17 Sep Arnout Hellemans | The Three Pillars of Successful SEO
Too many entrepreneurs want all the bells and whistles for their ecommerce site… and the hottest digital marketing strategies to bring in traffic.
But, says Arnout Hellemans, they often don’t get the results they expect… because they forgot to perfect the basic foundation of their site and marketing first.
Arnout outlines what many businesses are doing that prevents them from ranking highly in the search results, and losing potential customers midway through the shopping experience.
The focus of the conversation is SEO – myths and best practices. Tune in to find out…
- A quick way to check if your SEO efforts are ineffective
- How to avoid crawled but not indexed blog content
- Why marketing A/B testing can negatively impact your SEO efforts – and how to avoid it
- Invaluable insights into optimizing your web page ranking and standing apart from the crowd
Mentioned in this episode:
Voiceover: You’re listening to the E-commerce Excellence Podcast with Joris Bryon.
Joris Bryon: Hey, this is Joris of the E-commerce Excellence Podcast, and today I’m really excited to talk to Arnout Hellemans. Arnout is a world-class digital marketing consultant. And he started his digital marketing career many years ago with search engine optimization, search engine advertising, and later on, he added 02 specialties as well. And on top of that, he’s a real Google Analytics ninja.
Well, he’s helped all kinds of companies over those years. And he helped them to grow their online business often with really creative strategies and, and interesting hacks. And at the moment, he’s working for companies like Sitly and Moviq and Royal Bank of Scotland, and he’s a popular speaker on digital marketing conferences. And to be honest, I’ve known Arnout for quite a few years now. And he’s one of the smartest digital marketers I’ve personally met. So I’m sure this is going to be a super interesting episode. Arnout welcome to the podcast, really happy to have you here!
Arnout Hellemans: Thanks. Thanks for having me Joris. Very, very pleased you asked me and looking forward to it.
From SEO and SCA to CRO
Joris: Yeah. Cool. Maybe just to start, I’d love for you to tell everybody a bit about your background. Where’d you come from and your career? And how did you get to this point?
Arnout: Um, well, basically, my background lies in telecoms, this is where I learned regular expressions and databases and that kind of stuff. And then I was asked on a big startup, we wasted millions, I wouldn’t say wasted but we spent millions. And it was great learning money. This was back in 2006-2007. But due to the financial crisis, it is the plug was food so but that company, which was called Novick got me interested in SEO and SCA, because agencies were doing it and I started reading about SEO on the SEO Malls Blog, which is now just Malls, but it used to be SEO Malls. And I started just implementing stuff.
And it worked. Yeah, that’s kind of how I rolled into it. And then I got a hernia so I was unfit to work for about two years, which was ouch. But it gave me plenty of time to read. So I read an awful lot of blogs, an awful lot of things. And I started implementing that. And when when I basically grew one of my clients, I at that moment started working on basically 10 folded in a year. I was like, oh, I might be onto something. And that’s when I started my own consulting business. So that was back in 2009. I think so about 10 years ago now. And I’ve been working for big and small companies alike. What I like about big companies is that the impact you can have is can be enormous. What I like about small companies is that it’s way easier to get stuff done. So I like the combination, as you said on the at RBS, Royal Bank of Scotland. Previously, I did lease plans.
Those are like big financial and insurance in the Netherlands. I also worked for big web shops, with loads of products. I just did a training this morning. A web shop in, in basically motorcycle parts in six countries. enormous amount of products, like they got 100,000 SKUs per web shop. And they’ve got nine of them. So yeah, interesting. I like those kind of challenges. And it basically brought me here. If you ask me how I made the move from SEO and SEM into CRO, was because SEO and SCA were becoming too simple.
It was just about links, links, links. And the other one was just if you set up the campaign correctly, you’re pretty much good to go. So I got a bit bored. And so I got more into analytics and more into CRO. This was where basically was the conference we first met in Estonia. That’s where I basically got into CRO. And now I basically tell people I optimize. I just optimize websites stuff, everything. And I love doing it.
Joris: Yeah, cool. We still have a lot of experience in SEO. And I love talking to you about SEO because I always learned a bunch. And I know a lot of people and most of our listeners are ecommerce ecommerce owners they struggle with, with SEO for ecommerce, what would you say is the best way to do SEO for the e-commerce site?
Arnout: Ooh, well there’s no such way as doing SEO for an e-commerce site. SEO is a mix. So it has to do with usability. So with SEO there’s three pillars. One is the technical part, which in a lot of cases has not properly been taken care of. So this means canonicalization, which means duplicate content, shallow content, and crawl deficiencies, internal 301s like all of that, which is the technical part, then the second part is your on page content has to do with product, this descriptions, images, all those kinds of things. So content wise, and then there’s the third one, which is popularity, aka links, branding, links, that kind of stuff.
And so the way I see it, those are three different pillars. And they all need to work together to get the right mix. So for some companies, what prevents them from ranking is the fact that even though they have, say 10,000 products, they’ve got 200,000 pages in the index, because of crawl problems. So who doesn’t know which one to rank. Whereas other web shops might have all of that covered, but they don’t they lack the authority on those topics, slash links, they don’t have links. So they that prevents them from ranking.
And the third one doesn’t have their page titles filled. And then every page title is just the brand name. And they don’t rank either. So I always tell people, if you want to talk about SEO, there’s two words that you can use in the start of any answer on SEO. And it’s, it depends. So I can’t say. And I can just say that what I see happening a lot is the technical difficulties, which means like, just do a simple search for all the listeners out there. You do IN URL, which is IN and then URL, semicolon, and then your site name. So basically your domain name and look at how many results they actually return.
And then look at like, how many pages does my site actually have? If it’s around the same, you’re perfectly fine. If it’s like three, four or five times as much, you have a huge problem there. So can’t say what is the bigger thing? But those are pointers to go to. As I said, you need to get a proper SEO person to do an audit, if you suspect it’s not working as planned. And then that person should look at all these three big aspects.
Joris: Yeah, because, yeah, absolutely. And they all work together. And if you have the technical stuff covered, then you can look at the on page stuff and the authority. But it really depends. If you have that covered. Or if you if you’re on the beach sucks or at the authority sucks. So to every every case is different.
Arnout: Yeah, exactly. Exactly.
Should You Create Blogs For the Singular Purpose of SEO?
Joris: Yeah. One way I know that a lot of e-commerce companies try to rank in search results is by creating a blog. And they sometimes only created for SEO purposes. And what’s your take on that? Is that a good idea?
Arnout: Um, oh, here we go. It depends. No, but to be fair, people that do blog.doc as a subdomain don’t even bother. Because, well, you’re basically creating a new webpage. A new domain, a sub domain is being in Google is being treated as a separate domain, not part of the same. And Google is trying to solve this. And they say solved it. But I’ve seen hugely different different things. So yeah, it’s that it, I think that is my main thing. However, if a blog is set up in a correct way, so as a subdomain, not sub domain, some folder through reverse proxy, there’s loads of ways of doing it.
And then the blog writes about, say, topical relevance. So things like the 2019 living room trends, if you do proper research into that, and you’re not trying to sell but basically become the authority in that, then or you do proper research and say, do research in New York and California and, or in New York, and LA, and see the differences in trends there. And basically use that on the blog, and you use that in outreach, NPR, and you do it on a sub folder, it’s awesome.
Because you, you might be able to pick up links from big news outlets. And then that helps all the products you’re talking about, in that blog post of the newest trends to rank. If you’re selling, say furniture or, basically things from the living room inside and stuff. So it depends. Don’t purely do it for SEO perspectives, do it to become the authority. And a big thing here is don’t just chunk out content and write something you’re proud of. I think that’s the way forward.
Joris: Yeah, so better have one blog post every two weeks, then one every day, but one really good one every two weeks. That is really relevant. Yeah.
Arnout: Definitely. Use the simple way of doing an interview. Say you sell to, you do high end gardening, for instance, gardening plants, or whatever. And you know that there’s a few people out there, like celebrities that have used it or whatever. And you just reach out to them and say, can I do an interview with how you came to use our plants in your garden or whatever. There’s so many ways of doing this. And building something that you can use to rank but don’t, don’t do another one on.
Another blatant product review that they can read anywhere else. If you do a proper one like product product comparison. At That’s awesome. Like if you have to say it, you sell five brands of blenders, or kitchen appliances, you use them against each other, you give all the proper pros and cons. You create video content around it and just create one page comparative the three most compared products. Boom, there you go. That will rank.
Joris: Yeah, absolutely.
Arnout: But I think the biggest rule is, do you want to give it to a friend to read? If you then say no, then you shouldn’t publish it.
Joris: I love that rule, because there’s just so much rubbish out there. Especially if you’re just.
Joris: Yeah. What are like, typical SEO mistakes that you see your clients make often.
Arnout: And I think one of the biggest ones is canonicalization. So
Joris: Can you explain that a bit?
Arnout: Okay. Yeah. So basically, what happens is, the simplest version is you, you go to a shoe shop. So let’s say shoes.com. I don’t know the brand, but whatever, shoes.com/sneakers, and then I can sort stuff there. So I can basically say, let’s do all the blue and the brown ones. So then it’s question mark, filter color, brown and blue. That generates a new URL, which is not slash sneakers, it’s it’s slashed acres, question mark, sort or filter, blue minus brown. And that page, and so forth. If I didn’t say sort by price, it says question more.
So you basically create an enormous amount of pages by filtering, and if you do that, you get a create an enormous amount of pages that Google needs to crawl. Whereas in all essence, what you only want is to have the sneaker page, or you could have a brown sneaker page, but in this case, just sneaker should then you should put canonical in that sorted page that says, If you encountered this page, you actually want to rank the slash sneakers page. That’s what a canonical does. A canonical basically says instead of ranking this page where the canonicals are, I’ll point you to the one you should put in the end. Does that make sense?
Joris: Yeah. And it’s a pretty straightforward fix, I guess. Because instead of having hundreds of variations of one particular page, hotel and Google, like a, ignore all these other ones just go to this one. That’s the one. That’s the that’s the main one. That’s the important one. I know there’s sometimes like, a debate about SEO versus zero. People are sometimes afraid that some tests that they running may affect their rankings, for instance, what’s your take on that?
Arnout: But if you talk about testing, yes, I can see where that comes from. I don’t think any of the tool providers will actually admit that there will ever be a problem. But I noticed something with one of my clients where so you have two pages, say, the normal landing page and a version. And then you use an AB testing tool to show the version. And Googlebot has been updated to run the latest version of Chrome. So and this tool, basically, in the documentation, they say, we don’t make any exceptions, whether it’s Google bot, or a user.
We’re just gonna, so we don’t exclude Google, but from serving different content, because that will be seen as cloaking and cloaking, is when you show a search engine, something else, then you show the user. However, what happened with us was our, say, B page, so not the original but the B version, we put in no index on there, because you don’t want to B version to be indexed. However, what happened was, Googlebot requested the A version, just normal crawl, but was served the source code of the B page, which included the no index. So what happened is it was thrown out of the index.
Joris: Okay. So, is there a way to work around that?
Arnout: Yes. And if you’re going to, so what I always would do is exclude Googlebot and Bingbot and all the other search engine bots from these tests. That’s the that’s the best one. If you can’t, or don’t want to do that, what I should, what you should always do is have on the be version on a canonical pointing to the eighth version, so that even if the B version is indexed, you will you will still rank the A version.
Joris: Yeah, certainly we do it like that, for instance, whenever we work with a spit URL test, but we added canonical, but what happened is, it’s probably not the same, I guess, when you have an AB testing to and you use the editor, or the AB testing to them to manipulate some elements on the page, like hiding some elements, removing elements, putting them on replace, missing, that’s not that much of an SEO we issue then is it?
Arnout: Um, well, that depends on how Google Book renders the page. So if you don’t exclude them from that, and you remove, say, the page shot, or you change the page title, and Googlebot gets, see got shown the rendered new version, then it might. However, if you prevent Googlebot from being in the test group, it shouldn’t matter.
The Importance of Link Building
Joris: Right. Yeah. That’s, that’s actually the easiest to fix, Right? And one other thing, I was wondering, and you already touched upon that briefly, links, I mean, and a few years ago, everyone was saying link building is dead links don’t matter anymore. And I never really believed that was with like, people who wanted to believe they wanted to believe it, because then they could give up on their, like, lazy mind numbing link building activities they were doing. But what’s your take on that? How important are links nowadays for your SEO?
Arnout: They’re still very important. Because as I said, in the beginning, it’s three pillars. So if two of those, if one of those pillars is not working, you can throw as many links at it as you want, but they won’t rank, right. So again, it depends, however, it also depends on the country, like in Norway, in Sweden, and Finland, in the Netherlands, even, you can try and get like high authoritative links. But getting too high authoritative links are getting 10 relatively shitty links to 10 will still rank you.
Okay? Better than the two really good ones. And it’s a different story in bigger countries, like the UK or the US. In the US, you basically want to have these relative authoritative links. Google did quite a few crackdowns on link networks and directories and all kinds, all that kind of stuff. But, but but getting local press coverage or international press coverage, or national press coverage, like those are really important ones. Yeah, I think it’s, it’s still very important. But again, those two words matter the most, it depends. They’re definitely not worthless.
Joris: When it comes to digital marketing in general, I know you have like a really common sense and no nonsense, no bullshit approach. You often talk about the basics, like people don’t even have, yeah, have covered the basics, right? Like, like page speed. Can you expand a little bit on that? What in your opinion, or basics and like, but like page speed?
Joris: Yeah sure. Absolutely, yeah.
Arnout: So I think a very cool one here is that Google builds a report on older Chrome users. And it’s called Crux. So you can google it on Crux, C-R-U-X report. And what Google has done is for older people running Chrome, they ping back home their load times and times to interactive. So for any website that is big enough to be in the data set, you can create a benchmark report. So what I usually do is I take my client and then their two biggest competitors, and I’ll check them out on how they perform compared to the other two.
Joris: Yeah. I see. Yeah.
Arnout: And I think, when people started looking at that, they can see the enormous potential. I mean, I’ve worked with the developer who built an e-commerce platform, with 500 kilobytes of code, full blown check out everything. And that can be done. However, most e-commerce sites are like two, three, max, which is way too big. You don’t need to have that. And, and I often see, because these people then come in, say, I want an AB test.
And I tell them, well, if you just fix the basics, you’ll probably have a bigger uplift than trying to see which call to action works better. Because the impact is way bigger than just on site conversion. It’s also about how many pages you have in the index, your Google rankings, your basically all of that will be influenced your checkout, if your checkout is way faster, way easier. People like it makes huge sense. And if you want to benchmark yourself, take I think an Amazon product page and see how that loads versus yours. Yeah, that’s, that’s what you should be aiming at.
Properly Gauging Page Load Speed
Joris: I think like one of the tools,we sometimes use as well as its webpagetest.org, just for one thing, and it is that you can just generate a video in the seconds that a page loads. So you can actually because if you look at just the data, it’s sometimes it’s abstract, it’s vague, and people, okay, it loads in 11 seconds, which of course is awful.
But when it then they see a video of that page loading in 11 seconds, and you have to wait for actually 11 seconds for that page to load completely, then all of a sudden, it becomes real, and they know they have to work on it. Because they would never accept that when they go to Cyprus for some reason. They expect to see that number in there. And the reason I call.
Arnout: Yeah, exactly. And you can do the same if you sign up for GTmetrix, it’s free, you just log in, and then you can tick video and they will create a video for you. And then Lighthouse, you can also see a film strip so you can see the elements that it takes and why it’s so especially if you want to talk to developers, because developers usually basically go to come to me and saying, well, it’s not this slow on my machine.
So yeah, there’s multiple tools to do it. Um, so those are some of the basics. The other one is page titles. For your most search for products, internal sites search, and left and not optimize? Like, there’s so much and but again, it depends. But these are the basics you want to solve as soon as possible.
Joris: One of the things I like about the way you look at digital marketing is that you understand to the unit, you’re really intelligent, you’re really good at Google Analytics. But yet, you often make a case to step away from the data and really start thinking about the people. Can you expand a little bit on that?
Arnout: Yeah, I think people don’t really understand data, I think you’ve been asked these questions a lot is what is a good conversion rate?
Joris: Yeah, all the time.
Arnout: So what’s your answer on that?
Joris: Depends one that’s better than last week. To always try to be better.
Arnout: Exactly. And so the case I make it with that is, I asked them, would you be happy? If you have a shop and 10 people walk in? And one buys? Would you be okay with that? Or imagine 100 people coming into a shop, and only five people bought. Would you be okay with that? So I tried to humanize the data. And because that makes people think about it. And this is not, of course, it’s if you walk into a shop, you feel obliged almost to buy something, you feel a bit embarrassed if you just walk out again, most people do.
Whereas with a website, you can just click it away, and it’s done. So expect it to always be lower. But with these kind of things is if you’re if you’re in a brick and mortar store, you can see people get frustrated, they can’t find it. So you walk up to them say can I help you? And they go like, yeah, I’m looking for this, and then you just point them in the right direction. On the website, we send people to 404 pages.
We frustrate them in the checkout process by not telling them how many characters the password should be, or the validation rules are or whatever. So unless we start looking at it as real people will just look at it ah well, yeah, it only happens like in one out of 10. Well, in a real in a brick and mortar store, you wouldn’t be okay with one out of 10 people being frustrated in your store, would you?
So I want people to look at data, use data, but really understand what they are the real people on the other side. And if you don’t understand that, if you don’t think like that it’s very, very hard to create that mindset in your company. Everybody just should, its customer first it’s how Amazon grew humongously. One of the reasons was they put the consumer first. And they didn’t they weren’t looking at a data and numbers.
They use that to build a better experience. And that’s the way it should be. So it’s always try to use an analogy and to explain it in a real proper store. So what’s happening, people basically don’t finish the checkout. It’s like them being at the basically at the counter, almost willing to pay and then turn around it. Shouldn’t the chat ever happen? That’s right, there you go.
So I think those are some real life, real life examples on how to look at that. The other thing is that I see loads of people reporting on ranking rankings. So search engine rankings, SEO rankings. Google just did a big update, the medic update. So some, some companies I know were hit. I reached out to talk to them.
And so one of the questions I asked him was, so this was a company selling prescription and non prescription medicine online. And I asked him, like, did your revenue take a hit? They literally lost 60% of traffic pretty much overnight. And I asked them need your revenue, take a hit. And they said no, the revenue went up with 20%. Okay, so there wasn’t a problem, however, that he was reporting on traffic, traffic took a huge hit.
And what basically happened was, Google was showing them for informational queries. And Google probably got better at understanding when they were relevant and when they weren’t relevant. Yeah? And by having that, it could well be that they didn’t really have a problem, they got 20% increase in revenue, with less traffic conversion rates went through the roof.
Joris: Yeah because all that irrelevant, non-converting traffic that came to the site is not coming anymore. So that doesn’t matter.
Arnout: Exactly, exactly. And since conversion, right, that’s the other part since conversion rate is based on traffic and conversions. And if you if you send a ton of shit traffic to it, whether it be paid or whatever, your conversion rate will go down. It’s all session based.
Joris: Absolutely. And if the intent of the traffic is is is not really relevant, then you can have all the traffic and you’ll see the conversion rate going down, but doesn’t really help you. So I mean, well, it doesn’t matter if you have all that traffic if you’re not selling to them anyway, because you’re not relevant to them. I mean, we don’t do zero and zero, I hate that term, because people are so focused on conversion rate conversion rate, and I hate it.
Arnout: Let’s create a new name, let’s create a new name. We always call it continuous revenue optimization. Yeah, that’s way better way better way better.
Traffic Versus Conversions Versus Revenue
Joris: Because it describes what we do at the end of the day, we just want to look at your revenue, if your revenue goes up, who cares about completion rate? Of course, it has to be profitable, you can increase your revenue and lose out on profit. But I mean, who cares if your conversion rate goes from 2% to 5%. But you’re actually not making more money or if it goes from 2% to 1%. At the end of the day, you’re running a business, and it’s about the money you’ll make it’s not about the conversion rate. But anyway, that’s me ranting here.
Arnout: No, no, but. Well, I fully agree. I couldn’t agree more. I an even better one would be profit. So it’s basically profit optimization.
Joris: Zero anyway, now and we’re stuck with it. It’s not going to go away. I thought like, okay, let’s.
Arnout: But I fully agree, I think zero, as in conversion rate optimization is not the right. It’s basically what SEO was, in the beginning, people were talking about I want to rank rank rank. And I’m like, well, but rank for what? You’d rather have to write traffic with the right intent on your page? Because you’ll sell more, and you’ll spend less. Yeah, you see it now. Yeah. You see, when you Google a half-term, you know what ranks? Wikipedia with the explanation of what it is. It’s for a reason, it’s because people expect that company expect to buy anything at that point.
Joris: True. It’s all about the intent. But anyway, one of the things I definitely want to talk to you about, I mean, we’ve known each other for many years now. And one of the things that I admire about you is that you have that unique combination of the technical know how, also great strategic vision and a very creative mind, which is a very unique combination.
And every time we talk, I learned something new, something unconventional, sometimes even almost brilliant creative hacks or techniques. And, and those always stuck. So okay, frustration metric, you have a frustration metric that you’re working with in Google Analytics for city. Can you tell us what it is and how it works?
Arnout: Well basically it was proof of concept I tried doing, because I truly believe that nobody should have a frustrating experience anywhere in anywhere, whether it’s offline or online. But an offline we can kind of see it, when a door opens the wrong way. When I don’t know just we bump into them every now and then. Some are easier to spot than other. Well on a website. So let’s let’s do this little thing. What is the most frustrating experience you ever had to go through on a website?
Joris: Many, especially being a conversion optimization specialist, I see a lot of things go wrong. But I think like, checkouts usually are very frustrating, especially if it’s not working. And I don’t get any feedback, why it’s not working. It’s not helping me. And then I try again, and again and again. And I get stuck because there’s solution, they don’t tell me what I should do when I’m making a mistake.
Arnout: Exactly. Exactly I think it is for me as well. So basically, what I started doing, I started sending events whenever this happened, you know, whenever, whenever something fails, so here’s a little trick you can use. So you use event tracking you do in land validation. So when somebody hops to the next field, you validate it and you either get the green tick box or you say sorry, we, this field can be empty, or say something like, this email address has a space in which it can’t have, like those kinds of things. And make them as specific as can be. And then what you do you send an event saying category, check out the field name and the error.
Okay. Yeah, and you send us an email to Google Analytics. And, but if you have session playback tool, like Hotjar, you can send the same fields to Hotjar as well. And when you do that, what happens then is you’ll be able to see the frustrated user that encountered an error. So this is one way of looking at it. So the next one, so I defined any click for any field where there’s a validation error is a frustration. Right? And then the next one is clicks on non-clickable elements. If people click on an element, they that is unclickable. And then that’s frustration, because you expect something to be there. Otherwise, you don’t click.
Joris: Make sense. Yeah.
Joris: Cool. And you do that with a calculated metric.
Arnout: Exactly. I do that with the calculated metric. And but when you start thinking about this, this is way ahead. The reason I tried looking at getting into this is I want to see if I can prioritize which websites which pages or group of pages creates the most friction. Because those are the low hanging ones going through we should fix first.
Joris: Cool. That’s a really, really good idea. And I know you have plenty of others, but we’re going to run out of time. And, yeah, I know. I mean, we can go on for hours and hours, which we’ve done many times, which we’ve done in the past. Anyway, we’re running out of time. And just before we go, I want to make sure that people know how they can find you and learn more about you. What’s the best place for people to connect with you?
Arnout: Well, if they Google me, I should have my own knowledge panel. Which is pretty cool. So they’ll be able to find me. Yeah, Google thinks I’m a Dutch consultant. I have it in front and around 43 years old. And I’m that I’m from the Netherlands. So if you google Arnout Hellemans, there’ll be the knowledge panel with some videos that I that I’ve done. And then there’s my LinkedIn. I’m more than happy to connect to anybody on LinkedIn. As long as they mentioned, you and the podcast. And they’ll find my website. They’ll find my website, my contact, my Twitter, and like all of that, so yeah. I think that’s the best way to do it.
Joris: Yeah, what a better way to find an SEO expert by Google it, of course. Thank you so much for being here Arnout. This was awesome.