Übercontrarian Silliness on Comment Spam
Ben Hammersley (h/t: VK) and Ubernostrum (h/t: LW) offer two mutually inconsistent arguments against using the “new” nofollow tags, both of which lead to the conclusion that widespread use of the nofollow tag will lead to more comment spam rather than less. Both theories can’t be right, of course, but both
First, let’s dispose of the sillier of the two theories, Ben Hammersley’s. Hammersley argues that as the use of nofollow increases, comment spam aimed at boosting page rankings will increase to compensate. The underlying fallacy, and one shared by Ubernostrum, is that he assume comment spammers are currently pulling punches to avoid spamming too many sites or to avoid boosting their Google rankings too much – and can “turn up the volume” beyond their current spammings anytime they like, at no net cost to them. I say, bollocks. Comment spammers are already spreading their spam as far and wide as they can. They’re not waiting around for someone to make comment spamming less useful, only to spam more to reach some sort of equilibrium. Hammersley utterly flunks Economics 101 with this paragraph, that pretty well sums up his theory:
So as comment spam costs absolutely nothing to spread, there’s no loss to me if I spam sites with rel=”nofollow”. It might be, at a pinch, less efficient a method of getting readership, but it will still work: and efficiency is not a factor at all. Spamming costs nothing, so spamming sites with rel=”nofollow” doesn’t bother me. I’m not losing. Perhaps I’m not winning as big, but I’m not loosing anything either. There’s no incentive for me to spam those sites for the sake of getting Pagerank, that is true, but there’s even less incentive for me *not*to*. Why would I bother even testing the site for rel=”nofollow”? I might as well just hit it and leave. It’s less work for me, for exactly the same gain (some) and exactly the same loss (none).
Got that? Comment spam isn’t just cheap; it “costs absolutely nothing” to spread. Not only that, it costs absolutely nothinger to spread it even further than they’re doing now. So why aren’t they doing it now? Also note Hammersley’s new-new math: not only does nothing equal nothing, but “some” always equals “some,” too, even when one “some” represents a small fraction of the other. O-kay.
Ubernostrum’s theory is somewhat more sophisticated than Hammersley’s, but it’s still off base. Ubernostrum concedes that of course the widespread use of nofollow will reduce the type of comment spam aimed at boosting page rankings, but argues that comment spammers will go on spamming anyway for a different purpose: to actually be read by readers. Indeed, there may be some substitution effect, but not much. Anyone who’s dealt with comment spam for an extended period of time knows that most of it is not meant to be read, or even meant to be left up long enough to be read, only to remain long enough to be “read” by a Google or Yahoo! spider. Comment spam intended to actually be read will have to be written much differently, with fewer links and a writing style that makes it look to a human as though it were part of the original discussion. This kind of comment spam is much harder to devise, more costly, and most importantly, is not a viable substitute for the kind of comment spam that has been hitting us in waves.
I have no doubt that eventually, comment spammers will find some way to render nofollow ineffective. Until they do, nofollow is a nobrainer. Unless you specifically intend to allow your commenters to boost their Google rankings or those of other linked sites merely by commenting (as Kate and Lonewacko do), there is no earthly reason not to install the nofollow option.
UPDATE: Ubernostrum responds.
January 22nd, 2005 at 10:48 pm
Comment spam and the “no follow” option
XRLQ, the Mexican underground radio station that plays Captain and Tennille tunes twenty four hours a day in order to overthrow the Mongolian government, has brought up an interesting subject, at least to some bloggers.
The post has to do with comme…
January 23rd, 2005 at 9:44 pm
I get a few thousand spam emails a day. No, really. There’s no google involved in those spams, they’re simply counting on a certain small percentage of people who’ll click a link or install a virus or whatever.
I have a more thorough post on nofollow.
January 23rd, 2005 at 10:12 pm
No one disputes that some spam is designed to be read rather than to boost page rankings. That’s a given. It’s also completely beside the point.
January 24th, 2005 at 7:43 am
If they have bots posting, I can imagine it may be more of an effort to stop these bots than to keep them going.
However, it would be an effort to update the bots if posting systems change. Perhaps that effort is justified.
January 24th, 2005 at 1:32 pm
You should post this at kuro5hin. If ubernostrum appears in irc while i’m not quiescent, i’ll tell him to read this post, but there’s no guarantee that will happen …
January 24th, 2005 at 4:13 pm
My reason for assuming that comment spammers will “fall back” to postign for eyeballs is the simple fact that they’ve invested time and money in spamming as a marketing technique, and I don’t expect them to just say “aw, shucks” and give up. Spam which tries to get read and clicked on has been successful in email (if it weren’t successful, the spammers would eventually stop or go out of business), and there spammers have to spend a good bit of time and effort devising ways around blacklists and filters. Keep in mind that the profitability point for spam is ludicrously low; the numbers I’ve seen generally claim that a click-through rate of anywhere from 0.4% to 0.6% will make money, and I think they could probably manage that rate with comment spam if they wanted to.
As to whether spammers are currently “holding back”, I’m still somewhat undecided but leaning in that direction. It’s true that there’s an awful lot of comment spam out there, but it’s a drop in the bucket compared to the volume we see in email and on USENET. Now I know for a fact that comment spammers have access to mass-posting software and other tools just like their brethren in other media, so why is that? I wonder sometimes if what we’ve seen up until now wasn’t just testing the waters; after all, certain weblogging tools made it ridiculously easy to ride the comment-spam gravy train for a very long time, so all a spammer needed was the occasional burst of posts to high-PageRank weblogs. And that was good business sense from a spammer’s point of view; if you don’t have to make twenty million posts to turn a profit, don’t. Now that we’re closing some of the loopholes I think we’ll see the firehoses get turned on in earnest.
January 24th, 2005 at 4:56 pm
Sunk costs may influence irrational comment spammers in various ways, but they won’t influence the rational ones who, uncoincidentally, are the ones most likely to be successful in spreading their spam to all quarters of the Internet. Currently, there are almost no controls on email and Usenet, so guess what? That’s where most of the spam is. Web sites could host spam intended to be read, but they don’t provide any real advantages for that. Boosting page rankings is the one thing web sites can do for spammers that email and Usenet can’t.
More generally, look at the big picture. Nofollow takes away most of the benefit of one type of spam (the kind intended to boost page rankings), while creating no new benefits for the other (the kind intended to be read). From that, it’s axiomatic that the net impact on spam will be negative. If comment spam increases over the next few months (and I have little doubt that it will), it will be in spite of developments like nofollow, not because of them.
January 24th, 2005 at 5:49 pm
I think your analysis is off there; it takes away the benefit of one type of comment spam, sure, but in so doing it makes the other much more attractive to anyone who’s already committed to spamming as a strategy. The reasoning here is that one type of comment spam requires PageRank benefits to function, and the other doesn’t. Take away the PageRank benefits and it’s easy to see what will happen. Thus the net impact, insofar as it can be predicted, is an increase in the second type of spam.
January 24th, 2005 at 6:35 pm
That doesn’t make sense. No one is “committed to spamming” as an end in itself. Spammers spam to get something in return. Some spam one way to boost page rankings of certain web sites, while others spam a different way with the intent that their messages actually be read by real humans, and others still (most, I suspect) resort to both. Mounting a frontal attack on one kind of spam may or may not succeed against that type of spam, but it won’t make the other kind any more (or less) useful to the potential spammer. It won’t affect it at all.
What will affect it, of course, is better spamming technology. That’s almost certain to happen no matter what the rest of us do. The only question is whether we’re going to have better technology to fight back. The nofollow attribute is an important part of it.
January 24th, 2005 at 6:46 pm
It makes a lot of sense. Let’s say you’ve invested in some mass-posting scripts for different weblogging tools and forums, and you’ve gotten yourself a list of places where you can host and run these scripts. Then you find out you’re not going to get PageRank benefits from all those comments you’re spewing. Do you:
A) Give up, or
B) Realize that some benefit is better than none, and raise the volume of your posting to take advantage of the inevitable few tenths of a percent click-through?
I think the above is a no-brainer for anybody who’s already spamming.
January 24th, 2005 at 7:16 pm
James, what they’ll do is what marketers have done for centuries when an established method of marketing goes away. They’ll move on and try something else. When they hit upon something that works better for them than comment spamming in a “no_follow” blogosphere, they’ll concentrate resources on that.
They know perfectly well that with technology progressing constantly, anything they invest in will be good for only a short while (though comment-spamming and trackback spamming have been good for way too long already). It’s written into their cost-benefit analysis.
“I think the above is a no-brainer for anybody who’s already spamming.”
Just one problem: spammers do have brains.
January 24th, 2005 at 8:13 pm
Technology may be constantly progressing, but spam still infests our lives on a daily basis. It isn’t going to go away, and the spammers know that; they may switch to different modes of spamming or they may start spamming for a different perceived benefit, but they’re going to keep on spamming and nofollow isn’t going to stop that.
January 25th, 2005 at 3:33 pm
Oh, yeah, did I mention I am an idiot? Dddddeeeeeerrrrrrrrrrr…