It’s a well known fact in link building circles that Links in Old content simply aren’t as good as Links in new Content.
Taking some inspiration from a recent Whiteboard Friday I decided to test this theory. Cyrus Shepard, from SEOMoz, went through the theory that a link to your website from some old content does not pass as much “link juice” as a link from a new page; you can see the video below.
What do we mean by an “old page” when we talk about these old pages? From a technical, Google definition point of view, we’re talking about something that has been previously crawled and indexed by Google. Stale content, by stale we mean content that hasn’t been updated in a long time. It was written and it just stayed that way. There are no new blog comments. It has just been for two or three years the same way it was written. And old links. So this old page, all of the links that it got, it got years ago or months ago, and there are no new links coming in. That’s what we’re talking about when we talk about an old page. If it doesn’t meet these definitions, then it’s a new page.
Tim Grice, SEOWizz, did a study in March 2011 showing that links on Old Pages just weren’t worth it. Over a 5 week period Tim monitored the changes in search rankings by adding site wide text links in sidebars or footers, links inserted on indexed static pages with a PR 1 or more and finally he inserted links within entirely new content in fresh blog posts.
As you can clearly see the rankings for the “old content links” barely changed at all over the period where as the links within the new content rose quickly.
This got me thinking, especially the statement from Cyrus, that in order for Google to consider the page as new again you would need to make a significant change to it or build some links to the old content. But exactly how much of a change would you need to make to a page?
I decided to build links to some of my test sites using the same principles that Tim used, the old pages were already indexed by Google & they had not had any new links built to them recently.
- Link type 1 was my control this is a link where I only added the link into the text with exact match anchor text.
- Link type 2 I inserted the link and inserted 1 paragraph within the content
- Link type 3 I inserted the link and added two paragraphs
- Link type 4 I inserted the link and did 5 social bookmarks to the old page
No surprises that the link only and link plus 1 paragraph saw very little change in rankings but after seeing the great performance of both link type 4 and 3 so early on in the experiment I decided to edit two paragraphs of text and do 5 social bookmarks with another test page.
So not only did the page climb the rankings rapidly it has stabilised its new ranking; better than the social bookmarks only pages.
By no means was this a completely controlled and perfect scientific experiment as there was a new Panda update during the period of the test as well as the fact that the the content in the test pages weren’t all exactly the same. But as you can clearly see just by simply adding a link to piece of old content and editing just a small amount of text on an old page it has less value than adding a link, editing some text on the page and building a few links to the old page.
This will flag to the spiders that this page is now relevant and to recrawl the page. Which in turn means that the bots will follow the links within that text once again. In an ideal world it is preferable to build links within new content and shows how important it is to continue with content based link building methods such as using blogger outreach, guest blogging but if you are building links to old content i.e. broken link building it’s worthwhile taking the time to add some more value to the old pages.