Warning: Nofollow should not be used to prevent pages from being indexed. Google does not pass Page Rank but does crawl the nofollow-ed pages as part of its autodiscovery algorythm! To block pages use robots.txt rules and noindex in meta robots tag.
Apart from being a tag nofollow is a rule. Nofollow basicly tells the search engine crawlers: Yo, search engine crawler dude man! That link, I don't guarantee for it. I have it here but I'm not gonna vote for it! Automagically a sane person would ask himself: Why would a link that is present on your page, which you maintain, is not to be trusted?
This is the paradox in nofollow. You link to an untrusted source and you may send you visitors to to site but you don't guarantee it's relevant or even safe! So search engines allow people to send others anywhere but not take any reponsability. One should answer for the links on his site!
Some may say nofollow does not express confidence and only keeps page rank from going to that page. And what is page rank? It's google's algorythm to decide how much to value a page based on links it gets. It's a unit of measuring confidence throughout the web.
Nofollow is an evil thing that came out of the best intentions. As throughout history many horrors occured with greatest and best intentions. Everyone meant well but the outcome was not the one expected. Same thing with nofollow!
Nofollow tag was protection against comment spam. So blogs would automatically add nofollow to links in new comments and trackback so the spammers would hit in vain. Actually this fix did nothing! The main concern were not the links but the blogger's bandwidth consumed by spamming.
If you had the bad luck to reach a list of a link spammer you would have gotten hit even if you had triple nofollows on the new links and javascript redirection. Trust me! I never cared about nofollows. No one did. Lists of 10.000s of links could not be maintained. So you just hit'em all and pray for results.
Here is the real problem. By discarding any responsability about links people were allowed to link anywhere as long as they slapped a nofollow attribute on the links. On the other side people added nofollow on all the links so no site voted for each other. I think the flow of link juice has been reduced by at least 75% with this (actually more to 90%).
This is how nofollow looks like in a link:
<a href="http://www.i-dont-trust-this-site.com/" rel="nofollow">untrusted link</a>
and this is how it looks as a META tag:
<meta name="robots" content="nofollow" />
No follow has a simple effect. Prevents vote between pages hence locks the link juice flow hence does not pass page rank! (hence I used a lot of hence :)) The two implementations above have different functionality. The one in the Link only prevents from passing PR to that very link, the one in the META prevents from passing PR to all link on the page both on or off site and both having or not having a nofollow specified. So Link nofollow blocks links and META nofollow blocks all links whether they were or were not previously block at Link level on that page.
The real way to use it is for links within your site that don't need to rank: privacy policy page, disclaimer, shopping carts and so on.
Actually I mask outgoing links completely and block them with Robots.txt. So nofollow is a pat on the shoulder compared to the overkill I do to my links. Maybe when I'll grow up I'll share some link juice. For now this site has not much to share!
Post Feedback