Technical SEO might not be as fun, or exciting as producing great content and seeing the links pour in – let’s face it, not much is – but that doesn’t mean the results can’t be just as good.
Before I get into the case study of how I doubled a client’s SEO traffic from a few quick fixes with some technical issues, let me tell you a story.
Are you sitting comfortably? Then let’s begin..
In 2003, the British Cycling team were a bit of a laughing stock around the globe – although I’d love to see any of us regular humans do what they did!
Since 1908, the team had only won one gold medal (again, more than us regular people) at the Olympics and the at the world’s biggest cycling race, the Tour de France, the team had racked up a pretty embarrassing (for them) 0 wins in 100 years.
Let me make clear how bad things were for the team.
One of the leading manufacturers of bikes REFUSED to sell the British team any gear as they were worried their poor performance could hurt sales – ouch!
Now, I’m guessing you’re wondering why the sob story, and how does this relate to SEO?
The team’s fortunes began to change in 2003 with the appointment of a new performance director, Dave Brailsford.
His biggest philosophy was a simple one – it’s all about “marginal gains” Or, to you and I, what little things can be changed to move needle of fortune that little bit.
As long as the measurements improve fractionally, he classed these improvements as a win.
After all, 1% improvements in different areas add up to significant improvements in the long run…
To give you an idea of how little some of the improvements Brailsford made to what the team were doing, I’ll highlight just one of the changes that sticks in my mind.
He hired a surgeon to come in and explain to ADULTS how to wash their hands.
I’m not making this up.
His theory was, if the team don’t wash their hands properly, then the risk of them getting ill increases, as well as the possibility of passing germs onto other team members. If they are ill it means they can’t train, and therefore they could not improve.
It was all about the small victories
And it got them, within 5 years, to a huge improvement across the board. (More on this later – I won’t leave you hanging!)
Now, I bet you’re thinking “cool story, but how does this help SEO?”
Well, technical SEO is very simple. It’s all about the marginal gains. Every little thing I am about to cover on their own isn’t going to make much difference on its own, but combined it could – and has – led to one of my clients doubling the amount of organic traffic to their site..
Firstly, let me thank the client for letting me publish this. I won’t be revealing who they are – but once again thanks for letting me publish your story.
Back Story:
The client got in touch with me around the middle of March, telling me that their traffic had been flat and requesting a complete technical audit and server log analysis to see if they were missing anything.
Once the audit was complete it highlighted quite a few errors on the site. Some were basic, and some a bit more technical.
I was then asked if I could make the changes myself. I already knew the site was a WordPress site and the majority of issues I had tackled before. Plus, the client said I can use it as a case study and said he said he wouldn’t build any new links or update any content that I didn’t ask him to do.
The part about the case study got me interested. After all – I know this stuff works, and the client gave me the perfect opportunity to show you exactly how it works.
But, before I go over what I did, I bet you’re asking did it work?
Let’s look at the results before we get into the go any further – no point in reading what I did if the results didn’t come… And no, there are no Olympics in gaining organic traffic, but if there were, I reckon I’d be in with a chance.
The site’s organic traffic nearly doubled, as you can see from the screenshot from his Google Analytics. I don’t really like to use rankings report as success, as I’ve never been able to pay an invoice with rankings or deposit them in my bank account – but they DO matter for my clients.
Honestly, I would like to have shown the revenue increase for this site, but the client didn’t want to have this published, and I respect this. He also didn’t want to reveal the % of organic traffic to his site. This is why I had to black it out. (I should mention, however, I don’t like seeing graphs for traffic when you can see the segment, so easily could it be cheated simply by buying some traffic and passing it off as organic). I can assure you this is not what I did though.
Organic is a huge percentage of the traffic, so it has made a huge difference to his bottom line – something business owners care about. They currently don’t do any PPC or use any paid methods at the moment.
SEOs care about rankings report, owners care about the bottom line.
So, what did I do to make a difference over such a short period of time?
I have grouped these into themes and tactics rather than the order in which things were fixed. I get that not all of you will have the same issues, but employing some of these tactics could help a variety of sites gain that much-needed traffic.
Speed:
The site was slow. I mean really slow – it was on some shared $5 a month hosting. Nothing wrong with this type of hosting but when you are serious about your site and it’s getting a decent volume, you need to pay for decent hosting – no point in getting all that traffic when they can’t get to the parts of your site they want to quickly.
We moved the site to Cloudways and configured HTTP2 (the site was already https).
There are other hosts out there, but I like Cloudways. The site was originally loading in around 16 seconds and this reduced it by around half – just over 7 seconds to load.
The rest of the speed work was around critical path CSS above the fold, removing unused CSS, caching, and the usual bread-and-butter stuff. The site now loads in about 2.7 second. Yes, there is still some work – but without compromising the site’s functions it’s about the limit at the moment.
One of the other key things we changed (and I see a lot of sites making the same mistakes) was that the site used a Google Font. This is great in some ways as it means it can load pretty quick. However, the by default it imported ALL Google fonts, which meant a lot of resources were downloaded that weren’t needed. So, I stopped the theme downloading all the fonts – limiting it to the one that was needed. This gave the site a huge saving in file size.
Large Images:
This is kind of related to the above, but I identified that the majority of the images on the site were large and never had been compressed.
As there were a considerable number of images on the site and I didn’t fancy the manual task of downloading each image, optimising and reuploading using something like compressjpeg.com – I bought a plugin for £4.99 and it worked its magic. Sometimes, if a plug-in will save you time, it’s well worth paying for, especially the amount of time these types of plug-in help to save.
I am not sure quite how long it took. I set it going around 4pm and by the time I got back into the office the next morning everything was done – but I reckon if I’d billed by the time it took me, it would have cost a LOT more than £4.99.
Redirect Links:
While the site had no 404 pages found via the crawl, a considerable amount of the pages were going through redirects. In fact, there were some serious redirect chains on this site – the biggest one was 5 levels deep – which meant that Googlebot would’ve probably given up and gone home.
It took a few crawls to find every single one and fix them, but as well as saving crawl budget it meant Google wasn’t jumping through hoops. This also can help users as the page loads quicker without going through a redirect.
Non Secure links:
There were just a handful of these, but in there were a couple of links to the old HTTP version of the site. These were quickly fixed. I am not certain, but once they were fixed, Google didn’t crawl the same pages as often. My theory is they were crawling both the secure and non-secure version of the site as they were finding links.
It didn’t help that HTTP wasn’t 301 to https – that was another issue that was quickly fixed.
404 Pages in the Logs:
This one was a bit trickier to spot, and it was a good job I had done some server log analysis. Basically-speaking, the site had quite a few external links pointing to 5 different URLs which had been deleted.
The guy had done the correct things and removed all the links on the site which meant they weren’t found via a crawl using Screaming Frog.
It was only when doing some analysis in SEMRush that I realised why Google was crawling these dead pages. After all, I know Google doesn’t guess URLs.
They had external links pointing to the pages, but no redirects had put in place, so Google was landing on dead pages.
Simply installing the redirect plugin and adding 301 to the new relevant pages got that sorted.
Lack of Internal Links:
I’m pretty sure this had quite a positive impact, despite the fact that so much was being done at once. One place I know this had a positive impact on, was the number of people visiting a second page.
This site was one of the worst I had seen for short-changing on internal links. Around 70% off all the pages only had one internal link – and this isn’t great for Google.
Fixing this was a huge task. It took several days’ work as I didn’t just want to place the links anywhere – I wanted to place them on relevant articles and pages.
Schema:
The other significant thing I did was to install schema on the entire site and really in-depth. And not just a basic schema that comes out of the box. I really went full out on and included all the schema available to make it easier for Google to understand the pages.
And the Rest…?
There were some other minor things I tweaked and edited, but I didn’t want to make this article about every baby step I took. After all, every site is different.
And it’s important to note that in no way should you think that content or links aren’t important. They are super important, but by fixing the foundations of the site, the content and links were able to have a greater impact.
The purpose of this post is to show that marginal gains are important and even when a task seems small and you think it will have no impact it more than likely will.
Whether you can measure each baby step and say that each specific task resulted in ‘x’ change is something I’m not sure I’d spend hours looking at.
Instead, it’s important to ‘audit what matters’ – and that way you know that if there is an issue it will be causing you problems.
How Successful was Dave Brailsford’s Team Sky?
By the way, if you are unaware, and I know a lot of my audience aren’t from the UK, Dave’s strategy was, it’s fair to say, pretty successful:
In fact, they won an incredible number of medals (60% of all those available at Beijing 2008), and golds at every Olympics since.
The team also won several Tour de France titles under the “Team Sky” and Brailsford’s method for success has since copied by several of the other major nations.
If you really want to see the stark contrast check out the full list of medals won here.
Whether is cycling, technical SEO or anything else – the strategy of “marginal gains” could well be a good strategy to follow.
Again, as a final note, I really want to thank the client again for letting me publish this case study. It has been a great few months working together and I will be keeping an eye on their analytics to see where the numbers end up!
Hi Andrew,
nice work! Can you tell us a bit more about the “Schema” part? Did you use Microdata or JSON-LD? What WP plugin did you use?
Thank you!
Hi Mathias
It was Json-LD schema that was added. While there is nothing wrong with the plugin I used, I am currently testing a much more detailed plugin offering a lot more features.
Once I have finished the testing will share that one in the next day or so.
Thanks
Andy
Andy, By any chance, do you track the performance of the schema markup? Especially if we have multiple types of schema codes.
Love what you do, Andy 🙂
Keep more content coming. Log analysis rocks always.
Cheers
Suresh