Wikipedia has decided to nofollow all external links to help offset people spamming the service. In theory this should work perfectly, but in practice although all major blogging tools did this two years ago and comment and trackback spam is still 100 times worse now. In hindsight, I don’t think nofollow had much of an effect, though I’m still glad we tried it.
The solution is not to have external links. Seriously. When I finish the redesign of my sites, comments will be checked for spelling, grammar, length, complexity, and will not include author links. Any failure to comply will result in a delicious error message.
Well, I think it’s partly because spammers are thick and don’t realize that there are nofollows there. But they’re also probably betting on real clicks instead of search engine crawlers.
Either way, if Wikipedia really wants to fight spam they should find a way to integrate Akismet 😉 Works like a charm for me!
Depends on what the effect is you’re looking for. It may not have reduced spam, but did it prevent spam from taining SERPs?
The sad thing about the no-follow is that spammers don’t care, but legitimate people gets hurt.
But thank god for inventing Aksimet, blogging would be impossible without it!
They’ve actually been doing this for a while on other namespaces, like the Wikipedia: or User: namespace, but yeah — this is bigger.
Joost, that’s a very good point. If it did, then it would be worth it.
As a follow on from Joost’s comment, the only way to avoid being spammed, even if we’re looking at SERP, is to effectively have a PageRank of 0, no hits when a keyword search is run and or no method available for comments, trackbacks, pings etc from being made.
Given that’s not particularly useful to, well, anyone and given spam tends to be entirely indifferent as to whether content has a no-follow relationship or not, I can’t help think that this is an entirely odd move for Wikipedia – the ship sailed two years ago.
No-follow – a solution looking for a problem. 🙂
I’ve just hacked up a quick plugin that adds rel=”nofollow” to wikipedia.org links…
http://whatjapanthinks.com/wikipedia-nofollow/
Joost hit the nail on the head. Nofollow was never really about reducing spam on blogs, it was all about helping Google protect their indexes from pollution by blog spam.
Remember where nofollow came from… Not from the bloggers but from Google… Somehow bloggers got the notion that spam would decrease theough use of nofollow and Google didn’t attempt to dissuage anyone from this notion, widespread adoption of nofollow was going to benefit them first and foremost.
So here we are, we meaning bloggers, 2 years later, right where we were. The volume of comment/trackback spam has not been affected at all, and we were being naive if we thought it would…
http://www.johnkeegan.org/blog/_archives/2005/8/23/1161666.html
Forget nofollow!
Bloggers [heart] Akismet. 🙂
Ken, your plugin is a delicious zinger. I like. 😀
(p.s., Matt, I can read teh red text in the submit button below, Fx2)
As follow to Brendan’s comment on Joost’s comment:
It’s not about avoiding spam (or direct clicks). It’s about making spamming less of a blight on search engines. That is what it’s for. Nofollow might not help you avoid spam as a blogger (that’s what Akismet is for), but it helps me every single time I use Google. Which is often.
Matt – bravest thing I have seen said for a long time and have just blogged the same
Can we have a switch in 2.2 to turn nofollow off, rather than use plugins
Ken – I was just about to cobble a plugin together – I know a few people who want one – incoming link juice
NoFollow is a ridiculous concept that just refuses to go away. The more high profile projects buy into this, the longer it’s going to take to finally die out. As far as I’m concerned, Google is doing the web a great disservice spreading this idea.
On one of my sites, all content is blocked until an admin approves it. This STILL doesn’t stop spammers from trying to post the most idiotic and spammy links imaginable. The problem is that all this stuff is automated by armies of bots who don’t care about nofollow links and approval. Monkey sees form, monkey fills in form with spam and submits.
What would work for Wikipedia is whitelisting. External links have to be vetted by trusted moderators.
Arguably less work than removing spam from articles 🙂
Over time they could build up a blacklist as well to reduce the effort in whitelisting.
http://engtech.wordpress.com/2007/01/23/wikipedia-loses-the-google-juice/
Actually, I think implementing the NoFollow tag was hugely important. If you hadn’t done it, there would have been a whole other level of social engineering, inundating us with on-topic comment spam that would be very hard to separate out.
Perhaps I should add that I spelled out my case in some more detail at http://www.monashreport.com/2007/01/23/nofollow-does-matter-a-lot/
(the reliability of my trackback functionality is a disaster …)
In the process we loose all the linkage (forget pagerank for the moment). Say your wikipedia profile no longer says if you look for me on wikipedia use the about page on my website. What may be useful is holding back pagerank but not reducing the mapping of resources?
if Wikipedia really wants to implement a rule that will prevent spamming, I’d suggest they make a script that will automatically change the link from http://www.google.com to http://en.wikipedia.org/Google/go_http://www.google.com
something that will not reflect as a true link. if thats possible.
So this is the day when I’ll go back to using rel=”nofollow”.
For Wikipedia links only, of course.
I have actually been beginning to think that blogging tools should turn start turning nofollow off for trusted commenters and users. Those people who contribute good content to a blog really should be rewarded with a link that search engines will recognize. People who comment a lot all over the place deserve that recognition.