English Español Deutsch Français Italiano Português (Brasil) Русский 中文 日本語

Ross & Craig Site Audit (e-Commerce sites II)




Ross: Hello everyone and welcome to another SEMrush webinar. We're joined today in complete lockdown and isolation, and I'm here with, I don't know if you can recognize him, he is COVID-19 protected Mr. Craig, how are you doing Craig? 

Craig: All good, all good. 

Ross: I take it you're working from home this week. 

Craig: Last Friday we stopped, just took the decision, I don't want to catch this stuff, so from last Friday decided to make everyone work from home, which we do sometimes anyway if it's a nice day or good weather or somebody's away on holiday or whatever. 

Ross: Has your business been affected by it at all? I know you don't really do client stuff but you do training, but I imagine training sessions are canceled. 

Craig: I mean obviously training sessions, for now, are canceled because we're not going in to deliver them but it's hard to say because it's still early days yet, in terms of people. I've spoken to other guys and they're losing clients left, right, and center. Pay per click people stripping back their budgets. I've also seen various other people having to lay staff off already and stuff. 

I think in terms of business it will hamper things and slow things down but hopefully we're in a good position where we don't have to panic too much.

Ross: Hello everyone in the chat, it's lovely to see you all. This is eCommerce-based. Shall we talk about some of the websites that we've been given to audit today? 

Craig: Yes. Do you want to go first? 

1st Website: Over-Optimization, Improving E-A-T, & Better Internal Linking

Ross: Yeah, let me share the screen and we can go over the first website. Let's have a look here at Vitality Medical. I'm going to go through some basic things that I've seen. Craig, do you want to maybe talk through some of the top-level things that you've seen on this site and then I can jump into some of the tech that I've noticed? 

Craig: Yeah, so before you try to sabotage this website in any way or find fault with it, this website ranks really well and has a lot of traffic. However, one of the things that I think the website lacks is... well, we both spoke about it prior to coming on here...it's probably over-optimized as such. I think we both agree that there's probably a guy working on this website who's full-time who just optimizes and optimizes and optimizes everything. 

And that's not always a good thing, it's kind of borderline spammy. Although your rankings are good and it's working, I think it's really important to future-proof your website and potentially not make it so obvious.

I know that's probably a kind of contradiction, you either optimize or you don't. But you also said it as well, Ross, it's borderline spam with a lot of this stuff on here. I think that is something that eventually, whether it's in six months or whether it's in two or three years, it's potentially going to impact that website. Would you agree with that? 

Ross: I would. If they're watching this, I wouldn't necessarily call it spam but I would call it over-optimized. They are getting good traffic growth, as we can see from the SEMrush graphs, but it's really, something major's happening. I imagine this was some kind of migration that didn't go well or round about here there was no major algorithm update in February 2019, however there was one in September 2019, and it was a core update.

It looks like they've been negatively affected by a core update and they've made some changes and they've come back up. I would need to get some context on what did they do in order to fix the issues with the core update. 

Now, the fact that we're going on and seeing things that are clearly over-optimized. A good example would be this, so this is the deodorizer page and when we have a look here we can see that the titles just go on and on and on and on. 

The actual titles themselves, let me just hover over so we can have a little look. It's “deodorizer, skin deodorizer, deodorizer sprays, deodorizer spray, singular, skin deodorant spray, cleansing deodorizer.” Jesus. That's too much. 

You can actually see that Google's actually truncating that so it doesn't matter that you've got all those keywords stuffed in there. I would class that as over-optimization. Just pull it back to a single keyword or topic that you want to rank for. 

In terms of the descriptions as well, things like ”deodorizers on sale, buy skin deodorizer.” It's the same thing. That needs to seriously get updated. 

Craig: Yeah, it's probably going back to 2007 SEO, ramming the meta description full of keywords, no need to do it. 

Ross: Absolutely. That's one of the big things we noticed. In terms of over-optimization, that's one thing. They don't have huge control over their knowledge graph, so these guys are in medical and medical is subject to something called expertise, authority, and trust (E-A-T). Or at least they're a lot more exposed to it. 

Now, when I look at their contact information, it's really important that these people, as entities, are known to Google. And Google trusts who they are. When I have a look at their name, address, phone number information, typically for eCommerce we wouldn't actually care about things like NAP, because it's not localized. 

But in this instance, because it's medical I'm actually quite concerned. I'd just be very aware that Google does not understand who you are as an entity, and that needs to be fixed. That should be something that I'd look at straight away. 

Craig: I would also be inclined to say that because the website's ranking really well, there probably isn't going to be too much in the way of something glaringly obvious that they can fix as such. I think it's a combination of a whole bunch of stupid little things is what they need to do to then take that website up to the next level. 

But even if you look at the footer, they've still got the 2019 copyright on there. Silly little things like that don't mean anything SEO-wise but in terms of having a look at it and trying to be a trustworthy website that's selling medical supplies, you probably want to keep all those kind of things on point, as you said, because E-A-T is a big thing and that niche. 

Ross: When you look at that page and look at the structured data markup, there is person markup but one of the things that I would highly recommend is using sameAs schema. Interestingly enough, they've done it on their homepage so it looks like they knew they had an E-A-T problem and they've tried to fix it. 

You can see here, URL, Vitality Medical, sameAs all of these things, which is a really great start. I would be applying that to these authors because at the moment this is just some random person that's got nothing to do with medical. 

Revise the internal linking structure. This is another one for me. When I'm looking at the elastic tape page; this is a top-level page inside of their navigation and you're like okay, well if that's important then we should find all the other pages related to it and link back up to this page. 

What I've done here is I've just taken the domain and then the keyword they're trying to rank for, and I've pulled out all of these pages. These are the most relevant to and the most powerful pages related to elastic tape. But when you click on them, although there is plenty of opportunity to internally link back to those core pages, they're not doing it. 

There is good internal linking in a bunch of places, but it doesn't feel like there's a strategy behind it that makes sense. If I was to ask this particular SEO what was your thinking behind linking to these particular URLs...I think they'll probably end up saying, "I'd seen the keywords, I know internal linking's a good thing, so I linked to it." 

The downside of that is it's going to create a really weird mesh of things. When we look at this tape thing, for example, let me click into paper tape and then we click into a product itself, what I'm seeing here in this URL structure is absolutely flat. Which, from my point of view, that needs to change, and for good reason. I would want it to have it /tape/3mmicropore. 

Also, the way in which the internal links work, I'd want them to be all internal linking within the subfolders. Very, very hard for Google to understand what's going on if it's not properly structured.

I think that's probably everything from my point of view. Anything else from you in this one, Craig? 

Craig: I think you've pretty much covered a lot of the stuff that I would have said on there as well. There's not too much else wrong with it as such without digging deeper into Search Console and checking what's been indexed and what's not, things like that. 

But from the outside looking in, reasonably good job there but always room for improvement on those kinds of small things, and those small things can also make a big difference. 

I know people potentially look at that and suggest that internal links, how much an impact do they actually have? If you look at a guy like Andy Drinkwater...he's not done so much recently in terms of posting on his Twitter, but the difference of him just implementing a good internal linking structure can certainly move the needle. 

I'm a great believer in internal linking and that, just implementing that will make a difference to your website if you do it properly and with some form of structure and powering up the kind of top pages on your website. And by all means, I think Andy's also, if you Google “Andy Drinkwater internal linking structure”, I think he's got a fairly comprehensive guide on what makes a good internal linking structure. 

When is a Flat Site Structure Bad for SEO?

Ross: Got some questions coming in. "Why does flat site architecture have a negative SEO impact?" Do you want to give us your opinion on flat structures versus folder structures? 

Craig: I think again, when it comes to business and online and everything else, you want to future-proof your website as well and make things easier for Google to understand what goes where. 

Having the correct structure to a website rather than just yourdomain.com/thebestbiscuits or whatever it's going to be. Doing that, and as you scale, it's just going to be garbage. It's not under a biscuits section where it should be yourdomain.com/biscuits and then chocolate biscuits, then shortbread biscuits, or whatever the hell.

But there are counter-arguments. I've got websites that have a flat structure and they still rank well but I think long-term as that website grows I'm going to run into problems and eventually I'm going to have to bite the bullet and structure it properly.

I think from the get-go where you can structure it in a nice simple and easy format, and I think whatever weight Google does give to it, there's also arguments, Ross, who say, "Well, I've got these websites and I changed it and we had a positive impact." Lots of people talk about the positive impact. How much of that impact does it have is anyone's guess, though. 

Ross: It probably changes with the size of the website. You made a really good point, if you're scaling up and adding more and more content pages, more and more product pages, it's likely that it's going to have more of an impact the larger your website. 

If you've got a million products hanging off of one root domain, it's not ideal, and when it comes to clustering all of the different topics together, it's just a clean way to do it. Whether or not it's got a huge SEO benefit, well that's up for debate. Whether it has a good UX benefit for customers, yeah, it almost definitely does.

2nd Website: Improving Commercial-Intent Traffic and Heavy-Lifting Optimizations 

Ross: All right, what we got next? Do you want to put one up on screen and go through it? 

Craig: Let's go for it. Share screen. On to Trifecta Nutrition. This is actually a really nice looking website, I really like it. These guys, as I was saying to you before Ross, this is the one with special delivery of foods for guys like you who are too lazy to cook for themselves, so used to be the mummy's boy or whatever. 

Ross: All of that applies to me. 

Craig: Nice, clean looking website. An eCommerce website that can deliver food packages to you and obviously these are not just any food, they are good quality, organic foods that are going to help you look a lot better. 

The website itself, again, it's one of those websites, Ross, that if I go to SEMrush and put this in, it has an outrageous amount of traffic, so it is already a successful website. You can see there, over the past couple of years they have grown arms and legs. 

They had a slight dip, probably the same as that last website, round about the September time. And then from there it's been good steady growth. I think by all accounts it looks like they've got a good structure in place, in terms of doing some paid search as well, so it looks like a good, solid, legit business, plenty of backlinks and everything else. 

It's going to be quite hard to find fault with a website like this, but I'm sure we can come up with something. I’ve done a quick audit, a score of 67%, which is never good. I'm just going to quickly double-check there wasn't anything too glaring there. 

Nothing outrageous there, but what is quite important is the internal linking, and we spoke a bit about that previously on the last website. This is a website that doesn't have a great internal linking structure, 318 pages only have one internal link going to each, and obviously you can see here from the score that you're getting with internal linking, it's not great. 

With any website, as Ross was saying about the other website, you want to take the opportunity to link back to the most relevant and the kind of power pages on your website, but obviously they've got to be relevant as well.

It's very hard to pick holes in a website like this, Ross. I'd be curious to see if you think there's any glaring errors or any of that.

Ross: One of the most obvious things is when we first look at the site visibility graphs it's very easy to be like “oh this is amazing, these guys are doing really well.” But when we actually dig into what that traffic is and where it's going, not a lot of commercial intent with a lot of this traffic. 

There's huge volumes of traffic for things like how to get shredded, macro calculators, things like that. But none of that is a high converting. Things like keto meal plans or weight loss meal plans...they're not doing as well for a lot of those commercial terms. They're killing it on a bunch of the informational stuff. 

I imagine the strategy there is to get someone to convert into an email list somewhere and then slowly give them the benefits of using the service and then upsell them, which is great. And also if they're doing retargeting...the free, organic traffic will make the retargeting much cheaper and much more aligned with the actual audiences themselves. 

One of the things that I would say, we're now into the “nice to haves”, heavy lifting optimizations. When we run Lighthouse reports for them, (so for anyone who doesn't know what Lighthouse is, Lighthouse is essentially a speed test), first contentful paint is essentially how long does it take for the website to look like this here. 

You've got really high numbers like this, to be honest with you, five seconds, seven seconds isn't that high...lots of rich images and things like that, so fair enough. What I'd be recommending is get a separate image hosting service, which handles all the different images sizes. I would get that on a completely different service and have that piped in because it looks like most of the savings can come from real basic things like image optimization. 

The fact that you've got an 18-second load time for all the images on a core, commercial page. That's something I'd seriously consider is starting to optimize those images, it's really basic and easy to do. 

You can start seeing here that we're looking at main thread work, so what essentially we're looking at is all of the different files that need to actually load to make the page work, and there's absolutely tons of them. I would be looking at that. 

The project here isn't your standard SEO stuff where we're going in and fixing what's broken and things like that. We're looking at what their next big development cycle is, and we're starting to think about the physical infrastructure that they're using on the website. 

Because not only speeding it up will help with SEO, it's also going to really help with conversions as well. Having 18 seconds for all of the images to load just feels a bit lazy and you're shooting yourself in the foot there, it's a great site, it's really well put together. I had a look at their schema, it's phenomenal, it's all absolutely perfect. 

The other thing I'd be considering is looking at how we're doing keyword targeting here and what words are physically on the page. You use something called Surfer SEO, Craig, is that right? 

Craig: Yeah. 

Ross: Surfer SEO essentially is looking at the top 10...essentially all it's doing is looking at the top 10 rankings. Pick a meal plan, so we're jumping into it. What we'd want to do is we'd want to compare, of all the words and topics that are actually on the top 10, I would want to compare that to what's on this particular website. 

I'd really start going through it with a fine-toothed comb because you're now at a place where you're either going to do big, structural changes, like speeding the backend of the site up, or you're going to be doing very small tweaks to copy and we're going to start testing things on that kind of a level, but overall I would say that these guys have done a standup job. 

Craig: That's probably the best website out of all the webinars I've done and site audits, that is probably one that's in the best overall shape to be honest. But as you say, life doesn't stop there. 

One thing I do want to ask you, Ross, you were talking about the images on their website and all that kind of stuff. Is there a tool or a service that you could recommend that they could use? What would you be using if that was your site? 

Image Optimization Tools

Ross: The Type A Media website uses something called Sirv, which is essentially just, it's an image server and it's a dedicated image server. There's a free one called Thumbor, which does it as well, so we're wanting to host that off domain and then sideload them into the website in an ideal world. 

Thumbor does automatic scaling based on what the browser window is, so if you come on a mobile it will know that you're on a mobile and give you the smaller mobile images, whereas what happens a lot of the time is on a website, it will typically have one single image file and it will let the browser scale it. What's happening there is you're slowing down the process because you're making your user or your customer do more work.

Whereas if you put a call in to say that this is a mobile browser, it will physically pull the correct image sizes in. You're not on a mobile phone loading a massive jpeg, you're loading one that's much smaller. Those were the two services I would use. 

A lot of guys just stick them on AWS, so Amazon AWS, you just host them there and load them in like that. These are bigger infrastructure based questions so this is something where you need to get developers involved and you need to really look at the company's long-term plans to get it fixed. 

3rd Website: Slow Loading Times, Old WordPress Themes, and Basic Errors

Craig: You've got another website, Ross, so we've probably got time for one more and then we can go on to questions and answers.

Ross: The next website is something called Graystone Properties, so I wanted something that felt a bit more lead gen. This is meant to be an eCommerce thing, so it's technically eCommerce but not really it's more lead generation. 

Now, I'm still loading this page, Craig. Whatever this image is here, there's some serious problems with it. And when we jump in and have a little look and see what's going on, we can see that there's pretty massive speed problems going on with it.

This is because they're using a WordPress theme, that's a very well-known theme. And because of that there's a ton of stuff that they need to fix and remove. Their time to first bite and their time to first content is just painfully slow. 

If the site doesn't load and work then it doesn't matter how good your SEO is, no one's going to see it. When I dig into this site, one of the things is a UX problem. If I click on Communities it's actually just sending me back to the homepage, it doesn't actually go anywhere. If you've got clickable properties, make sure they actually go to the physical pages. It's a bit of a nightmare when it comes to navigation so let's see if we can get this. 

The other thing is I don't really know what they do, which is weird because once the images actually do load there is tons of content on there. It says they've got pet-friendly apartments, cool, but then we look at the things like the titles and all the meta, it doesn't say anything about pet-friendly apartments. 

I would ask, what are you trying to rank and what do you care about? The usual stuff is wrong with it as well, so there's absolutely no internal links whatsoever, and for some weird reason, I guarantee the guy who bought this website was like oh no, you need to link to our partner's LLC website, which is totally pointless. I'd just remove that. No internal linking whatsoever. 

All of the copy is pretty much duplicate across the entire thing. For all your different apartments, ideally what you'd want to have is actual content that was original to the place. Do you think they've got any schema, Craig? Would you like to hazard a guess? 

Craig: I would say absolutely not because I've just literally clicked on a page and it's got the URL as the title tag. 

Ross: Yeah. There's like a huge issue here and their CSS is broken. It just looks very cobbled together. It looks like they've got a WordPress theme and threw a bunch of stuff together. This, for me, is a “burn it to the ground and start again project”. They're not ranking for their brand name which is obviously a major issue. 

Craig: So would you burn it down and start again? 

Ross: The problem is this theme is absolutely crippling you. The reason why themes are a bad idea or plugins are a bad idea, if you don't maintain them, a couple of reasons: they'll get hacked very easily. There was tons of news over the last week or so about people hacking plugins in order to inject links. 

Also, because they have to keep up, so just because you've got this version of Avada, there's 100,000 other versions of it. They need to have CSS and JavaScript files that apply to every single version that they've ever created. And Avada's an old theme, it's been going for years and years and years. So we're probably on number 15 or 20 iterations, but guess what? They have to give you the CSS and JavaScript for every single one.

When you're loading an Avada theme website, you're not just loading what you need to make your website work, you're loading every single iteration of what people need in order to make those themes work. That's why it's a bit of a loser in my book there. 

Craig: Yeah, it's a common thing, though. With certain WordPress templates or themes from Themeforest and stuff, they are cluttered, which leads on to the next question from Gini, Gini good to see you, and so she's asked, "How common an issue is this with WordPress themes that you audit?" 

Ross: It's constant, it's an absolute epidemic. People go for WordPress because it's fast and it's cheap. You know the old scoping triangle, do you want it fast, cheap, or good? You can have two. That tends to mean if it's fast and cheap then it's not very good, and that's what we're seeing in this instance. These people have paid less than probably $3000 to get this made, because it's very broken. 

When you're using WordPress themes, what I would advise, a) don't get a theme, just get someone to actually code the backend. Our Type A Media website; it's not a theme, it's a custom WordPress build. It doesn't need constant updates because we're not using plugins, we're not using any page builders or anything like that. The developer built a page builder into the back of it. 

A lot of people add plugins to add functionality. At the start it's great because you get the functionality immediately but then they all need updated, then it starts getting really heavy and starts breaking things. 

There's really no need for a lot of this stuff in the web, like jQuery sliders on a brochure page. That slider that they had, there's no need for that, just get rid of all of that. Their entire navigation is based off of JavaScript, so if I turn JavaScript off guess what? All those dropdowns, they don't work. 

To answer your question, I see that a lot, and it's just because it's cheap and easy but it gets you in a lot of problems further down the line. 

Do Mega Menus Negatively Impact SEO?

Craig: We're getting some more questions here, Ross, and this one is probably best answered by you. Someone says we previously mentioned mega menus in another webinar, and how they can potentially negatively affect SEO. So what are your great strategies for menus? 

Ross: I like mega menus, if I'm honest, as long as they're done tastefully and well. I would always put your top core categories at the start and then I prefer category pages rather than massive dropdown menus. If it helps the user get to something that they need, great. If it is just there because you're having trouble indexing pages or ranking pages then I'd say not great and that's kind of bad.

I like secondary navigation on the left-hand side of websites, so if I was to click in, to use Craig's example of biscuits from earlier. Click the biscuits page and then when we get into the biscuits page there's a left nav which has got chocolate biscuits, digestives, jammy dodgers, et cetera. And that's where those secondary links are. 

Because when it comes to Google bot, actually hitting the link then following it, and hitting the link and then following it, there is a bit of benefit for having it all in a big mega nav but if it's a smaller website I would just stay a million miles away from it. 

Make sure that when you turn JavaScript off, it still works, and make sure that if you've got a massive site with tons and tons of navigational links, site maps are really important and properly siloed site maps as well. Do you have a view on mega navs? 

Craig: I mean I would normally stay clear of them. There's nothing wrong with them but for the reasons that you said, making it simple and easy to navigate, it's a much better user experience and stuff for the customer as well, where I think on a personal level going on to a mega menu can be a bit of a pain. People can end up clicking the wrong stuff and as you rightly said, I'm not a fan of massive dropdown menus either, it just interrupts your flow and everything else. 

I'm sure it's okay from an SEO point of view but I think more from a user experience point of view, trying to keep things simple and I actually, I've been looking at trying to update my own website. Now, I don't have a mega menu or any crazy navigation as such, but what I am seeing across the board is if you look at a guy like Neil Patel...

Ross: I like Neil Patel, I respect him as a businessman. 

Craig: Yeah, very clever guy but if you look at his website it's very, very clean and minimalistic now. I think we're coming full-circle now where people are now trying to strip things back and make things a hell of a lot easier rather than a website rammed full of mega menus and trying to have everything on one page. 

For me, Neil Patel's website at the moment, very clean, very tidy, just very simple. You wouldn't go looking at it and go, "Wow, that's amazing." But mobile-friendly, fast, everything that you would want is there and I think minimalistic is the word I'm going to use here. And minimalistic and mega menu doesn't tie together, in my opinion. 

Product Variant Pages on eCommerce Websites

Ross: Caleb is asking, "For a business with four unique beverage products," hopefully alcoholic beverage products, "Four flavor variants. Should they have four PDP pages with a variant switcher or four collections with four flavors." Thoughts on that? 

Craig: I think when you have any kind of product, whether it's beverage or shoes or whatever it's going to be that's got variants like different shoe sizes or different flavor variants, what you don't need to do is have a page for every flavor variant. 

You should be able to have one page that ranks well and then obviously people can change the flavor or whatever. I think a lot of people make that mistake and they start to duplicate content or content cannibalization and everything else. 

Ross: Yeah, I'm kind of the same. The only time that I would disagree with that advice would be if there is search volume. If you're on a jeans page and you've got blue jeans, black jeans, red jeans, with all the search volume for all of that, so they should be separate pages. 

We've got Joovani asking, "What's your thoughts on blog commenting, guest posts and press releases? Do you think they are spam or do they work?" 

Blog Commenting, Guest Posts, and Press Releases

Craig: Press releases is more your bag, Ross. A real press release clearly still works. Guest posts still work very well. Blog commenting, absolute garbage. 

Now, the press release one though, as I say, the reason that I'm highlighting that is Ross does press releases and PR and everything else very successfully but there's press releases that Ross does and there's press releases, and the press releases that you may be referring to is some crappy submission website where you can put a press release and it spins it out to 500 websites and that's garbage as well. 

Ross's way of getting the high-quality PR links, definitely is the number one place that you probably would want to try to get a website and then obviously guest posts and whatever is down to whatever you want to call it. 

Guest posting is very much a real thing out there and I know people say, "Ah, that's garbage and whatever." They do work and that's it, it's as simple as that. But blog commenting, I've not heard anyone actually asking that for a while. I think the last time someone asked that they also asked does social bookmarking still work. These kind of things are dated and I wouldn't bother with those crappy spam tactics. What about yourself, Ross? What's your opinion? 

Ross: When he's like, “press releases”, should we put a bit of content on one of these press release wire websites and just spam them out? It's probably not going to do you any harm, if I'm completely honest. But be careful with your anchor text. 

Sometimes when you put a press release on a wire website, local news will pick it up and you'll get secondary links. I think as a tactic in isolation, yeah, probably fine. Blog commenting, no, just don't. 

Guest posting I can confirm for a fact that it works very, very well, just be smart about it when you're doing it. Make sure that there's at least 1000 traffic going to the site as my kind of main prerequisite for anything. In terms of DA and DR and all those metrics, they can be useful but I'm less bothered. I just care if they've got traffic or not. 

Ross: We're coming to the end now Craig. I think everyone knows where to find you but for those uninitiated that have been living under a rock, where can we find you? 

Craig: You can find me on craigcampbellseo.com. What about yourself, Ross? You still using the Type A Media brand? 

Ross: Just go to canonicalchronical.com and all will be revealed from there. Craig, thank you very much for taking the time. 

Craig: See you guys later.

All levels

Check out other webinars from this series