This test is aimed at determining whether using relative links will reduce the power of link juice flowing to the target page.
Test Downloads
Transcription Below
A relative link vs and absolute link. An absolute link is a www.mysite.com because you are showing exactly the location of that link. A relative link is just when you do a backslash (/) on a site meaning, the homepage. That is a relative link. You are not putting in the whole part – the www, the domain, the .com – you are only putting the part just to identify within that site.
There were some discussion whether relative links were bad for SEO and that they would harm your page. This test was done by Ryan Hough. The idea was that you’ve got two links going to internal links - link passing to a relative url and an absolute url.
The idea here is when you are setting up your internal linking structure, are you going to be passing less juice when using relative URLs? If so, the page with the absolute link would jump it.
As these pages began to rise through the results, the relative url never jumped. The absolute url is always below. What this is telling us is that there is no difference between the juice between a relative link and an absolute link. What is interesting is, in a completely separate thing, if you are using the relative url, it is entirely possible that you could overcook your internal anchor text.
In the situation where you’ve got an EMD (exact match domain) and PMD (partial match domain) and you’re pointing your links internally, you are using that keyword again and again and again, you could potentially overcook your anchor text that you are using simply by using that emd over and over. The relative url would be a way to get around that.
In your internal linking, you would be using the relative url that does not contain the domain. It is something to think about. If you have a page that has a ton of links coming in internally and you are seeing weird performance – it is not moving as it should and it is behaving in a situation where you might have too many links coming in that are exact match anchors.
The relative link would be something to consider and changing some of those out to see if you can get some kind of movement up. Otherwise, I would not sweat it too much if you have an old school site that is all relative urls and you need to turn around and change these to absolutes. You are fine as you are.
The Moz article that was cited is worth reading because I have a couple of different reasons for not using relative urls but overall, they didn’t really hit me as anything that I should stress about.
Dori: That is really interesting though just in the fact that you have a client that you think might be hit with a penalty like an overcooked keyword type, to go in and take out that exact match and change them into relative. I never really even thought about that. That is pretty advanced stuff.
It’s the kind of thing where you are looking at a page and you think you are not overoptimized – you are looking at your optimization vs others and you are not spending any spammy links in, have a good profile, but it is behaving as though it is under a penalty, look at your internal link structure – in the situation where you’ve got an EMD, especially, and see how that is set up. If you do not have an EMD, this isn’t anything that you need to worry about.
I think that is also what might be in the situation where people think EMDs do not work, it might be that EMDs with good internal linking structure are harming themselves because of their links structure. It is not that EMDs don’t work, it is that a large EMD site could be a problem because of all its internal links.
Dori: Wow. I never would have thought of that. You are right, that could be exactly what is going on with people that say EMDs don’t work anymore because we know that they do. They could be overcooking themselves.