Use edge/CDNs to stitch together and publish content and build SEO results

Nick says: “My number one tip this year is a continuation of last year’s tip - it’s to better understand the edge platform for deployments. That’s where you’re deploying content or stitching content together. You need to understand edge as a way to change the content and experiences that you’re delivering to your users.”

What do you mean by stitching together and publishing content on the edge?

“Edge is a CDN (or Content Delivery Network) platform. This can be Akamai, Fastly, Cloudflare, etc. These are the platforms that deliver your content to the end user, no matter where they are. We’ve been using these platforms for many years now to successfully deliver content - not only locally, but faster – and we have essentially been doing this for performance. Now, these platforms have become more and more advanced, and their computational power has become more advanced.

This relates to Moore’s law, which we’ll all know as geeks. The CPUs have become better and the memory’s better, so these platforms can now handle much more complicated requests. We first started using edge to do requests that happened in the header, like redirections. Now, as the capacity of the edge servers and the edge nodes has increased, we can start to do more complicated things.

One of the things we can do is take the request, take the entire page, and unpack it and change it. We can pull in different content from other places, and we can essentially change the entire page that the user then receives, and then deliver it on to them. Your edge platform can be doing anything that you could dream of - in terms of changing that content request and serving it to the end user - which is incredibly exciting for everyone who’s working in the edge area.”

Can you take an original piece of content published on a web server, pull it apart, and then insert personalised content in there?

“Absolutely you can insert personalised content, but even just generalised content. Let’s say we have an eCommerce site. For years, SEO teams have talked and dreamed about getting more content onto product-level pages, but the pushback has been from the dev team. The dev team has said, ‘No, we can’t change the template, we’ve got a long backlog, we can’t do this.’ If we come in on the edge, we can take those product-level page requests, we can inject blocks of content onto those pages, and then serve it to the end user. By the end user here I mean the user, but I also mean Googlebot - because Googlebot and other search engines will see this as the request.

We could be using the edge to insert the content that we always dreamed about, to rearrange the nav menu in the way that we’d always wanted, or to just completely rearrange the page in any way that we wanted to. This could be a permanent change that we are making for our end users (or for Google) or this could be a test that we are running over a period of time. Maybe we wanted to change the nav over a six-month period for the entire website. The edge platform is a way that we could do this.”

For an eCommerce store that has some kind of seasonal offering that you want to give greater emphasis from the navbar, would this be a quicker way of doing that than your existing system?

“It could definitely get you out of that kind of problem. Say you wanted to insert a whole different set of links into the nav for Black Friday – the edge would be a way to do that. It is a whole development platform, essentially.

If you think about your origin server, the edge is an entirely new layer that you can do this development on. As you make those changes, and that development, it then gets served to Googlebot or the user as the actual request. It’s a second chance for us to change the page that doesn’t involve the origin server.”

When and where is the edge most useful? Is it for presenting a slightly different message to users at different stages of the buyer journey?

“It can definitely be used for that. You’re touching on personalisation, and you do have those options with the edge. You can serve different content to different users, different groups, or even to people in different locales. Some of this could have a huge advantage for SEO. We do kind of veer away from SEO when we go down this path, though, because it’s more about personalisation.

For us, part of the personalization process that we’ve always been interested in is: how do we personalise for Googlebot? For this question, the intersection between SEO and edge has been particularly interesting.

Googlebot is a user of our site and, historically, we’ve always been quite worried about that. We don’t want to present something completely different to Googlebot than we present to our normal users, because that would be cloaking. There can, however, be circumstances where it’s advantageous for us to fix a certain problem for Googlebot that we couldn’t justify fixing in our normal development workflow. There are things that we might choose to do for Googlebot that could be advantageous to that crawling and indexing of the site.”

Does Googlebot ever know that any of this content has been stitched together?

“No, Googlebot will always see the end result. Even in the future, Googlebot has no way of detecting that this request has happened on the edge server.

However, all the usual rules apply. You have simply got another opportunity to change requests, which is the same as changing it on the origin server. You still have to consider what you’re doing for your users, what would be considered cloaking, etc.

That’s where the whole area of personalised content has become much more complex over recent years. We want to give a personalised experience to different users, but we also want to be able to control the experience that Googlebot has - because that is the experience that gets indexed and ranked. There is a grey area there, but the rules haven’t changed. The edge is just another means of deployment, or another way to change the request.”

What are the benefits of using the edge to control and edit content instead of a conventional CMS? Is it generally faster and does it help SEOs that are dealing with old CMSs that don’t have the same functionality as modern CMSs?

“All of the above. You may also find that there are blockers in terms of development, and the dev teams that you have on the origin server. They may have a very long list of backlog items that they’re working through. You may find that you’re dealing with multiple CMSs, which we often do as SEOs. We find that we have more than one CMS, or we’re mid-migration - and having a layer that goes over all of these CMSs can be a huge advantage. You can start to make changes that affect every single request.

You may have legacy CMSs (I’ve worked with clients that still have Perl CMSs, for example) that you simply can’t do anything on. The ability to add SEO functionality to old, legacy parts of the site, as you upgrade them or move them across, is fantastic. There are lots of advantages to doing this.

A lot of the advantages for SEOs are ways to get around blockers, but there are also advantages that are particularly pertinent right now, with the macroeconomic climate that we’re in. We know that a recession is around the corner, teams are under pressure, and we’re asking to do more with less, which always happens.

The edge is one way of getting around this. You’re using architecture that’s already being bought by the company. They’re looking to leverage that expenditure, and it has usually been brought in because it helps you deal with security, bot management, content delivery, or even performance. Leveraging that existing spend is very attractive to senior management. Also, you have the ability to do this without any additional software or platforms. You’re adding things to a platform that is fully covered by your security policies. You don’t have to worry about any kind of performance impact, because this is how your content is served. You’re not going to take these requests off to a third party, or a third-party platform. You’re serving them through the CDN that you already have set up.

There are a lot of reasons why people are doing development in the edge environment, because they see it as an extension of what they can do on the origin server. You can develop in both the origin server and on the edge.

You can also deploy very quickly because you don’t have to worry about all the interconnecting parts, the politics, or anything else. You can fix something on the edge in a three to six-month period that may take years, or be so blocked that you can’t possibly see a solution.

You may want to do that permanently, or you may want to do it as a test. We talked about changing the navbar, and the SEO team could have the hypothesis that if you had more control over the navbar - you were more reactive, and you could start to put links on there quicker - then that would help your SEO. The dev team might have fed back and said that it’s impossible to have that kind of control over the nav. You can then implement a quick nav change on the edge, and the SEO to improve that by having additional control over adding and removing links. That can then benefit SEO in all those different areas. You can choose, then, whether you keep that functionality on the edge, or say that this functionality needs to be implemented on the origin server because you have made the business case for it.

You’ve proved why you needed it and you’ve shown real-world results, so you can say why it should be prioritised in the dev queue. It’s a very powerful additional tool to get things moving within a company. With the macroeconomic climate that we’re in, companies are looking at being able to make these changes faster, and the edge is a really good way to be more agile.”

Are there any cutting-edge edge/CDN developments that SEOs can take advantage of in 2023?

“It’s hard to know that they’re happening because you can’t actually detect them on the edge. Large companies and large brands are doing more work on the edge than ever, and they’re all very interested in this area, but it’s hard to point to many specific examples.

It’s very interesting work for SEOs. There’s never been so much opportunity to make large-scale differences for the brands that we work on. In my work now, as an edge architect, I’m largely brought in to solve the unsolvable - which is my remit in a lot of the work that I do.

One of the advantages of being a smaller firm was always that they were more nimble and able to make changes quickly but, with the edge, enterprises are catching up. This will particularly help large enterprises because they are slow. They have so much politics, development, multi-CMSs, etc., and everything that happens in enterprise creates these blockers. The edge is particularly interesting to them.

However, companies of any size can benefit from it as well. Cloudflare (which is one of the leading CDNs) is essentially free for any company, and there are an increasing number of people producing scripts that can run on the edge. I think Torque (my company) even got the Majestic tags and functionality working through Cloudflare, which we released on the Majestic blog. You’re going to find a lot of people making things work in that environment, which can benefit every company of every size.”

What shouldn’t SEOs be doing in 2023? What’s seductive in terms of time, but ultimately counterproductive?

“I’m going to say AI content. The reason I’m saying that is because I’ve actually worked a lot with AI content. We produce it, and it is even one way of deploying the edge. You can create content through the AI content production process, and then use the edge to insert that into certain pages to get it working for Google and users.

What I’ve learned from that, however, is that AI content is complex and if garbage goes in, garbage comes out. A lot of people are looking at AI content and thinking that it’s very easy, because you can simply produce content, so it’s going to reduce the costs that you spend on writers. I would disagree with that quite strongly. To get good quality AI content, it’s going to move the emphasis from writers to editors. It can produce something that’s very high quality, and grammatically accurate, but it does need an extra bit of work and workflow to keep the quality up.

AI content is very interesting, and it’s going to be increasingly interesting in terms of producing text and content for SEOs over the next few years, but it’s not a shortcut. It’s going to involve you reallocating those resources into workflow. What’s the thinking behind it? What’s the quality of it? What’s the meaning of it? What’s the intent behind it? Don’t fall into the trap of thinking that you’re going to cut costs dramatically by producing everything with AI.”

Nick Wilsdon is the CEO and Founder of Torque. You can find him over at torquepartnership.com.