Making the rounds this week: IO Error’s critique of “nofollow”, the link add-on that was supposed to stop comment spam, but hasn’t even slowed it down. He suggests that it was never intended to: it was simply a sneaky way to lower blogs’ rankings in search engines.
Now, I don’t have a problem with the idea of being able to tack a note onto a link and tell Google not to treat it as an endorsement. But the way it’s been implemented in most blog software—blindly applying it to all links in comments—is overkill. Legitimate, useful, interesting comments should factor into the resource’s ranking, for all the reasons IO Error provides.
Where rel="nofollow" does help with comments is in devaluing the spam that slips through your filters overnight. If a search engine bot happens to crawl your page between the time the comment hits and the time you see it and remove it, nofollow will at least make the spam less effective. Of course, you only need it active for a day or two (depending on how often you check). Once you’ve cleaned the junk out, you want what’s left to get the rank boost it deserves. I’ve been using the No Nofollow plugin to do this since March.
In short, I don’t think that rel="nofollow" is a bad idea in itself. It’s just being used the wrong way.
When nofollow was announced, we added it to MediaWiki and set it to include it on all external links in wiki pages by default. I didn’t expect it to kill blogspamming and wikispamming overnight, and of course it hasn’t; there are a *lot* of sites still running older versions without nofollow support and it’s still worth a spammer’s while to run automated mass attacks.
As a website app developer, I do feel responsible for establishing conservative defaults. There are a lot of sites out there that are set up and then poorly maintained, or are hardly touched at all except by roving spambots. Having the default case be to keep the linkspam out of the search engines is, to me, a better choice for the public good than not doing so.
For sites that are groomed and maintained, where linkspam will actually *get* removed, then it’s less important. It can be disabled, or a “smart” aging/removal of the tags would be useful where implemented. But for linkspam on half-abandoned sites that will sit there until the ISP goes out of business, an automatic aging system will just put the spam right back into circulation. In my eye, that’s the most important target point for the use of rel=”nofollow”.
Good point. I wonder how many Geocities guestbooks from 1996 are full of casino ads these days.
I guess that’s the key issue: Is the site maintained? If it is, rel="nofollow" isn’t needed to deal with spam, except on a temporary basis. If it isn’t, it solves half the problem. (All of the problem, if curious people don’t visit the site and click on the nofollowed links.)
Maybe a quarantine with mass-editing and whitelisting capabilities? Pull up a list of new links once a week and mark them all? Require user input, but make it easy. Just brainstorming here.
I’m in general against nofollow but I think it will also help bloggers who don’t know how to fight comment spam any other way. Another side benefit to bloggers is that they can use the tag selectively to keep certain links they use from getting pagerank. Nofollow maybe will work as long term strategy but as now i’m happy with captcha and other kind of plugins
Couldn’t agree more: if the blog owner keeps the house in good order then nofollow shouldn’t be needed. However, the temporary usage of it as a default for blog comments that are awaiting moderation by a site owner might be useful though. When the comment is approved then the attribute can be removed; it could be thought of as a kind of an automatic gatekeeper for blog owners who don’t pay as much attention to their sites as they’d either like to or are able to. So saying, I won’t be using nofollow myself; I’m intending on keeping close tabs on my blog when I eventually get it live…