I have been experimenting recently with web site wide links to see if the horror stories are correct, whether they maybe help or whether they are only a waste of space.
So, I picked a random phrase on a PR0 weblog and searched Google on this term. The term was actually a claim sentence for a paid blogging system, so it us used on a load of websites, but none of them are maybe optimised for it. At first, my PR0 site was 35th on Google for it. Then, I linked from the home page of a PR3 site to the post page with this phrase and watched the results.
Once the PR3 web site was visited by Google, it itself appeared on page 2 of Google's results, about about position 13. A day or two later and the PR0 site jumped to 15th and then slowly up to 10th position, overtaking the PR3 web site.
Now, I cannot explain why it continued to climb up after the initial jump, but it did. But the effect of the single link is not what I am trying to look out.
I left the web sites alone for a week whilst I monitored the movements, watching them balance out. Then, I put the same link onto the rest of the PR3 website in exactly the same position.
Now suddenly thousands of pages of the web site had a link to the PR0 blog, using the random phrase as an anchor text. This really was a web site wide link if ever there was one. Not much happened, for all of 2 days! Then suddenly the position of the PR0 site started to crash off.
Within three days the web site was down to 39th on the search engine listings, even if just 2 dozen of the pages on the PR3 site were reported on Google as containing the link.
So, this points to some things.
First, by introducing a web site wide link the effects of the single link are knocked out. A single link on 1 page could give a benefit, but put the link across the website and there are no benefits.
What about it dropping to 39th, from the original 35th? Does this mean a website wide link is damaging? I think not. Out of those web-sites above it in the listings, 2 of them, are the pages from the PR3 website, plus a couple of new high ranking sites have also displayed the claim sentence. So that explains why it fell further than it climbed.
The second point to notice is that a website wide link is anything but a link on every page. Only around 23 pages are cached with the link out of some thousand pages across that web site. So, it looks as though if you get more than about 5 - 10 links from any one website in a short time period then the links are totally ignored.
This means that if you are writing WordPress templates in the hope of getting plenty of links in, you are perhaps wasting your time! The same for plenty of other tricks. Whether further site wide links when they are built up in excess of time on an older site have the same effect, I cannot say. That experiment will take plenty longer!