Blog

Howdy, Moz fans, and welcome to another edition of Whiteboard Friday. This week we're going to chat about structuring content, placing content, and placing links, specifically with regards to some things that have come up over and over again in the SEO world, but still seem to be a challenge for many of us who play in the field.

One of the questions that I'm going to start with is around subdomains and subfolders, because this just comes up again and again and again. I think one of the reasons it's emerged in the last few years is, unfortunately, some statements by Googlers themselves -- a statement a few years ago from Matt Cutts, and one, I think last year or two years ago, from John Mueller basically saying, "Hey, Google has gotten much better at identifying and associating content that's on a subdomain with the main domain, and you don't need to worry about placing content on two separate subdomains anymore."

I am sure that Google has actually made strides in this area, but this question still has the same answer that it did years ago. I'll show you some examples.

You're asking, "Should I put my content on a subdomain, or should I put it in a subfolder?" Subdomains can be kind of interesting sometimes because there's a lot less technical hurdles a lot of the time. You don't need to get your engineering staff or development staff involved in putting those on there. From a technical operations perspective, some things might be easier, but from an SEO perspective this can be very dangerous. I'll show you what I mean.

So let's say you've got blog.yoursite.com or you've got www.yoursite.com/blog. Now engines may indeed consider content that's on this separate subdomain to be the same as the content that's on here, and so all of the links, all of the user and usage data signals, all of the ranking signals as an entirety that point here may benefit this site as well as benefiting this subdomain. The keyword there is "may."

I can't tell you how many times we've seen and we've actually tested ourselves by first putting content on a subdomain and then moving it back over to the main domain with Moz. We've done that three times over that past two years. Each time we've seen a considerable boost in rankings and in search traffic, both long tail and head of the demand curve to these, and we're not alone. Many others have seen it, particularly in the startup world, where it's very popular to put blog.yourwebsite.com, and then eventually people move it over to a subfolder, and they see ranking benefits.

But even more telling was a recent example from just a few months ago on the iwantmyname blog. Timo Reitnauer wrote on the iwantmyname blog a piece about how he had moved their content actually from the main domain to a subdomain, which is not usually the way we see things going. But, man, when that happened, ugly, super, super ugly. You can see his traffic graph. He graphed his Google search traffic and showed, from his Google Analytics, a nice snapshot of when they made the move, what happened to their search traffic, how long it took to recover. It actually still hasn't recovered. It's been five or six months now. So very, very frustrating for them, and they're going to move it back over. I think maybe they already have moved it back over.

But this was great in that this piece went to the front of Hacker News. Lots of folks from around the startup and technology worlds commented on it, shared their experiences and opinions as well.

Bottom line is it's really dangerous to put content on a subdomain still. I believe John and I believe Matt when they say that Google has made strides in this direction. The problem is they're not good enough or perfect enough to rely on that factor, and so I'd really urge everyone to keep your content on one single sub and root domain, preferably in subfolders. That's how you're going to maximize your potential SEO benefit. This is one of those technical SEO things that just hasn't changed for many years now.

Next up, a question around rel=canonical, and 301s, especially cross-domain rel=canonical, meaning people are pretty comfortable with the rel=canonical that sits on your own website on your pages and maybe says, "Hey, the print version of this page should actually be considered the same as the web version of this page. Or the mobile version of this page should be considered the same as the regular version." That's fine.

But folks have more questions when it comes to cross-domain rel=canonicals and content that perhaps they own because they own multiple websites, or they have licensing agreements across those, or they have business development or partnerships, those kind of things. So they're wondering, "Hey, I've got content on multiple websites, and I want to move some of that content, or I want engines to interpret it all as coming from my site. Should I put it on the other website and use a 301 to redirect it, or should I use a cross-domain rel=canonical to say the old page is now the new page?"

Well, actually this is one where, from a technical perspective, the engines are doing a pretty darn solid job. Google is doing a very good job. We've seen Bing make strides here with cross-domain rel=canonical. They seem to be doing a pretty good job as well. I haven't tested them as intensely though recently.

Basic story is with the 301, other site.com/a can redirect to your site.com/a, and both visitors and engines, anyone requesting the old page, get the new page. The only difference with the rel=canonical is that when a visitor requests the old page, they're still going to get it. They're still going to get that othersite.com/b. Search engines, however, are going to get the new version of the page, or they're essentially going to consider these to be one and the same.

What we've seen is that, in both of these cases, the ranking signals seem to be passed very similarly, if not perfectly similarly, very similarly. It's hard to detect any difference there. But the rel=canonical can give you an option whereby you say, "Hey, I want to maintain the branding or some unique aspect of something that happens around othersite.com, and so I wish that I could have visitors be able to still go to that page, but have search engines know, hey this is actually just a copied version of this one, and if you're going to rank one of these two, I'd prefer you to rank this one."

That's a great use for the cross-domain rel=canonical. But this is much more a user experience and a branding experience issue than it is a technical SEO one, because both of these work pretty darn well.

Then the last issue I'll cover today are around some of this content and link optimization stuff is: What if you've got multiple websites all linking back to your main site, and you're wondering does that or would that help my SEO? I can't tell you how many folks, surprisingly even folks who are very savvy, who have done lots of other stuff in the technical and web development worlds, are thinking about this from a SEO perspective.

I can understand where it comes from. Basically, you have this understanding that more links is a good thing and that more link diversity is a good thing. So you think to yourself, "Hey, maybe I can capture more links and more link diversity by having more slightly different websites. I want to keep my main site all about one particular topic or one particular niche of that topic, and I want to have these other niche sites that maybe I have some great domain names in my portfolio or some really brandable ones. Maybe I've picked up some old domain names that I've bought, or I've bought entire properties outright. So what I'm going to do is I'm just going to add a site-wide link or many links from these pages all back over to my main site."

What you're hoping is that this will amplify your ranking signals and amplify your opportunity. The opposite is true. In fact, what's happening is you're creating a barrier for the full link equity for brand, user and usage data signals, and any potential social signals. You're creating a barrier that's stopping some of those things from passing fully here.

Let's just imagine that you've got four links over here, and they are all pointing to mysubsite1.com, which you're then thinking, "That's great. That's exactly what I wanted to have happen, and now mysubsite1 is pointing to my main site." You're actually losing most of the link equity, the value, the ranking power that would be passed if only two or three of these links had linked directly over to your main site.

As we talked about with the subdomain/subfolder issue, by collecting all of the ranking signals on one sub and root domain, you create the best possible benefit. This concept of domain authority -- I don't necessarily mean the number in Moz's Mozscape Index -- the concept of domain authority is that basically as a domain becomes more popular, as it inherits all of these ranking signals, could come from links, from visibility, from branding, user and usage data, all the kinds of signals that a domain inherits, it passes those on to all of its different pages. But it doesn't pass them on to other sites.

That's true for each of these as well. They're inheriting signals that they're not fully passing on here. I'd recommend that you 301 redirect all of these and have one main site. It simplifies a bunch of your work and streamlines it. It lets you focus on building this one brand. Branding is so powerful today and online visibility as a whole, not to mention SEO, that this is really a best practice.

I would get rid of those subsites as best you can. There are still reasons sometimes to have a microsite or a different website for branding purposes, or if you're going to sell that site separately, or if it's a completely different team working on it. But from an SEO perspective, everything on one sub and root is really ideal.

All right everyone, look forward to the comments, and we will see you again next week for another edition of Whiteboard Friday. Take care.