entity extraction on queries, on keywords, on the reviews that users leave on Google Business profiles or Amazon, just collecting a big data set of what users are talking about and then utilizing that as a, of course, not replacing your entire content strategy, but as a component of it, I think is quite important because it does show that you care about what users are talking about and actually addressing those issues as well. And I think also just having the mindset of sentiment analysis and incorporating that into your content strategy, incorporating that into the way that you make decisions about your products. So I think this particular process is a little bit harder for SEOs that are in organizations where SEO is considered a side function. So I think here, specifically in order to become a little bit more abridged with other functions of marketing and in business in general, we kind of have to go out of the mold of being just an SEO or just a data specialist and actually tie in certain components of the strategy all together in order to help other departments make better choices. So I think although sentiment analysis is not a direct SEO project, it can certainly influence other business decisions that can improve the organic traffic of the site. And it can also improve conversion rates and other aspects that can bring in more revenue. So this is why I'm kind of honing that aspect a little bit more because I know how important it can be for companies, especially enterprise companies that have a lot of data at their fingertips and it's not utilized. So yeah, here are just a few, but and 1 last thing, I have spoken about this before, but just utilizing SERP analysis data, especially right now when we have so many different changes that are happening in the SERPs and just utilize it. Do SERP analysis audits frequently as an e-commerce store. That's quite important. Being tracking feature snippets and making sure that your website is optimized for structured data as an e-commerce store. I think that is extremely important nowadays. And this is how, you know, Google decides how to display you in a more advantageous way. So definitely something to do. Surprising how many

Richard Hill [00:28:33]:

SERP results have structured data in them, whether that's product reviews, whether that's a lot of FAQs seem to really dominate on the, when I'm searching for barbecues, it's a lot of FAQ data, price data, availability. Yeah, it's been an hour, probably it's a bit of a tip for you guys listening. You know, it's quite, I shouldn't say easy. I hate using the word easy, but it's reasonably straightforward to go and take over an FAQ, sort of structured data listing,

Lazarina Stoy [00:29:07]:

you know, with a- It's also extremely easy to automate the process of extraction of question and answer pairs using data science. So I actually have a script of doing that. You just copy paste the text of the article that it just extracts the QA pairs and automatically puts them into the structured data markup. So definitely using, people typically think of data science and machine learning, and now everyone's referring it to AI and it's just sounds so much more complex than it is. But people think of it more of analyzing data, crunching numbers. I think in our industry, it's a lot more about productivity improvement, automating different processes, and just doing analysis of text. So if we can think of the opportunities where we can be more productive, and I think structured data is a wonderful, wonderful example of this because we already have the dictionary, if you wish, the schema that we should use, and we just have to extract the different components that need to fill that dictionary. And so this is the perfect automation project and like the perfect data science project. And as an e-commerce brand, if you don't have your structured data sorted, you know, even things like product schema, if that's not okay on your start, and this is the place where you should start, right? Just being included in Google's shopping graph is what you should be doing as the first thing, even though we don't see it yet in the SEO audits and the tools.

Richard Hill [00:30:46]:

But this should be, I think, a top priority for a lot of stores. Yeah. So some quick, fairly quick wins there, I think, on that last piece, particularly, and you touched on your scripts there and scripts. So that leads me on to my next question, really, really nicely. So sort of effective tools for SEO or scripts or, you know, be good to talk a bit more about, you know, so, you know, if you've got thousands of products and you want FAQs on them all, you know, you've obviously got, yeah, I'll take it away.

Lazarina Stoy [00:31:20]:

So, uh, I know a lot of people listening might want me to say, just use GPT and automate it. I'm not going to say that. So in preparation for this podcast, I was thinking about, you know, I have to mention it somehow, right? Because it's all the rage. But in my eyes, or at least in my world, GPT, API and chat GPT have been extremely helpful for coding tasks. So especially when it comes to data science, if you are a beginner and you think that your coding skills might be the reason that is holding you back from trying or testing out scripts, or even just coming up with an idea of how to automate something. I think that ChatGPT is the perfect way for you to get started, because It's very, very good at those code scripting tasks. And most of the things that I've heard as ideas that people have that they can automate something are not as unique. And I don't mean it in a bad way. I mean that if it's not unique, that's good, right? There's a lot of resources out there on, for instance, how to automate schema, how to extract Q and A's, how to scrape data, how to scrape the user reviews and all of that stuff. So, uh, if you, if you have a project like that, I think child TPT would be the perfect way. And now coming back to the FAQ example, I think that starting with a SERP analysis, seeing what actually ranks for those questions. And then, of course, you can use a generative AI to kind of help you get started with the content. But I would always recommend having a product expert, you know, first of all, write something that is unique to the brand. Having an SEO expert, of course, do an analysis and include all of the different features that need to be mentioned or keywords or entities and so forth. Just having that SEO expertise incorporated in that. And yeah, I think It's also important to be able to actually do changes in bulk. There are, you know, this is often a hindrance. Okay, so if we have a thousand pages that we have to insert FAQs, how are we going to do that? Of course, you can build them in Google Sheets or somewhere else where it could be semi-programmatic as well. You could use templates, you could do all sorts of things, but if you can't upload them to the site, then you obviously have a problem. So yeah, I don't know if That's a very long-winded response, but I definitely think there's different approaches to it. Some might be more successful. So let's say then our listeners,

Richard Hill [00:34:09]:

they've done a SERP analysis, they've collated a list of questions and answers that clearly are ranking already or there, you know, and so they're now trying to sort of take over those positions by rewriting those and creating different things with different entities and features, etc. So then they've now got their Google sheet of 50, 000 Q and A's, FAQs. What now, right? Yeah, this is our world we're in. You know, I actually have somebody full-time writing, pretty much full-time, about 30 hours a week, writing FAQs for me, you know, with using a lot of tech, not just, not just by hand, but different tools like also asked and various APIs and different things and, um, but uploading them then to say a Shopify, a Magento, have you got any, to automate that uploading, anything you could give us any ideas on that? Is there any ways to automate uploading sort of?

Lazarina Stoy [00:35:14]:

I think always just, you know, we're coming back to the discussion of working with developers and having the budget for that, because I think the response here would definitely depend on the complexity of the site and kind of, you know, what platform they're using, what kind of rendering they're using and all of that stuff. So I definitely think here it would be best to kind of go with developers for it and having the SEO and developer hand in hand and working on that project. I think the importance here is that, the opportunity identification is becoming easier and easier in our world. And then the challenge is actually doing something that is not only just to compete, but also to do it the best way and do it by the rules that Google wants and all of that stuff. So it's becoming a lot more complex to actually implement and to do things the way that they're supposed to be done as opposed to just doing them. Because I remember a few years ago, FAQs were, I think, the best strategy that you could do. Or at least everyone, it seemed like everyone was getting a ton of traffic doing that. And now, obviously, with the recent Google IO update, we're seeing that Google are kind of taking over that question answering function. So my little projection is that we might actually see that particular snippet getting reversed or at least not as frequently shown, just because if Google is successful generating responses based on the data that we have super greatly marked up with our schema, with FAQs, they can just pull that up and gobble it up and just spit out a better response, right? Because they have that ability now. Then, of course, we're kind of in a losing position and we have to do something different to play by the new rules, right? So, yeah. Great. A little bit of a tangent there.

Richard Hill [00:37:18]:

But just 1 last part on this, then. So let's say tomorrow you come to your your your fire up your laptop. What are the 3 tools that you usually go to? Or what are the 3 things you're working on? You know, is it Google Search Console? Is it, you know, you use SEM Rushmore or are you using, I know you're gonna probably say some of the own, your own things that you've created as well.

Lazarina Stoy [00:37:48]:

I would use OneSpot to just combine everything. My own scripts would be 1 thing. But to keep it less mysterious, I will definitely say if you have an interest in automation or you have ideas how to automate things, you can very easily create scripts nowadays, like I mentioned. So getting a code buddy, even if it's a chat GPT or it is a real person, just start thinking about things that you can very easily do. When it comes to tedious tasks that you don't want to be doing, there is a way to automate them most likely. So with that said, I'll move on to the other 2 slots because You've only limited me to 3. So Google's tools in general, I'm a big fan of. I think they're doing a good job, not only in terms of releasing tools that are accessible to people and beginner friendly, but they're also doing quite a good job with the documentation of these tools, which I appreciate. So Google Search Console, of course, Looker Studio is a platform that I use not only to build dashboards and kind of performance reporting, but also when you have a data set that you have to export different chunks of, sometimes it's a lot easier to go through the route of BigQuery and Looker Studio as opposed to doing manual Python-based work on that big data set. So I appreciate that a lot more. Of course, working in Python, recently having GPT is quite helpful for a lot of the tedious tasks or even just things like automating tickets creation from an audit that you do. Those are things that GPT can help a lot with. And I'm refraining from using GPT for just because I know in like a month or 2 they're going to be like yeah here's the fifth version so I'm trying to keep this updated as much as possible. And yeah in terms of the other kind of tools I like both SEMrush and Ahrefs. I think they have their own strengths and weaknesses. For a small business, I've found SEMrush to be quite useful and a little bit more friendly in terms of the navigation. I think, especially for a small business, having the content assistant is also quite helpful, something that I don't see quite often. And yeah, there's a lot of up and coming tools every day that we see in SEO. So I try to test as much of them as possible. I see some benefits of some of them, even the things that they do better than the big brands. So I would definitely encourage

Richard Hill [00:40:43]:

that testing and experimentation as well. Great, that's a great list, great list. So crystal ball time, we're sat here in 12 months, you know, what's the future of data science within SEO?

Lazarina Stoy [00:40:55]:

I think, I think specifically in SEO it's going to be bigger. So right now, there's a little bit of overwhelm right now, specifically in terms of AI experts, I feel. And I am personally a little bit mentally strained, going on LinkedIn and Twitter because of that. And I feel that that might be making some people not want to give data science a shot or machine learning, or even just experimenting. So I hope that more people start exploring things beyond chat TPT as potential data science projects and start working on actual data science and machine learning and not just, you know, experimenting with GPT. But at the same time, I think it's good that people are starting to wake up a little bit to that niche. I think also, hopefully, the voices that are real within that field, people that are, you know, machine learning engineers working in SEO or SEO data scientists. I hope that those people's voices get amplified a bit more. 2 people that come to mind straight away are Jess Peck and Brittany Mueller, amazing educators, super passionate about keeping people in the right track when it comes to data science and SEO. So I definitely appreciate that, you know, and voices like that being amplified a little bit more. So I'm hoping that happens in the future as well.

Richard Hill [00:42:38]:

Yeah, yeah, there's going to be there's an awful lot of AI experts right now, isn't there?

Lazarina Stoy [00:42:44]:

Lots of noise, I would say. Honestly, I feel a little bit, you know, overwhelmed. And

Richard Hill [00:42:53]:

yeah. Yeah, it doesn't matter where you go. I think I've actually removed Facebook off my phone and I went on my iPad a couple of days ago and I've got Facebook on my iPad and it was like every other thing was just AI adverts for AI courses and this expert is now an AI expert. AI everything, yes. And if we have a couple of minutes, there's something actually very interesting on that topic.

Lazarina Stoy [00:43:15]:

Within the history of AI in general, there have been periods just like what we're seeing today, where there is a lot of rapid development in a certain area of AI, and there is a lot of promise and a lot of hype. And I can tell you straight away, the last time that this happened was with Watson, early 2000s or before 2010, where the promise was that this technology is going to be incorporated into all sorts of different areas. And of course, there's a lot of investments pumped into the field. And at some point, you know, right now, we're not seeing that yet. But in the past, what has happened is that at some point, this deal gets saturated and the research component cannot keep up with the business component. And the business component gets disappointed that the investments are not getting the return that was promised or the technology is not working in the way that it was expected. And then the AI winter comes. So investments stop. There is no more investment for research. And then the research kind of stagnates a little bit until the new hype comes along. So, So far, I think there has been 2 or 3 times in the history of AI. It started in the 60s, not like some people think, 2 years ago or something. Like it's quite old fields, if you wish. So in the history, this has happened a few times, and I'm hoping that this doesn't happen this time. But at the same time, I'm still seeing, you know, as someone that has gone into research on these technologies, I am seeing, you know, the misrepresentation of what these technologies can do. And this is kind of the first sign that people are going to be disappointed at some point. So hopefully this doesn't have this huge enough frivolous effect to actually soap investments into the field. Because I think investments into this field is great. But you know, the hype component of it will not lead to anything. So that's great little tangent in history on AI.

Richard Hill [00:45:32]:

Great finish to an absolutely fantastic episode. I like to finish every episode with a book recommendation. Do you have a book to recommend to our listeners?

Lazarina Stoy [00:45:42]:

Yes, I have a book to recommend that I'm reading right now and it's from a fellow Bulgarian, I think, I'm almost 99% sure. So it's Elements of Data Strategy by Bojan Angelov. So he is a consultant that has worked in many big tech companies and has worked in project management and data and consulting in general. So I find that book to be very good for anyone that is interested in data in any capacity. So I know that is quite a big umbrella of people at this very moment in time. But it has a lot of different components to it, like structuring a data team, optimal way of structuring processes, and projects as well. And overall, how to make sure that what you're doing with your data strategy makes sense and it's not just chaotic, right? So an example of chaotic data strategy in our SEO world or e-commerce would be if you don't have dashboards in place, if you're not analyzing your data, you cannot think of an AI strategy, right? If you're not even understanding your own data, you cannot think of incorporating advanced technologies, regardless of how trendy they are. So I think that thinking of processes and systems is very well represented in that book. So I think

Richard Hill [00:47:07]:

I'll have a look. I'll add that to the list. And we'll tag it up in the on the show notes. So the guys that want to find out more about yourself and reach out to you, what's the best way to do that?

Lazarina Stoy [00:47:16]:

Yeah, it's just lazarinastoy.com or at Twitter is at lazarinastoy.

Richard Hill [00:47:22]:

Fantastic. That's it. Thank you so much for coming on the show. Thank you for having me. It's been a pleasure chatting to you. I think we'll do a, maybe 12 months time. We'll get you back on and we'll have a, we'll see where chat GPT is.

Lazarina Stoy [00:47:38]:

Yeah, I'm sure it'll be a lot different then, but we'll see.

Richard Hill [00:47:42]:

Thanks for coming on the show.

Lazarina Stoy [00:47:44]:

Thank you so much. Bye. Thank you for listening to the eCom@One ecommerce podcast. If you enjoyed today's show, please hit subscribe and don't forget to sign up to our ecommerce newsletter and leave us a review on iTunes. This podcast has been brought to you by our team here Bye bye.