fbpx

Technical SEO & Site Migrations with Chris Green

Next up on my podcast is Chris Green. Chris is someone I would consider one of those technical SEO guys who love a good site migration with sites that have tens of thousands of pages and there is nothing at all wrong with that.

I’m rather jealous I don’t have the attention span to do big site migrations or fancy technical audits but we all have our strong points in this industry and getting the technical side of your SEO right is as important as anything else. Or if it’s a site migration went wrong, you can say goodbye to your rankings and revenue.

I was keen to get Chris on so we could discuss a bit more about how he goes about this type of work, what tools he uses and a bit more about the processes of a technical site audit, or a big old site migration. The podcast is 40 minutes long, I’m sure we could spend all day talking about SEO with Chris picking his brain and taking advantage of his willingness to share useful information with us all.

Below is a full transcribed version of the podcast.

Craig: So, welcome to today’s podcast where I’m joined by the very famous Mr Chris Green, dressed up especially just for me with a lumberjack shirt and the beard, I know I’ve got a cheek to slag anyone with.

Chris: This is outdoor-grade lumberjack gear, I’m wasted it in office.

C: Yeah, definitely. So, you are currently in Footprint Digital just now?

Chris: Yes, I am.

C: All going well in there? For anyone who’s listening, who has not seen you kicking about talking or whatever you may frequent, what is your day-to-day job in there?

About Chris Green

Chris: My day-to-day job in here is, well, it’s a mix actually, my title is Head of Marketing Innovation, which sounds impressive but in practice doesn’t mean a lot to some people. Actually, what I do is I help the delivery team, I help the sales team, I use my knowledge and expertise across our products and services, work a lot with our partners and doing some very cool R&D bits as well.

Chris: So, a bit of everything. I’ve worked agency-side for nearly 10 years now and I’m just utilizing my skills and knowledge across the base. I do a lot of training, I work with SISTRIX, ambassador for OnCrawl, and then handhold a host of other tools that I do a lot of work with. So, lots, I do lots, I have fingers in many pies if you put it that way.

C: But that’s obviously a line of success. Obviously you’re doing well and you’re getting all these other opportunities that come off the back of it. But for today’s episode you were looking to go down the route of discussing what you think, well, I think is a really good topic which we’ve not really covered on a podcast before, which is site auditing.

C: Off-air we were speaking about it and you like, or you said you liked migration, which drew that face from me when I was like, Ugh. Migrations are pain in the ass. Site audits are a pain in the ass. But they’re massively, massively important. I think for me there are not any super technical SEO, my skills lie in other areas of SEO I would consider. But I know enough about audit and stuff. I wouldn’t claim to be as geeky as you about it.

Chris: You need the shirt for that.

What is involved in a website audit?

C: Yeah, I need the shirt and beard, and you know what, it’s just boring stuff. But for me, a lot of people are still missing a lot of the basics when it comes to site audits and every time I audit a website or inherit a website or someone asked me to look at a project. You look at it, going, man, no one’s looked at this website before and people seem to be like, I’ve been audited and people have suggested this, that.

C: The first thing I wanted to discuss as me doing an audit, like running it through DeepCrawl, OnCrawl, Screaming Frog, SEMrush or whatever, is that a site audit or is that me just scratching the surface?

Chris: Yeah, an audit is many things to many people really. And there are degrees of it. If you work with tech SEOs of any kind of seniority, an audit to them is going to be a week, two-week or even a month affair — this big kind of forensic deep dive into everything. Whereas for me and my experience, most of the auditing that I do is with presales, so it’s before that introductory meeting, is what we are working with.

Chris: And my goal for that kind of presale audit is teaching someone something they don’t know about their own website, which varies – some businesses, they’re not really tech-savvy, that’s really easy because they might not even be aware of what their website is, but if anyone’s actually any good and they’ve got a marketing team and you can teach the marketing managers something they didn’t know, then from a sales perspective, that’s fantastic and that goes a long, long way.

Chris: The bread and butter of any SEO tech audit as we call it, I’d say, running a site through Screaming Frog or one of the various crawlers, OnCrawl, DeepCrawl. That’s the starting point of an audit, but that is but scratching the surface and it depends on the size of the website, the complexity of it and what the person’s looking for. If we just assume that my traffic is not enough, how do we make it better, then that becomes a Whac-A-Mole game of okay, what bits are broken or do I need to fix?

Chris: And actually identifying problems in an audit is the easy part. Pretty much all audit software out there is really good at saying when something is behaving unexpectedly or as it shouldn’t be. The actual hard part is, okay, well what is the fix for this website in this situation with the dev queue that it’s got with the budget that they’ve gotten, the crappy CMS that they’re running off, usually.

Chris: A successful audit manages to take all of that potentially weeks worth of work and gives it to a client in a way that they are awake long enough to digest that, A, this is good and B, they should do what you tell them to do.

Chris: That’s the perfect audit and anything in between that, I mean, I’ve not seen an SEO delivering identical audit to someone else, even within the same organization. Everyone does them differently.

Chris: Which isn’t a huge problem as long as takeaways are upfront and can be followed. That’s the take we usually have on it.

C: So when someone typically comes to you or the organization you’re working in for an audit, do you guys implement any audit or is this purely just a, here you go, here are all the problems, here’s what we’ve looked to resolve, go and get it done by a developer.

Chris: I’ve been in a fortunate position where I’ve often worked with the developers or with the partners who are doing the development work, so very often now we get called into audits when one of our partner agencies is dealing with dev or about to do the design or specking out the new project. We’re usually, we’re coming into a process that already has a developer with the inclination and means of fixing stuff.

Chris: The worst kind of audits are those that are seen as a standalone project because very often at the point of sale it’s not made clear that the audit doesn’t fix a thing. The audit is just a shopping list of issues. If you’ve spent, I don’t know, I think 10 grand, 15 grand, whatever, on a really big audit, well you need two or three times, four times the amount to fix the stuff that you’ve identified usually and very often that doesn’t get done.

Get Audit Changes Implemented Quickly

Chris: I think it’s no secret that within the SEO industry particularly a lot of the organizations you’re working with, the more of these recommendations are probably going to get ignored or stuck in a six-month-long dev queue. And that’s the nature of the beast.

Chris: Sometimes some audit side, just literally business case ammunition for, our website’s buggered, we need a new one, but you don’t need to spend that much on an audit for me to tell you that. It depends what the decision-making process is like, but you can usually tell the health of an organization by how used an audit is or how needed an audit is.

Chris: I think a terrible website and infrastructure usually means that there’s some decision making going wrong somewhere along the lines and they’ve got a bigger problem they need to fix elsewhere.

Do Small Sites Need Website Audits?

C: Yeah. So in terms of the sites that you’re normally working with when you’re doing an audit, what size are you talking, how many pages do these sites have because I’m assuming that widely varies, you get some local guy with 50 pages and potentially an e-com website with a million pages. Is it just the same process, it’s just a million pages that’s got to be audited or do you tackle bigger websites differently from smaller websites?

Chris: Yeah, that’s a good one. The vast majority of websites I used to work on are WordPress. You could usually tell what you’d need to audit and what would be broken just by the virtue of the platform. And you get to learn how people develop on WordPress, so you can usually tell are these blind spots are going to be missed, I’m going to need to take extra special care for that. The smallest site audit that I’ve ever done was on a four-page website, which I actually argued we didn’t need an audit for because, to be honest, I can cover that in a meeting.

Chris: The largest site, one of the ones that I’m running pre-sales on is plus 10 million pages and the approach on that, I mean that’s across over 15 different host names and actually the audit process on that is not how do we audit, what’s on the domain, it’s actually how do we partition here.

Chris: What are the CMSs underpinning? And I mean CMS is plural. There’s more than one. There are so many different systems and teams and businesses underpinning that and actually it’s not just as simple as, okay, well, Google sees this is the domain. We’re going to audit that domain because you can’t.

Cloud-Based Web Crawler

Chris: It would take even a crowd-based crawler weeks to finish that and how you approach that really kind of, well you have to take it as it comes. The biggest variables usually in an audit process are how many template levels there are because actually you don’t need to see every single page on their website to understand where the problems lie, especially if a lot of the content is dynamic or pulls from databases.

Chris: eCommerce is a brilliant example. You could have a page that has a quarter of a million products, but it could only have 10 templates. So one product, one issue, will be amplified probably across the whole base. Could be a faceted navigation system, could be a poor category. But that’s it.

Chris: You don’t have to check that on every single page, in theory. And I say in theory because if a website is built interestingly or badly, maybe I put that, you find consistencies and you find bits of content hidden or you find stuff that’s not on the CMS that’s elsewhere. So yeah, it varies wildly. Anything under a thousand pages is easy to audit though.

Chris: You can look at every single page, you can crawl it within, well not even half an hour, and you’ll know what you’re dealing with very quickly. Anything over 10 to 15000 takes a little bit more thought and you need to start asking for awkward things like log files and it takes them a long time to give you.

C: Yeah. So what are your go-to tools for an audit? I know there’s a bit out there. For me, I like to compare so I don’t like to just trust what SEMrush’s crawler says. I like to put it through Screaming Frog and maybe even DeepCrawl and then compare the data and then go from there. You use a similar approach or what tools are you actually using?

Best SEO Auditing Tools

Chris: Yeah, my starting point is almost always the same and that’s actually the search results. So what’s actually in the index, what’s in the robot’s text? I emulate a mobile device on the browser to see what the mobile site is working with. And then if I’ve got access, I go into Search Console and look at the top-level numbers as Google sees it.

Chris: That’s only because the size of the website and the complexity of the build may determine which crawler I need to go to or which data source will work. I need to size that challenge up first. I’d say that almost always I go into something like SISTRIX or SEMrush, similar things there. There are quite a few tools out there just to see, okay, well if there’s anything from Search Console that I’ve missed, what else can I see? How many host names has this been? How many pages is the site recorded to be at now? What has it been in the past? And then it goes on a crawl.

Chris: Usually if it’s a really big website or if the mobile website is that different from desktop, I tend to run a mobile crawl first and then compare it with the desktop following that. You can do that on Sitebulb if you want it desktop based or you can do that on DeepCrawl or OnCrawl or similar.

Chris: But I almost always use Screaming Frog as well just because like most people, that’s the one I learned to crawl on. I guess that’s my training wheels. Obviously the only complexity there being is the size of it and you know you just have to come and set things up differently. But the crawl will start the process.

Chris: And then in terms of how do I go then breaking things down, I click around the website and have a look and you can tell quite quickly. Are the titles garbage? Is the heading structure rubbish? I’ve got a nifty plugin that I keep forgetting what it’s called that tells me which frameworks the site’s built on, if it’s Angular or Vue or React or whatever, just to work out, okay, well what is the technology this is built on?

Chris: And then I also look at things, is it on Apache or is it on NGINX or IIS. That gives you an idea, okay, well this is what the service stack looks like. These are the problems I’m likely to encounter when going through it. And then I try and break it. So where is it likely to go wrong?

Chris: So can I break parameters or filters or search boxes and pieces like that because they’re the kinds of things that a crawl won’t pick up. Not really. They’ll pick up bits of it, but if you test it proactively, you can start to, have I got a really big problem? Yes or no, right at the start.

Chris: And then if I’m fortunate enough to have log file data, then one of the best things to do really quickly to understand what is the site like is run a crawl of the website and then check against the logs and what is Google getting to that isn’t existing in the crawl and vice versa. And then you start to see, okay, well what is the exposure of this website like?

Chris: Realistically though, anything under a thousand pages, I don’t generally touch logs because it’s just not that big a deal. Anything over 10000 pages, I insist on it. But you’re lucky if you get them. I’d say probably two in every 10 times I get access to them. And even then, you get a week’s worth of data or maybe a month’s worth of data if you’re lucky just because logs are big. And most people don’t realize that access logs are useful later on. So don’t just delete them.

C: Good tip. Don’t delete them. The workplace is the platform that you most frequently use, or people frequently use, not you personally, but everyone’s got Workplace, in my experience as well. What are the most common problems you find? Something that comes up time and time again with Workplace and that people can fix without having to go and spend an arm and a leg on an audit with guys like you?

Chris: Oh, right. Yeah. The real easy stuff. Most common one, I had one today, is their little checkbox that says please don’t index the website under readability. Robots block it. You still see that day in day out. There’s one site we were talking with earlier, that had been like that for two years and no one revised yet. They’d wondered where the traffic was.

Website Audit Quick Fixes

Chris: And you could say a savvier or less scrupulous individual might just go, I’ll do the audit and give me log in details and I’ll fix your website. And lo and behold, in a couple of weeks you can start to see the success. But other stuff, basics, not installing SEO plugins, so Yoast is my go-to, although I know a lot of others use different ones and then using that to make sure that, taxonomies, media files. The junk that WordPress creates needs clearing up and getting out of there.

Chris: And so few people are conscious of that. A 10-page website, how much of a problem is that? I don’t know. I’ve worked with a publishing site recently that had been around for 14, 15 years and there was a lot of junk in there that they just weren’t aware of and actually is where a plugin like Yoast can sometimes make things worse if you don’t know what you’re looking at.

Chris: They’re site maps that generate will tell Google where all of that noise is. So if the developers coded elements or post types that are elements of pages, they’ll all have their own URL somewhere and you won’t know they’re there. Google will know and they’ll get indexed, and all of a sudden your nice, relatively streamlined site becomes bigger and just full of more junk.

Chris: And again no indexing or blocking, well you just run through the taxonomy section, you just check, I do want and I don’t want. It’s a real simple task. If you’re building a new website, you’ve got Yoast installed and the developer knows what it’s doing, you can SEO that website within about half an hour.

Chris: The rest is just down to the underlying CMS is really strong but it’s just what we do to WordPress that makes it bad. So what plugins have we put on there? How much garbage is there? Is it slowing the website down? Is it a security risk? You can break core functionality, so you think that canonical is going in there but have you broken the template, is it actually garbage. I think they’re the most common things.

C: Yeah. It actually makes WordPress websites boring to SEO or work with because there’s a plugin that will fix most things…

Chris: …assuming it’s been built right. But whenever I’m throwing up affiliate sites or doing anything like that, I always go to WordPress because I’m not a developer and I know I can get something working within an hour or so.

Chris: You can’t do it on scale without someone smart doing the legwork for you and you just end up with a crap internal linking structure.

C:  Yeah. So I think Workplace is a great platform, and as you say, if you’re not a developer, it’s something you can throw up something very quickly. But if you were to have a choice, if someone was to have a standard website from a technical point of view, would it still be Workplace

Chris: Yeah, I think so. I’ve worked with Drupal, Joomla, Magento, WordPress, and if you have a bigger sort of enterprise platforms… Fundamentally it’s just what you know, right? A lot of people will say that .net websites are terrible. I think Kentico is one of the CMSs built off that, and actually it’s not, it’s as good as the developer that’s built it. And it’s as good as what you know. I’ve seen far more badly built WordPress websites than any other platform, but that’s because anybody who isn’t a developer, like myself, can say, Oh I can build a website and it’s probably not going to be that good because you don’t know what you’re doing. You stick with what you know.

Chris: I love and loathe Magento. In some ways, Magento can be great to work with it, and other ways it’s an absolute pig. But again, I understand it. I know how it works. So if you know how it works, you can usually find out how to fix it.

C: There have been bits of bollocks with that, but talking about migrations, migrations is another thing that you enjoy and… Obviously audits and migrations go hand-in-hand anyway. We actually had a migration last week with our website and migrated over to UKFast. By the time the Monday came… The website was up and running on Friday, then the Monday came, I think there was some automatic plugin update, some PHP update with the server and the whole thing blew up. And we were looking at love hearts across the website and everything else.

C: There was a server issue with UKFast. We actually took them away from UKFast because they couldn’t resolve it, believe it or not, but migrations very rarely go easy. And this was a fairly small Workplace website, migrated over onto another server and things went tits up. Things like maybe the PHP version being wrong and so there are so many things that can go wrong with a migration. People think we’ll use the all-in-one migration plugin for Workplace, import the website, job is done, there’s never any problems. When it’s bigger websites, I’m assuming more problems arise.

Planning your site migration

Chris: The bigger or more complex the project gets, just the longer you need to spend planning the thing. So for example, that instance of moving it from wherever it was to UKFast, I would probably have moved the whole platform or created a copy or a version on it and soft-launched or tested on a different domain may be password protected, just there first, just to make sure the stack is compatible. If upgrades eat like on the server or the box or you know even plugins or app-side level stuff is causing issues, then that’s the website has been built badly, usually.

Chris: Someone’s hacked a core file or something. It’s just not going to work right. If it’s planned well, I’ve been fortunate to work with some really, really great developers and often their intelligence has made me look better because these processes, which are stressful, even though I like it, it is stressful. If you have someone that knows what they’re talking about, that helps plan it, that’s great.

Chris: I think where I’ve learned how to work is find someone smart on the project team, make good friends with them, and make sure that they don’t run off is the key one. And the second is just the project management side. This is expectation management and it’s slightly more, dare I say it, boring handholding, less actual SEO stuff. More account management elements. Just make sure everyone knows what the hell everyone else is doing.

Chris: If you’re going to your developer or your SEO at the 11th hour saying, here’s a new feature or we’re about to launch, it’s probably going to go badly. But that’s just simply because we don’t have time to do what we need to. In one of the talks I gave fairly recently, I had an analogy of a ship heading to pier about to crash a cruise liner. And the idea is the sooner you start turning, the easier it is to avoid the pier.

Chris: That’s the analogy of, well, in the 11th hour you get told that you have to change direction. Well, you’re probably not going to succeed. And then you’ve got awkward phone calls to project stakeholders and whatever saying, I’m sorry you’re going to lose money because of this. And by that point, no one enjoys that.

Website Migrations

C: It’s painful stuff, migrations, and something that I try to stay clear of, the website, the thing’s anyway. I think it’s best left to people who are experienced in that field. But as an SU, sometimes you just get railroaded into things to help people out with specific situations. But yeah, migrations are never one of my things.

C: But, in terms of migrations, again, I’m fair, I wouldn’t say weak obviously. Moving a Workplace website over to another server is fairly simple and easy to do regardless of what way you do it. But are there any other cool tools that can do it quicker, am I missing a trick? I can do it manually. If someone sends me a zip file, you want to connect, I know how to do that or I can use one of the default plugins but is there another way of doing that or not?

Chris: If you’re going from platform to platform… So WordPress to WordPress, then yeah, there are quite a few. I still prefer taking full export and pushing the full export into the new platform. And then obviously FTPing any files over. I think actually there is some benefit to doing it manually. With a provider, you need to know what’s there because if you don’t know what’s missing. And the problem with that is if you’ve got a WordPress website, you’ve hosted any content outside of the CMS, all of the automated solutions will miss that.

Chris: If you’re going from one platform to another, it depends on the direction of travel. So there are quite a few solutions, although I can’t quote off the top of my head, Drupal to Joomla or Joomla to WordPress, there is like an ecosystem running there, but they’re all variations of importing one database from one system into another and varying degrees of complexity.

Chris: There’re little hacks that I tend to use though is I usually have a Screaming Frog or another scraper to crawl or to extract content, pull it out into CSV format and then you use CSV importers rather than out of the box. So I choose what elements I pull in by which elements of the page I scrape and pull in. And there are quite a few developers that I’ve worked with that are quoted hundreds of thousands of pounds to migrate database content, writing scripts and finding the best way. And I’m like, well, no, the CSV importer in WordPress is really straight forward.

Chris: 40 minutes later there are 10 years of blog content that’s gone in without a hitch. So that’s a pretty cool thing to do, which I probably should charge more for if I’m being totally honest.

C: That was going to be the next thing. For migrations, audits and everything else, your… I can’t sit here as I say claimed to be an expert in those fields, but obviously I’ve been in the industry long enough to know that the price for these things wildly vary and obviously I’m not going to sit here and say tell me a price, but how does someone know whether the price is value for money because there’re guys out there who say 10 grand for a full on-site and it could be a bag of garbage and they’re probably getting some SEMrush report, which you know if the person paid $99, they could defend themselves at the push of a button. How do you determine what is value for money when it comes to audits?

Chris: Interesting. In terms of the cost, it’s usually time-based, this is what time it’s going to take us. In terms of what’s value, because you’re right, you can spend, I mean you could spend a 100 hours, 200 hours auditing a website and actually get the same value as spending 50 just because those 50 were well-placed.

Chris: The critical part is working out well what do you need to actually gain from this. And there have been times where you’re auditing a website or you do the pre-audit and you’re like, well no matter how much I do to the onsite technical of this site, you’re not going to actually get where you need to. And it’s having that conversation up front and early because the recommendations will be, you need more links, you need content for better links or you need whatever strategy you’re going to need to acquire those.

Chris Green:
And the technical, it’s just going to be variations of a theme. You’re going to be tweaking some nuts and bolts. The worst technical I’ve seen, you’re missing alt tags on your social media icons or crap like that, which makes no difference. Agreeing with what is in scope and out of scope to start with, but then going, well what is the output? You’re going to have five, 10, 15, 20, what have you, quick-win recommendations and I think the best audits have a degree of that upfront.

Chris: Not to give it all away, but to say, well based on what we can see, this is what we’re going to provide. If you’re doing an analytics portion of an audit, but you’ve got a website that gets like a hundred users a month, like what is your time in analytics going to be well spent in doing, because it’s not going to tell you a lot.

Chris: You can tell if goals are set up or if you’ve got crap data, but that’s it. Migrations are slightly easier actually, especially if you’re running eCommerce because you know what the average order value is, assuming analytics are set up, what your conversion rate is. So you know the financial cost of losing traffic. And if you base that on what happened last year or you can use some forecasting to say, well, we reckon that this subset of pages is going to have 3000 visitors, you might lose 10000 quid if you drop that.

Chris: That is an easier sell. An intensive, when you size up a migration, this isn’t exact, but I tend to find it so anywhere between 10 and 20% of the total project costs for a website needs to be migration and SEO.

Chris: You want it to go well. Now there are so many caveats and I don’t want to say it depends and fall into the SEO cliche. But if it’s anywhere over that, then you’re almost spending as much on this as you are development or design, which very often struggles to provide value. If it’s significantly under that, you’re under-investing and you might be at risk.

C: So eCommerce websites are really simple. If you’re doing like lead-gen or you’ve got a brochure website, that’s much harder to put a figure on because what does it do now? How much value does the website generate as it stands? At a minimum, migration should be protecting value. It’s just making sure that just building a slightly better-looking website doesn’t make rankings go up.

Chris:  It’s standard, which I hear quite a lot. You want to counter that early?

C: Yeah. In this industry, in general, for anything, way of living, if anyone’s lost interest, make sure you do your due diligence and make sure you’re approaching the right people that actually have history and experience, guys like Chris, who have been in this industry for a long time and obviously you speak at a lot of events. How is the speaking thing going? Have you got a lot lined up this year?

Chris: I’m doing a lot more workshops and tuition this year, so I’m doing a couple of days lecturing at Greenwich. We’re running our own Academy in the summer outside of Colchester in Essex. I don’t know if I’ve got any big events in the pipeline. No, none that have been announced. So I’ve just been a bit so off the mark if I’m being totally honest. I’m terrible at saying no. So I don’t know if anyone is looking for people to speak and it’s more than two weeks in the future, then go for it.

C: Yeah, I think people pull out of speaking events and stuff, so Chris’s one of those guys that can obviously fill slots if anyone’s looking for speakers. Obviously spoke at Brighton and many other places. Optimize, which is a decent event as well. I know you’ve been to a few of the same ones I’ve been to. Optimize is always a good thing as well.

Chris: Say hi to Andrew, he’ll give you beer and pizza as well.

C: Yeah. I think the second one. I don’t think the beer was… It probably was flowing fairly well, but I think it’s obviously getting better.

Chris: Yeah. I need to get back there sometime and revisit it. I think it was more than two years. One of those or it was one of his first ones I’d done.

C: That was a great event. Great to meet people down there as well. Need to get back. But sadly we are out of time, Chris, but it has been a pleasure having you.

Chris: Good talk as well. Literally, wasn’t as boring as some of them.

 

seo profile image

Craig Campbell

I am a Glasgow based SEO expert who has been doing SEO for 18 years.

  • social media icon
  • social media icon
  • social media icon

Online Courses