Of interest, I’ve have had a growing suspicion that new link-votes that are suddenly inserted into older articles have less value than if the link-vote uploaded fresh with a new article.
In a recent test I did I placed a link-vote in an older article that had an existing PageRank value of “3.” This link-vote was targeting a high-level key-phrase and I pointed it toward a non-optimized related website and the website moved after 3 weeks from not even ranking in the top 1000 to ranking at the 91st position.
After waiting a few weeks I decided to delete the link-vote and within 3 days the website vanished from the top 1000 results.
From there, I waited another week and then I placed a link-vote (on the same domain) on a freshly written article and simultaneous uploaded them both at the same time. What I saw after that was rather surprising as the same website suddenly re-appeared at the 41st position the very day it was first crawled.
What this told me was two things:
1) The obvious: the fresh link-vote placed on a fresh article on the same old domain influenced the search engines more so than the link-vote placed on the older article that had an existing PageRank of “3.”
2) While it is common knowledge that link-votes take about 12 to 21 days to be realized in the search engines, the second link-vote that I placed on the same domain was realized the very day that Google first crawled the fresh article.
Any thoughts?
Posts
3 comments:
Interesting study. I'd be curious to see if the same behaviour is observed consistently, or if there were other factors involving the strength of the linking pages. One test isn't conclusive, but it's food for thought.
Ever since Google's favourite paid links whipping boy TLA brought InLinks out of private testing and into public release, we had to know Google was was going to try to filter out that type of link.
A couple other questions about your test:
1) Did you add text or just create the link from the existing page copy? (More on this when I know the answer.)
2) Was the anchor text for the new links relevant to the content of the old & new articles?
Hi Ryan,
I developed three new sentences within the existing paragraph and placed the link-vote within it. The article was about 5000 words long and occurred near the middle of the article. While the 3 new sentences where directly related to the link-vote the greater portion of the article was only vaguely related.
Regarding the new link-vote placed on the new article, the entire article was focused on the history of the particular topic and the link-vote occurred 3 quarters of the way down on a 1000 word essay.
So there may be two things going on: 1) the new article was completely related from head to toe. 2) the article was freshly posted.
Your thoughts?
There is a theory that the PR score of a page is divided across keywords or phrases representing the theme of the page. That theory holds that a combination of the strength of the key phrase on the page and the strength of the incoming links to that page relative to the text of the outgoing link impact the strength.
That explanation works much better when I draw it on a napkin.
I think freshness does matter some - it's only natural that Google would want to be wary of InLinks type link selling. I also think Google needs to be careful not to punish sites who are trying to improve their internal linking structures by linking or siloing a site based on existing content.
I guess we just need to test more!
Post a Comment