E-Mail:

Ten Suggestions for Google's Blogspot

Emotion. Raw emotion.
That's what caused me to write something about the Blogspot spam problem this weekend - independently from Jarvis writing independently from Haughey writing independently from Cuban writing independently from Bray.
Google has responded with a reasonable explanation for what's happening on their end, although their course of action seems to be a bit mysterious. I'm noticing phrases like “one thing we can do” and “we can also make it more difficult.” Can or will? Regardless, I think they're gonna be making some changes soon:

Just as a first step, we're publishing a list of deleted subdomains that were created this weekend during the spamalanche.

And in a recent IM exchange with Jason Shellen:

I spoke to the team today and they have some good ideas as well as two or three initiatives in place that should go live soon.

Today, it's the egofeeds that suffer - but what about tomorrow? I realize that CAPTCHA isn't the perfect solution, but there might be some other options on the table:

  1. Employ a blog spammer. Beat them shitless once they've been hired (hazing spammers isn't illegal), but then employ them to help you figure out what all those asshats are doing. After they've told you the secrets, beat them shitless again.
  2. Probationary Period. Only allow new users to create a limited amount of blogs. Say, only one for the first three months. Then, if that goes well, let 'em create six. Then, if THAT goes well, let 'em create six more. Offer carrots. “You've had your blog for a year. Happy birthday! You now have XYZ access!” Oh, isn't that nice. Slap rel=”nofollow” on all outbound links for the first X months. I realize rel=”nofollow” isn't perfect, and it's not going to stop people from doing what they're likely to do anyway. However, give spammers one less reason to come to you to do their dirty deeds.
  3. Sponsor a Blogger. The “new” way to get a Blogger account is to be referred by someone else. Then, if the invited one screws up and starts spamming the blogosphere with bunk or copied content, revoke both accounts - the inviter and the invitee.
  4. New banner button. You're showing me ways to get my own blog at the top of every single Blogspot page. However, since I don't have the Google toolbar installed *GASP* - I can't report spammers! What's stopping you from putting a “Flag” button in the banner, too? Let anybody report the problem, easily and quickly. NEVERMIND, I just saw it appear in the toolbar - and may have missed it before! Just flagged 10 out of 20 I visited… blatant problem children. Consider this point a note to all you do-gooders.
  5. Take every experience seriously. Yesterday was not the first time anybody reported on the Blogspot honeypot problem. Even with Dave Sifry reporting that only a small portion of today's blog are splogs, how will the ratio sit in a few months? I can tell you, I get far more email spam than I do legit emails these days - and I get a lot of emails every day.
  6. Cross reference your databases. I know you're tracking this stuff effectively in Google. You probably know which URLs are spam and which aren't. You likely know which ones are link farms with self referential links and which aren't. When a new Blogspot blog points to something that Google already knows is a “bad seed,” automatically suspend that Blogspot account. You're Google, man…
  7. Reward flaggers. Send 'em a t-shirt for every X thousand VALID splogs they flag. Give folks an incentive, and… it seems that everybody goes ga-ga for a t-shirt. Seriously. Even let 'em parade their VALID numbers somewhere on their blog. “I've killed X thousand splogs to date… and all I got was this lousy t-shirt.” But it's NOT lousy - it's a good idea.
  8. Audit randomly. Say, a few times a year, you do some kind of “Hey, we just wanted to make sure you're enjoying our service” message interruption when someone tries to access their account. It's a random hoop to jump through, and if you do it right, the true users won't mind a bit. Flag the account so that they have to answer a question to regain access to their account, via the Web or otherwise. XML-RPC folks will be forced to log in and answer a random (non-invasive) query.
  9. Flag “hot” keywords. If you don't know what those might be, talk to your buddies over in the AdWords department. Any kind of pr0n subdomain references should be instantly earmarked for review as well.
  10. No more dashes. Seems a good chunk of splogs have dashes in the subdomain. Eliminate that as a possibility. Seems that most “real” users don't care about using dashes.

Dunno. Just some suggestions from someone who truly cares…

14 Comments

Not sure about the dashes one unless they implement catergories. I uses dashes in some of my blogspot blogs as I am trying to fake catergories (badly mind you).
Example:
Main blog: mollyzine.blogspot.com
Podcasting Catergory/Blog: mollyzine-podcasting.blogspot.com
HTH
Molly

Awesome - I'm especially fond of anything that involves giving away t-shirts (and rewarding the flagging of spam blogs).
We have definitely been learning lessons (and, in fact, working directly with) our fellow Googlers who have been fighting spam for quite some time. The ideas about cross-referencing, hot words and sharing tools are ones we are already implementing.
For what it's worth, we're seeing that the spammers are serially creating individual accounts in order to create their blogs … it's not multiple blogs under a single account.
Thanks for all the ideas.
Jason Goldman

Chris, I have mixed feelings about your ideas.
1. Employ a blog spammer
Not sure how realistic this item is, but it sure gave me a chuckle. :)
2. Probationary Period.
Limiting accounts in the way you've suggested doesn't seem realistic. How would Google limit this? E-mail addresses are infinite. Cookies are easily ditched. This item, IMHO, would only negatively impact honest users, not sploggers. Also, your idea of no-follow is, IMHO, completely useless. I believe that most sploggers are creating splogs not for PR, but rather for AdSense earning.
3. Sponsor a Blogger.
Agreed. I've already suggested this in my own Blogspot Anti-splogging Solutions post.
4. New banner button.
I don't think the “report a splogger” tactic is very useful over the long haul. At present, sploggers can create new splogs faster than people can flag 'em and Google can investigate.
5. Take every experience seriously.
6. Cross reference your databases.
7. Reward flaggers.
Duh, sure, and n/a :)
8. Audit randomly.
If the sample is modest, the usefulness is minimal. If the sample is large and frequent, the disturbence is substantial. Not a good idea.
9. Flag “hot” keywords.
Might be somewhat useful. By forcing sploggers to obfuscate keywords, their AdSense revenues would undoubtedly suffer at least a little.
10. No more dashes.
Attacking minor things, not root issues causes and causes isn't worth it.
* * *
My own list, summarized from my post:
1) Charge $1 to establish an *indexed* blogspot account. Still give away free (non-initially-indexed) accounts. As an alternate option, offer free upgrades via mobile-text-message (a la Gmail)
2) Do invites (blogger invite a blogger)
3) Automatically or, upon request, offer to index any free blog that's met a certain threshold (minimum posts, minimum time alive, etc.) This'd allow people without credit cards, without a cell phone, without friends on Blogger, etc. to still get their content eventually indexed.
All of this would be a major pain for Google, likely, and a minor pain for honest bloggers… but I think as a comprehensive plan it'd kill 99% of splogs.

Fair series of comments Chris - and understandably born of both frustration and concern. Unfortunately, this is going to be a problem with ALL “free” blog platforms and I don't really see a universal solution. I guess as long as the cycle is broken SOMEWHERE, the system becomes less attractive to the spammers. What about also implementing spam ID on the directory, search systems - like pubsub , technorati etc? I think it's going to need a multi pronged attack if it is going to be solved. I would agree absolutely that we must act NOW.

7 Sounds good to me… Everyone (well, most everyone) likes showing a counter or two or three on their blogs… and then there's that t-shirt…

None of this will reduce blog SPAM. All that it will do, is remove market share from Google. As such, don't expect them to implement most of it.
-Randy Charles Morin

It scares me a little what will happen by this time next year when RSS might be a tad more mainstream with Vista and Office coming out (even if you're not a windows user you have to admit most people are).
Can you just see how it's gonna be for those NOOBs who go to blogger and set up accounts. And, worse, the NOOB slimey peeps who just wanna earn a buck will splog to the third power. If you think these guys are bad, wait til all your neighbors, your dentist and the man who drives the ice cream truck all have blogs.
I sure hope they figure this out soon. An idea:
Why don't we all just pool our money, hire the splogger, then we can all beat him senseless…

Netcraft has a nice summary article of the splogstorm:
http://news.netcraft.com/archives/ 2005/10/17/ google_draws_fire_over_blogspot_spam_blogs.html

The answer isn't in the blog hosting sites, that's not where the money is. No, it's in the ad content providers. Google Adsense, Yahoo, adbrite, and others are the villians here. They've created a product that is easy to exploit, abuse, and manipulate to make some easy dough. If I were an ad publisher, I'd be very pissed off that someone is allowing this fraud to occur on their network. If Google et al don't start self regulating soon, either gov't will step in or the internet will route around this damage. Their business model will be changed or lost either way.

i thought you'd be happy to know that your words are being heard at google, most definitely. i'm sitting at the zend/php conference listening to a keynote by adam bosworth, vp of engineering at google, and he made a comment about getting “yelled at about spam in blogs”. ;)

I like those a lot, someone has just posted this on Blog Party and I think that there is not much hope for blogger unless Google starts taking it seriously. They don't seem to take their blog search engine that seriously either. The results are horriblee and so not Google!

to me from where, and why. I dunno, it’s a habit I aquired long ago when I ran an unsuccessful dot-com back in the boom days (boom? more like bust). Anywho, sblogs (or, to the somewhat less initiated, spam blogs) have been around for a while, and as Chris Pirillo pointed out, Blogspot is notorious for attracting these creatures of the blight. Ah, yes, the sblog roams free at Blogspot. Not that Google isn’t interested in doing something about it - let’s face it, they have to pay for all that idiotic bandwidth

I think we should have hierarchical labels, similar to how Wikipedia has hierarchical categories. For instance, the labels “cats” and “dogs” could exist in the following hierarchy:

I. Animals
A. Pets
1. Cats
2. Dogs
B. Wild Animals
1. Ostriches
2. Elephants
II. Plants
A. Houseplants
1. Hostas
2. Azaleas
B. Wild plants
1. Dandelions
2. Crabgrass

Actually, Wikipedia has more of a web of categories than a hierarchy. So, for instance, “Elephants” might fall under the categories of both “Animals” and “Wild,” and the category “Wild” could have the subcategories, “Wild animals” and “wild plants.” Play around with Wikipedia’s categories sometime. You’ll see what I mean.

That kind of scheme, if properly implemented, makes it much easier to find what you’re looking for.

Someione should take you and beat the ******* **** out of YOU for not having a life!

What Do You Think?

 
Blog Widget by LinkWithin