Yes, this is yet another post about Google’s Preventing comment spam. So Google have run a little level-of-optimism test three days ago. The results?
Well, the good news is — there are optimists.
The bad news is — it will have zero (like this — 0) effect on the amount of comment spam. Or multiply zero by three, if you count MSN Search and Yahoo! too.
I think the title of the Google post has a little bit bitten off it, full version might sound like, say, “Preventing comment spam from influencing PageRank”. Because this is that rel="nofollow" attribute does. It does not catch spammers and boil them in the hot oil, nor does it put them into a jail, neither does it make their web servers crash and burn — it simply does not take some link into account when calculating PageRank.
Yes this is great, because now people googling for “cheap Viagra” or “enlarge your johnson” will get results not skewed (hopefully) by comment spam. Good for them. How often do you google for stuff like that? Good for you too.
The fly in the ointment is, that bloggers seem to overestimate importance of PR for spammers. More than a year ago Movable Type and Blogger offered redirect as a way to fight comment spam. Did that help? You wish…
Links in e-mail spam I get are never counted for PageRank. Despite server-side spam filters, client-side spam filters, I still get spam e-mails. Why is that? It is because spam works this way — distribute links as wide as you can, hoping that some will click them and some of those clicking will be tempted to try some cheep-stuff-long-penis product. There is extremely small percentage of the those who will, but — huge numbers of links distributed multiplied by some tiny percentage of interested still are very well worth spamming.
Ok, this move might lower PR for spammer sites. So it may affect SERPs and spammers will get a bit less traffic from the search engines. So what? They will still have links on popular blogs read by many people. And if there is a link one can click it, right? And “solutions” involving redirects make this even easier — I will not click on the link “online-poker-bla-bla.com” , but if there is some redirection involved, I will not know where does the link lead, until I click it.
The sad point of this is that “solutions” like these are not solutions at all, they just witness another victory for spammers.
You start crippling and hiding your e-mail address — you are defeated; you cripple your comments system making it more difficult to use for normal people, make real links invisible via redirects, force people to solve graphical puzzles, make links in their comments weightless by adding some fast-baked attribute — you are defeated again.
Spammers will try to inject their links everywhere they can (referrer spam, anyone?) and while they succeed on this we may
add as many attributes as we can, that is not a solution.
And I don’t know a solution yet. Do you?
Rimantas,
for what problem do you want a solution?
a) make your blog spam free;
b) do (a) effortlessly;
c) do (b) for every blog site;
d) achieve (c) once and forever;
e) achieve (d) for every form of communication, not just blog.
Also, would you consider solving a problem 90% victory or defeat?
If you want (e) and 100%+, then I’m afraid there is no silver bulet. But if you think that every step which makes spammers extra problem without harm to the good guy, it’s worth doing, then read on I have few [teoretical] thought.
First, (a) is achievable by allowing only proofread and approved comments. I don’t know is that acceptable according to moral codex of blogger. But goal is achievable 100%. If 90% solution is acceptable, then maybe you can let your readers to help your other readers to avoid visiting spammers by implementing some vote system.
Granted, this requires resources to implement. Can you automate that? Of course, it requires testing, but how about checking comment and contents of the links in it against SA’s database which you use for e-mail? After some training I think good portion of spam would be filtered out. Remaining could be dealt using manual methods.
Goals (c)-(e) are somethere beyound horizon, but to aproach to them I don’t see any other way than methodically cutting opportunities for spammers and lowering their revenue:expences rate.
And if all your post is about yet another case where marketoids won over techies, then I’m saddened by that fact no less than you. Do you have solution for this problem? :-)
Three rules for the spam game:
1) you can not win.
2) you can not draw.
3) you can not leave the play.
Greetings,
Antonio, from Malaga (Spain)