How would you prefer to report Twitter spam?

The above comment assumes that enough clicks on the "report spam" link will automatically suspend the account. Unfortunately this is one of those places where human interaction is still the best possible filter.

How about a "report spam" link, which then submits the job to a queue of Human Interaction Tasks on Amazon's Mechanical Turk? This will place the spammers under human scrutiny.

Or how about a queue where Twitterers can go and check all the possibly-spam accounts, and give their vote on whether they think it's spam or not? This will keep it in the community, and people will feel that they have a hand in improving the service that we all use.

You could even hook up a Bayesian engine to the vote queue and let it learn from the humans :-)

As discussed in the previous thread, one person's spam isn't another. I would actually offer 2 different options to users that block people: Mark as 'spam', or something like 'I don't care for this person's content'. That way you don't get people who just don't want to be followed by someone reposting lumping their complaints under spam. This would allow you to weight a spam score that, when triggered, could shut down an account for a review.

A block link should be included directly in the email notifying you about a new follower (maybe call it 'review this follower', so you're not implying you need to make up your mind about blocking a user from the notification email along). It would then lead to a page with the 2 options above. The page to block a user should include that user's last 5-10 tweets, so someone can get a feel for the kind of updates they do, including their homepage URL.

Lastly, if you're going to act in any way on spam users by removing their access or accounts, there should be a clear posting (not hidden in the TOS) that explains what spam is to people.

I'd like to see this is a few places. As others have mentioned, it should be on the profile page, probably next to the 'block' button.

I'd also like to see it on the New Follower email
(see http://getsatisfaction.com/twitter/to...)

and I'd also like to see it on the Followers page, next to the 'follow' and 'block' buttons (for people that I'm not following yet but are following me)

I agree with Joost's comments related to the nature of the spam. Personally, I am most offended by the New Follower spam but I know that others are offended by too many links back to a web site / blog and automated updates for things like announcing that they are live on Qik, come chat! So at the risk of making this overly complicated, we should have some kind of drop-down box and/or text box that allows us to describe the nature of the spam.

There's a new form of spam. @mytweets is exploiting the most used hashtags, like #proxxima, #nob , etc , to spam people who track those strings.

I guess it should be safe to block anyone who posts the same link N times in a row...

Hi Crystal, [who is the duct tape of Twitter],

I think a 'report' link like we see on some forums would work, Except, I wouldn't punish people instantly. Some are pure spammers. But many misunderstand too. I once had a friend/folllower ask me for a link. And a stranger I never asked to follow me yelled 'hey that's spam'. Huh?

Thanks for all your terrific help since the beta!!

PS- Any update on http://getsatisfaction.com/twitter/to...

*oops, sorry link shouldabeen-> http://getsatisfaction.com/twitter/to...

what about we tweet @reportspam @spammerID. then you have a number of reports, times , so on. (@reportspam is you, and @spammerID is them, the bad spammers)

why am i a cupcake?

I think there needs to be a report spam button on the profile page.

It needs an inline command (ala @/d) so I can do it at any time, whether on the road or on the laptop.

You need fast reaction, so that spamming is never worth while, even if bot automated. That means automatic reaction, but you want to avoid human reporting errors.

So add a "spammer @foobar" command. It flags the account and reduces # of posts account can make per minute. On the fly guessing here. Have each report reduce outbound by 20% of total. So 5 reports makes it hit zero. At some level (which you can decide internally) you get notified to check it out. Otherwise a given spam report will wear off after, say, 24 hours. Accounts that hit zero might require manual re-enable, up to you.

Oh yes, and you can only use the "spammer" command on someone who has recently sent you a message.

Open question on how to notify sender. They ought to be told, but should they be told who did it? I guess not explicitly, but keeping them from figuring it out by timing is too hard. So just send them a system DM alerting them that they've been rate limited by a user who thought they were spammer. If they aren't, no problem, they'll be back to normal in a day. If they are, take a hike.

Other considerations.

You could require a CAPTCHA in addition to (but I wouldn't do instead of) the 24 hour period.

Do you allow people to flag as spammers only @reply messages, or also messages from people they follow? After all, they can always unfollow. You might want to have a different mechanism in the follow case (e.g. an option on the unfollow form and/or command to give spam as the reason).

You need to consider both sides of the equation. Always consider what the next spammer escalation might be. In this case, my immediate thought is that they might try to overwhelm the system with false spam complaints. E.g. set up an account, and then report a bunch of people they follow as spammers. (Particularly ones who are reporting on their other bots.)

This isn't as much of a problem for the @reply case. But it's a problem for the general case of my following back someone, only to discover he's a spammer.

Short term, you could decrease impact for people who reporting multiple spammers in short period of time, but that's not great. Or you could only count reports from such a person if others do the same. But it all gets messy. Down the road I think the right solution to that is to assign a reputation to a person based on age of account, connectivity and activity, and then assign more weight to reports from accounts with good reputation, and have less tolerance of spam reports on accounts with low reputations.

Two things to always keep in mind.

This isn't a technical problem, it's a social problem. In particular, it's a war against an intelligent enemy. Don't treat it like static that you are trying to remove from a phone line.

There is no winning this battle. It's about tradeoffs between ease-of-use and security. And it's about strategy and anticipating what the next move will be when you close a door. All you can do is try and make the cost of spamming sufficiently higher than the cost of defending from spam.

I'd be happy to talk about this more directly. I've spent a lot of time on the email anti-spam side of things. The technologies are slightly different, but the overall tradeoffs and strategies are not.

@nazgul

+1 for Kee Hinckley's suggestion of an inline command. (As an aside, I think I prefer the keyword 'reportspam' to 'spammer' since the former seems more like a verb -- makes it more similar to track, follow, leave.)

+1 also for the idea of adding more weight to spam reports from accounts that have a higher (internal, Twitter-derived) reputation score.

As to notifying the spammer... it might be best to do it several ways:

1. have something displayed on their home page that reads: "A total of x people have reported your account as a possible spammer, y of them in the last 24 hours. Your account may be suspended if z more people report you."

2. Every time they login to IM, send them a message similar to the above.

3. Send them the same message as a direct message

4. Send them the same message via the email address they used to register

Since most of the true spammers are probably using bots and don't bother to login to their accounts via the web, it's possible that #4 will be the most effective (unless they use throw-away email addresses).

Following on wendelscardua's point, I'd love it if Track alerts can be enhanced to recognize that we've blocked someone and therefore don't want to receive their tweets, even if they use one of the words we're tracking.

P.S. I just noticed that since this is a discussion, I couldn't star (or otherwise mark as noteworthy) anyone's reply.

Related thread: Nasty techniques used by Twitter spammers

I have to +1 for Kee Hinckley's inline suggestion too.

+1 +1 +1 for the karma/reputation idea. If we are ever going to have an online society/economy, we need to start building up personal credibility in a real way. People are on the internet for the long haul now. I, for one, would like to see a reputation return on all the years I've spent on the internet helping people on various forums and IRC channels.

Oh, and if any GetSatisfaction.com moderators are listening: when are you going to support OpenID logins?

For that matter, Crystal, when is Twitter planning to support OpenID? Perhaps to associate one's existing OpenID with a Twitter account, for added credibility?

I would like it if there were a mobile command that I could send back like:
spam @annoyingFer which would automatically flag them as spammers (which should tally, so that if say 5 people flag them their account is automatically blocked), and this setting should also automatically block them from sending me any more messages.

I would also like one that is
stalker @creepydude for people who are doing creepy stuff that freaks us out.

And maybe one for:
addwhore @letsaddeveryone for people that just ruthlessly add people that they don't know. I don't want to have to make my profile PRIVATE but I had to because I kept getting people I have never met, nor ever heard of, adding me, and then never answering my direct messages asking them who they are!

This way you can have different recourse for different types of bad behaviour.

Thanks for listening -i

I think ev is looking into the [@]mytweets spammer.

addwhore spammers- adding me, maybe I will follow them and their randomness... the "spammer is now following you on twitter" emails bug me the most. I don't follow spammers.

Right now we can:
FOLLOW
TRACK
HELP
TRIPS
Maybe we can
SPAMBLOCK

as in send an update that says SPAMBLOCK <username> and they're blocked and tagged as a spammer for the admins.</username>

In last week or so I've seen a pretty dramatic increase in the number of "spam" accounts trying to follow me. I've blocked most of them, but would like an easy way to report suspected spammers.

And while on one hand its nice that you're checking to make sure I really really want to block someone it makes the process a little cumbersome...and then after blocking it takes me back home instead of my "followers" list.

I'm pruning my list now so have multiple accounts following me that I didn't even know were there.

A the same time, I'd hate to see a legitimate outside service blocked by a minority of people.

Ditto what mdy said.

Flag as spam.

I repeat- TRIPS??? what is this? I can't find ref to this on the nets

I like a point system approach. If you get a negative point value to a certain point, your account gets suspended. Almost like how PlusPlus bot works. Except if you get a low enough score, it drops your account.

@dacort (of http://Tweetstats.com fame) has an interesting write-up on using statistics to identify spam accounts.

A "report spam" button would be great and also giving me the details of this person in the email you send to us to accept them as a friend would be better. I don't want to have to go all the way to the Twitter site just to find out that this person is using a bot to find followers. It's dumb and it's easily fixable. I want all the person's details in that email, simple fix!

I'd prefer to have an option to approve all following requests. I can ignore spammers, and people who want to follow me don't have to contact me offline -- I can approve them right from the notification email.

And don't forget the iPhone! I do a lot of my Twittering (following, etc.) from the phone--make sure it's easy to delete spammers from there, too!

For a *great* list of Twitterspammers (Twammers), please see http://www.twitspam.com/. This is (I think) the definitive list of them on Twitter--and I have been sent follow requests from more than a few on that list.

I'm kind of alarmed at the idea of a 'follow whore' being flagged, because I set out to find the kinds of people/ groups I'd like to follow- news agencies, politicians, journalists, experts in various fields; and then people doing interesting things, or know SO MUCH about an area that I think is important; and there are SO MANY that I ended up with a huge list I'm following. I like it a LOT. It took a lot of time to find all these. I don't expect every one of them to follow me, and I would be surprised if they did. I'm really a nobody.

I mostly just watch how everyone uses twitter so that I learn and understand. I only post, at most, once or twice a day.

I also cull people who post twits that just don't interest me- my thought was that it would eventually condense and the value to me would improve.

I've used many 2.0 sites and have suffered a great deal from the floods of spam. In my experience, it has been the internetmarketingguruwannabees that are responsible for bringing most of them along with them, promises of great riches just for posting links everywhere.

They take great sites, like twitter, and destroy them AND their reputation along with it, and the honest and committed users lose a good thing and suffer.

So let me say a couple of things about this:

1. the solution MUST be community based, or the demands on the staff will be too great, and how twitter runs will necessarily change forever, as will the costs.
2. No one should EVER get a twit that they haven't personally chosen to follow.
3. Consider making an 'entry level' for newbies that observes how they handle it. Automated 'warnings' can be given for flagged activities, and graduation could be after an extended period of use.
4. Add a ratings system- overtly sexual things should be rated X right on the profile, so that people aren't accidently adding an X rated twitterer.
5. Add a spam function next to the star and reply on the twits. It should be on EVERY twit.

I think that Google's perceived 'reliance' on backlinks for pagerank is the reason people are spamming all the web 2.0 sites. If google didn't hold back links as so important, many of these people would just go away. I wonder how many great social sites will be destroyed by this alone before Google will change things?

Rating Twitter pages is going to be really hard. Yes, there are overtly sexual accounts. But one of the great things about Twitter is its mix of work and home, personal and professional. It's like a workplace in a world without sexual harassment. I really don't think we want to go down that road. When it comes down to it, the issue isn't about content, it's about forcing content on others. When you look at it that way it doesn't really matter what the content is--it's whether the sender is abusing the recipient.

Crystal, I just noticed that the 'spam report' option is no longer available on from here: http://twitter.com/help/ Please tell us that you guys are about to roll out a Report Spam button on the Profile page???

I am less interested in reporting spam than I am in identifying someone as a spammer as quickly as possible. If you haven't seen Merlin Mann's post at http://www.kungfugrippe.com/post/3137... check it out to see some good ideas on what statistics can be used to identify someone as a spammer.

If the email that came to me said "so and so is now following you. They have been using Twitter since... Follow x people and have x followers... " Stuff like that. Maybe even some of the other stats like %age of total posts that are “@” responses and %age of total posts that include a link, especially if most of the links are to the same site.

If this info is in the email itself, I don't have to click to visit their profile at all. I simply delete the email. (I doubt I would go through the trouble of reporting them at all. Sorry.)

Lately the amount of Twitter follow request emails I've gotten has been overwhelming, so I've just turned the email notification feature off. If the email could be more informative, cutting out the step of going to the person's profile to see if I want to follow them back, or if they're a spammer, I would be more inclined to keep it on and look at the emails.

Sarah

+1 for Sarah's suggestion. Having the data in the initial email would be great. I l think the email should have a couple of links: 1 - FOLLOW the new follower; 2 - BLOCK the new follower; 3- REPORT the new follower as spam/potential spam.

Excellent ideas! As I said, why has it taken this long to implement something?With the technology it could have been done since day one.

mdelfs said:

why has it taken this long to implement something?With the technology it could have been done since day one.

I'm just guessing, but here's what I think -- since we each have to explicitly choose to follow someone on Twitter, the thinking was that Twitter users wouldn't have to deal with that much spam, and therefore spam-handling wasn't a priority on Day 1.

Later, Twitter *did* recognize the problem and they actually took action by adding the Block feature (that wasn't there on Day 1 either). I believe that came about because some people were persistently unfollowing and refollowing people, just to generate a new 'user is following you' email. Block took care of that.

Now that the situation with spam has gotten worse with the spambots, I'm encouraged that Twitter is listening to our grumblings and is evaluating ways to address the problem.

A final note: in a recent FastCompany interview, the founders say Twitter has been prioritizing enhancements that will improve the system's reliability, and I for one can understand that. People will stop using Twitter if it's not reliable, and if no one's using it, then the spam situation won't matter.

I hate adds from people who are either spammers or just foolish people who follow thousands of others for no apparent reason. In my dream world, I would receive a monthly digest/report email of new follow notifications rather than each one in its own email. By each person's username, I would want to see the following:

a) Their twitter profile: Name, location, Web site, bio.
b) How many people they follow and how many follow them.
c) How many people we follow in common.

This way, I only get bugged once a month, and I can easily determine which if any of my new followers are worth following back without having to go to their individual twitter pages.

Whaddya say, twitter?

+1 for weskimcom's proposal for a "digest" email. I proposed something similar on this thread:
http://getsatisfaction.com/twitter/to...

Lets say weekly or bi-weekly and I like it!

Agreeing with several other people:

1. In the web interface, I think this link should appear near the "block" link.
2. I would love to be able to do this by SMS. When a freaking Toshiba TV follows you, it's pretty obvious that this is spam and it needs to go away NOW.
3. I am uncertain how I should deal with the somewhat-spammy friend whores. I've gotten follows from people who've tweeted "I'm trying to get ____ followers by the end of the week!" and have clearly friended every friend of every profile they've clicked on. This feels like spam to me, because they don't know me, they don't care about what I post, we have nothing in common, they're just hoping I follow them back so they can look popular. Is this spam? Is this okay user behavior? It dilutes what I'm trying to accomplish with Twitter. Is it time for a max number of new people you can start following per day?

I think it's pretty simple. If someone is following ten thousand people, and fifty-six is following him/her, it's pretty damn obvious that this is a spammer. Just put a cap on the amount of people you can follow; you can't realistically follow over a thousand people, unless you only use twitter for posting and ignore reading the people your follow.

Are people having problems with spammers hitting them with multiple follow notifications? That's the only rationale I can come up with for the current "block" functionality. I get follow notifications from spammers, but I don't follow them back, so they can't really bother me at that point, can they?

Honestly, I don't care how many follow requests I get from those users who are "following 1000s of people but only followed by a handful of newbies." These are simply spammers/morons whom I will never follow back and therefore will never have to hear from again.

What I *do* resent is that my email notification simply tells me that so-and-so is following me, and I have to actually go to the person's profile to determine whether or not they're someone I actually know or if they're just a spammer/moron.

This is a waste of my time -- let's put more useful information into the notification emails themselves! If I see that I have a follow notification from someone who's following 1,000 people, followed by 12, and doesn't follow anyone I do, I can instantly delete the email and never have to think about this person again.

Better yet, how about an email notification that looked something like this?

****

Hi, [your username].

Spammer/Moron ([spammer/moron username]) is now following your updates on Twitter.

Spammer/Moron's stats:
Name Spammer/Moron
Location Spamtown
Web http://www.spammermorons.com
Bio I'm a spammer/moron!
Following 10,351
Followers 12

[link]Follow Spammer/moron with notifications on.[/link]
[link]Follow Spammer/moron with notifications off.[/link]
[link]Block Spammer/moron.[/link]

Check out [spammer/moron username]'s profile here:

http://twitter.com/[spammer/moron username]

Best,
Twitter

--
Turn off these emails at: http://twitter.com/account/notifications

****

Whaddya say?

weskimcom's suggestion is made of win, but in a perfect world I'd never even get that notification, because it's from a frickin' spammer. I say, implement weskimcom's suggestion ASAP (since it's real easy), add spam ratings to ban/block notifications and @'s as we move on (since that's much more difficult)

Edit: Also, reading the scrollback more, digests sounds like a great idea as well. I say implement both, but if I had to choose, I'd choose to get more statistics in the mail rather than to get a monthly digest of almost useless info.

Wes' suggestion is killer--that's exactly what *I've* personally been hoping/crusading for--just give me *all* the person's details in the email so I don't have to go visit the Twitter site *just* to say "don't follow" or "block" or whatever--this is crazy since I sometimes have up to 10-20 of these requests a day, especially while I'm on the iPhone. I love Twitter, but, I don't feel the need to go to the Twitter website a few times an hour for no good reason!

I am one of those that follows FAR MORE than follow me. As I said above, I'm following a LOT of people- politicans, news organizations, journalists and etc. that I don't expect to follow me back. All I want is to keep up with what they're saying and doing.

The idea that I'm a spammer or a moron because of that is insulting and just plain wrong.

There are LOTS of people that put a lock on their account to prevent followers, and if there was one I didn't bother that person no matter how interesting they are. Locked accounts are just that- they don't want followers. Therefore, if you don't want any, just 'lock' your account.

Spam, as Kee mentioned above, is abusing people with information. Following is NOT dumping unwanted and useless information in people's lap.

How I deal with all the tweets from all these people and organizations I'm following is my problem, not yours.

If the information I tweet to people that don't want it is spam, then I deserve to be flagged and/or stopped.

I see Twitter as a 'continuous people's news feed' which is from real people, much like a news feed is from the bbc or elsewhere. Will I read EVERYTHING the BBC issues? No. Will everything they put out there interest me? No. Should I quit following their tweets because of it? No.

If Bin Laden were tweeting, would I follow? You bet. Would I expect him to follow me? No way. What for?

Yes, followers and followings get out of balance. Who cares?

Margaret, that's cool that you follow as many as you do for your own reasons--but--let's be honest here, I'm not concerned with you--I think we can agree that we are targeting those who follow FIFTEEN THOUSAND people for example with no reason and no possible way for a human to actually follow that many people. Most of these people are using TwitterBots to just randomly follow people which I think is 150% wrong. THOSE are the people that need to be stopped, not someone legitimately following many people.

Margaret, how many people are you following on Twitter? If you're following a thousand or more people and are actually able to get something out of that, I stand corrected and withdraw my initial suggestion.

I am following over 4k and I have almost 1k following me. I admit it is much but i have been culling as I see that many of them aren't tweeting things of interest to me or use it for specific/personal reasons only. I also cull people who carry on conversations all day long on twitter by the hundreds of tweets.

Yes, I've gotten a LOT from watching that feed, which I leave open the entire time I'm on the computer which is almost 18 hours a day.

It is MUCH more interesting than anything the msm churns out as news. I am a people person, and the real people have much more interesting things to say that are much more significant than the media crap we're expected to swallow.

Admittedly I don't talk a lot, I perhaps tweet once a day, maybe twice. I do all my talking on my blog instead. I only tweet when I have something valuable to say (I think).

I would REALLY be upset if spammers ruined this for me. I like it far too much.

So...as we have said over and over, a simple email with the critical details of yourself would allow the person to either accept or deny your follow request--I think that's all people want. In other words, I don't care if you follow me, but, if I'm expected to follow you back, I want some details as to who you are and that you aren't using a bot to just snipe people for your own marketing purposes.

mdelfs, I think you mean the internetmarketinggurus. As I said before, at other 2.0 sites it was the 'gurus' that brought along the spammers, no matter how many times they deny that. They are looking for 'deep links' and for them, the more the better.

Their reasons for doing that are much different than mine are.

I don't know how to address that problem. I haven't seen anything work with them.

The ONLY thing that will make that stop is if Google quits relying so heavily on backlinks for their page rank and search results.

I agree with the Report Spam button. I don't think the potential of abuse would be too high - it would be good to have some people check spam reports so that legit accounts don't get mixed in (for example, pressing the wrong button).

I would like a button or link anywhere easily accessible. I don't like having to dig through to the help contact form to report spam. Putting a link somewhere higher up (and unobtrusive) would be ideal. I also prefer having spam reports reviewed, maybe after a certain threshold have been reported, for Twitter staff verification.

a "report spam" link on the profile is a must, but I use twitter mostly from my gmail account, so there also should be a way to report through IM services

Somehow, I want it INSTANT! I just got hit by someone spamming about DIVORCE websites and I am assuming she got my name off the timeline and somehow immediately took my whole list because they all started posting about the SAME SPAMMER. I don't mind when my FRIENDS post something about a website they found interesting; they do it rarely and I don't consider it spam. I want to be able to go to the spammers profile and report them RIGHT THEN AND THERE. I blocked them but they are still out there spamming if no one cuts off their account completely. It needs to be more hassle than it's worth for them to constantly be setting up new accounts.

would it be weird/confusing to have two buttons- "this is TOTALLY DEFINITELY SPAM" and "I think this may be spam, somebody check it out please?"

I'd like two features. First, let me set the context. I use Twitter to keep track of my pals, and to let them know how I'm doing. I don't mind that my comments are public, but I mind a lot that a few times a day I have to deal with messages from Twitter each day notifying me that spammers, self-promotional jerks, and confused people are following me.

So one, I'd like to be able to mark my account as a personal one, rather than a broadcast one. In fact, the personal account should be the default, and the self-marketing/brand extension/attention whore mode should have to be manually turned on.

The personal mode means I get notifications of subscribes, but that you guys are very choosy about who you let follow me. If they have a suspicious ratio of followers, if their account is new, or if they aren't connected to people connected to me, then you rate limit them heavily and manually review their accounts before lifting their follow rate. In broadcast mode, anybody can subscribe,and you don't get an email upon every change in your fan base.

Second, so that amateur anthropologists like Margaret can still follow lots of people, give her some way to follow people quietly. Maybe that means she can follow without generating a notification. Maybe you introduce a new category of following, introducing the friend/fan distinction. Or just create a whole new app called something like TwitterWatcher that lets her keep large lists of people to spy on without doing anything that implies a personal relationship. Or something in between.

unwanted email = spam. I WANT the emails that say "your friend has finally found twitter and is following you!" I DON'T WANT the emails that say "spammer/jerk with 15,000 follows is following you too!" I don't care if they can read my tweets, can see me in the public feed, are selling snake oil- but it's the email that drives me up the wall. The extra time it takes for me to visit twitter and go "ok, nvm, spammer" annoys me. If added info was in the email, then it would rock. The email wouldn't be annoying.

Oh, but if we don't have to visit twitter.com, we don't give revenue for any pay-per-impression ads that there might be.

AMEN Crystal!

from the getsatisfaction rules:
"If you see content that crosses the line of acceptability (including spam) please mark it "Flag for Review." We'll remove flagged items if we agree that they're no good."

How about something like that for twitter?

My 140 character response was, "Twitter sorely needs to 1) tell me mutual friends of people who add me and 2) let me set an obvious question about me that adders need to answer. Right?"

An additional report spam link would be great too. I'd rather not get the spam in the first place though.

So when is twitter going to take action on this? I just started tracking a few new topics and am finding a ton of new spammers. Try tracking "live mesh" for example.

Guys, if you are using Firefox, this Greasemonkey extension is AWESOME! It tries to watch the ratios on the user's Twitter page and alerts you if they are normal users or Twammers (Twitter Spammers) based on the followers they have. Try it out, I love it! http://www.splitbrain.org/blog/2008-0...

I would prefer not to have to go to their webpage to report spam. 90% of the time, I can tell just by their username.

Anybody who signs up to follow 100 people in one day is most likely a spammer. They should be flagged. (And then they should be flogged. :-P)

I'm coming in late to this discussion and don't have time to read everyone else's comments, but I will throw in my 2c anyway. :P

1. Add a "report spam" link on each profile page and "your followers" page.

2. Add a "spam [username]" SMS command to allow people to *report* spam.

3. Points (1) and (2) need to be complimented by people using the tools they currently have. If someone is following you and you don't like them or think they're spam, BLOCK them also.

4. Modify the tracking function, so that if someone sends the SMS command "block [username]", you no longer receive messages from them based upon your tracked keyword list - even if they're not someone you follow. This will make point (3) much more effective, ie., you can block someone who is spamming you via a tracked keyword of yours (as well as report it as spam using point (2)).

4.a. Include a function on twitter.com to be able to view all the people you've blocked (under your account settings - keep it off the profile page). Drop the font size right down, add in 2-4 columns and let's get 100-200 users per page, as it's only going to be for quick review and no one will want to sit there clicking "next page" after 20 users each time just to see who they've blocked.

5. Any spam reported needs MUST require the account in question to be investigated by a HUMAN. There is too much potential with being able to even simply report spam for people to either "as a joke" or actually vindictively have them and their friends continuously report someone's account as spam when it may not be.

6. Although it would be useful to back all this up with automated reporting and investigation (eg., bayesian filters), the final decision still needs to be made by a HUMAN working at Twitter. The community input should only be for reporting only. Again, as per point (5) - we need to limit upfront the potential for abuse. Automatic closing or even simply restricting account usage to x number of updates an hour/day based upon community input or automated reporting/filtering will undoubtedly result in the limiting or shutdown of innocent people's accounts. This is just as unacceptable having spam itself.

7. If someone at Twitter investigates an account and determines it to be spam, then that person should no longer be able to use their account, and all their followers should be automatically removed. Send an email to those followers to let them know someone they follow has had their account marked as spam, so they are no longer on your friends list.

8. Include a function on twitter.com to allow you to simply run a report on your contact (following and followers) list to determine who is spam by the number of times it has been reported as spam. The minimum number to show up in this report should either by very high (in the thousands** - again, to prevent potential abuse by people unfriending you because you have simply had your account reported as spam when it may not be) - or based upon the current external tools such as twerpscan.com where it looks at your following:follower ratio.

** Don't worry, if it is spam, then it will get the numbers easily enough from multiple circles of people. If this number is low, it could be achieved by one circle of people (back to that potential for abuse thing..!).

9. I personally don't see the point in including a "report spam" link in the email you receive to let you know someone is now following you, as wouldn't you have to check out that person's profile in order to determine whether it's spam? Use the link that would be provided under point (1) to do that.

For those who hate the spammers, check out TwerpScan: http://twerpscan.com/

I would love it if Twitter integrated something like this, so I could say up front that people with bad ratios couldn't follow me unless I followed them first.

What's wrong with a 'Report Spam' link, and after a certain number of flags from unique IPs, a member of staff checks the reported account over?

I'd like to see a simple "Report Spam" link on Twitter.com, and possibly right in the email "[spammer] is now following you on Twitter." A simple report spam link there would be helpful too.

Another addition so it's easy to report spam using the api and all Twitter post devices/websites, etc, is a simple "reportspam username"

Just a suggestion.

I would like to add one statistic to my previous suggestion for more robust reporting in "new follower" email notifications:

http://getsatisfaction.com/twitter/to...

That would be a % figure that tells me how many of a person's tweets contain URL links. While there are legitimate twitterers who routinely link to blog posts and whatnot from their tweets, I've noticed that a common spammer technique is to post tweets claiming to be about some useful piece of news but always with links back to their site, presumably to drive up their hit counts.

Hate 'em...

Oh, and in case no one has mentioned it yet (I did a quick skim but I didn't see it), it would be great if the Twitter mail that notifies us of new followers also includes the date when that user joined Twitter, similar to the date that we see when we send a 'whois' command via sms or im.

Generally speaking, the spam accounts are fairly new, so the join date will be another indicator (i.e., I can see how many people they've followed since the time they joined).

My 2 cents:
* I would like an easy way to report when accounts are bothersome. Not just spammers (like "successstories" which seem to be one) but also 'twittergatherers' or whatchacallum. I mean people who just seem to hoard twitterers. It should of course be clear that reporting doesn't automatically mean that those accounts will leave your followers-list.
* I am worried (as others have already said) that the punishment for alleged spamming is immediate removal. I would like to see human intervention and maybe a 1 warning system. For example: I use stumbleupon and twitterfeed to let my friends as well as my own memory know when I found a nice site. This can be seen as spam. However, I do not feel this is spam! And many others would agree.
* the idea of voting in some way what is spam might be nice or it might be chaos, depending on the way it is executed. A lot of 'is so' 'is not' can occur.
* It might also help to be able not just to block a person (which means 'never again') but also to unblock without having to follow someone first. Maybe a way to remove someone as a follower (but not block) will be more helpfull in many cases in addition (not as a replacement) to the block button.

Anyway, these are just my thoughts on the subject. Please excuse my English if it isn't all that clear or correct (I do have a splitting headache and English isn't my native language).
I am curious as to where this all will lead...

Like others, I'm worried about a completely automated mechanism where an account would be removed when a "sufficient" number of "spam" votes has been received. A few thoughts:

* In the web interface provide TWO buttons: one "spam" and one "not spam" so people can "vote" either way. When the "spam" vote is higher by a certain amount or percentage than "not spam", FLAG that member's updates so others can jump in to either confirm or vote "not spam". One person one vote, you could even change your mind by voting the other way, but you still get one vote. Then human review by Twitter after some period, if the "spam" vote is still (considerably) higher than the "not spam" vote - otherwise just "expire" the voting and flagging. This would go some way towards preventing someone being removed as spammer merely because a bunch of people dislike the member, and might also reduce the amount of effort needed for review by/on behalf of Twitter.

* For the IM/phone interface provide two commands that are equivalent in action to the "spam"/"not spam" buttons on the web interface.

* The IM/Phone interface would also need a command to block-in-track so that a "keyword spammer" won't get through anymore (regardless of whether the blocked account is/will be seen officially as a spammer)

I realise that "block in track" as an option is a separate discussion, but it ties in closely, so I'm mentioning it here merely as a reference.

By allowing voting both ways with some checks and balances in trigger levels and elapsed time, followed up with evaluation, I think all-too-hasty banning of what looks like a spammer (or someone who is merely disliked by a group) would be prevented. It wouldn't be fast, but I think that's a good thing. With many "not spam" votes, a review might not be necessary; but for a review a suspected spammer might be given a chance to defend their case (unless it's too blatantly obvious, maybe).

You can do "yes" voting without a reputaion system, and even that probably would not be enough. Otherwise the spammers will simply set up bot accounts to vote yes.

The key to coming up with any anti-spam idea is to consider what the spammers will do in response. Often the response is worse than the current problem.

Always keep in mind that the spammers have essentially infinite accounts and bandwidth and need only an infintesimal response rate. Solutions that slow them down or cut them off quickly are best.

Kee Hinckley brings up a good point about "what will the spammers do in response?" This is definitely something to keep in mind when looking at how to tackle this problem - and voting accounts as either spam or not spam is going to be exploited.

As an example, I've read various reports of spammers hiring people to overcome manual intervention sections of sites where completely automated creation of accounts was failing - for example, answering challenge questions, solving CAPTCHAs, etc (sorry, it was a while ago so I'm not sure where I read it, just remember that I did!).

A good system would be a Time/Rating and double check system. This would be fair to anyone, even if a mass of people decide to take out someone they just don't like:

Rating system: Everybody starts at 0. If the rating goes to -25 (negative 25) it's real bad.. if it goes up, then we all like that person behind the profile!

Double Check: The Twitter people get a warning and take a look at the profile in question (and the messages!) and from there on decide whether the profile should be removed if it's with very poor rating.

Solution: Let the rating system warn the Twitter people and from there on let them look. Of course, the rating warning for the Twitter people should be in a certain time limit.. for instance: one week of continuing negative ratings, they look.

Ask others to help and translate when messages are in another language.

Also try to see if it's possible to use Akismet on Twitter. Works very good on Wordpress blogs..

It seems easy to find the obvious automated follow-spammers:

> 10,000 followEES (people the spammer follows)
< 1000 followERS (people who follow the spammer)

You could also add blocking stats to really eliminate any false positives. e.g. > 5% of your followEES block you.

This would protect people like Scoble who follow lots of people, but flag follow-spammers who load the system for no one's benefit but their own.

Lycos mail deletes your account if it hasn't been logged into after X amount of months, I like that method.

I'd say put a "flag" link on the user's profile. It's hard to find the spam report now.

Someone who's marked as a spammer shouldn't be able to follow anyone, with or without notification.

Having to manually clear out porn spammers and such every week or two is NOT helping matters.

I just wanted to say that I am most annoyed by the current solution: Not informing me that a spammer is following me, so that I need to hunt through the list and identify who's new and spamming. I generally like puzzles, but this isn't one that I appreciate.

I'm not sure if it's too late to be replying to this topic (or if this idea has aready been covered). But I just found this topic *after* posting my own idea.

http://getsatisfaction.com/twitter/to...

It's a little idea for a spam repository like Akismet for Twitter.

Seeing a surge in spam followers. I would be happy to help @Crystal. You may email me anytime.

Something has to be done and soon, this is a potential wrecking ball for twitter. Once you start getting offered £3-5k a week and get offers of scantily clad girl in the same minute there is something going seriously worng.

A report spammer button needs to be implemented. I also like the idea of the Akismet Twitter spammer......

I find that reporting spammers via the http://twitter.com/help form and selecting 'spam request' as the 'request type' works quite well.

When Twitter receives the spam report, they investigate, and usually around 5 days later, they give you an update on their findings and what they've decided to do.

@crystal also mentioned in a separate thread yesterday that they've just 'removed a large number of re-tweet accounts', so they're definitely taking action.

It's just too time-consuming right now, period. Going through & blocking spam accounts (both those that are just repetitively linking to their site as well as the retweeters from the public time line) currently takes several clicks as well as forcing me to log into more than 1 account that I manage.

Trying to report all of them via the help form or an e-mail to support or help with a particular subject line is even more time-consuming. This is completely KILLING the joy of getting new followers. Please help!

Twitter should/could do the following *right now* to make things easier:

- Include a "block" link in the email.
- Include the # of people this "person" follows, along with the number of people who follow them. This helps us determine if they are legit or not without having to click on their profile.

They changed the Twitter Help Request page. One can no longer choose 'spam request' and simply submit. In fact, when one submits a request there is no drop-down to indicate that one is submitting a spam account for consideration. They're also requiring one to provide a name and e-mail address (phone # optional) when submitting a help request.

It's been a few months since this thread was started. Crystal, would it be possible to get an update on this? We know that you guys are taking action against spam accounts - which is great - but the process for reporting spammers through the Help page is clunky, hard to find for most people, and time consuming. Are there plans for a more streamlined approach - e.g. a Report Spam button on the Profile page? Also, there is a Twitter account called @spam that claims to be associated with Twitter HQ. Is this a new channel for reporting spam?

Here's my suggestion,. in the form of another topic I'd started, unaware of this one. RE: FolloWhores, as I like to call them. http://getsatisfaction.com/twitter/to...?

Thanks!

I mentioned this on Twitter earlier:

"@crystal Regarding that how to report spam topic on GSFN: Possibly a Twitter account anyone can DM with a username of a spammer?"

I'd rather keep the conversation about twitter spammers here, so I'll post a few more of my thoughts here:

Before Twitter gets an easy way to report spammers (maybe a "report spam" option on the Twitter block page?) it'd be a good idea to either:

a) Have a Twitter account that will allow anyone to DM it with a (potential) spammer's @nickname, so as to not publicize and clutter public twitter feeds with spam reports.

b) Setup a GSFN thread where anyone can mention a spammers name and a member of Twitter support will receive an email with the spam report immediately (using GSFN's follow function). It doesn't have to have a reason behind it, and it wouldn't require Twitter support to respond to each request - just investigate the account and delete it if necessary. After a bit of thought, I *do* like this idea more than the DM one I originally suggested.

And of course, that "spam report" option/button on the block page would be useful and more user-friendly than the ideas I posted above.

Thoughts? :)

Plenty of solutions offered in this thread, some easily implemented, but yet none are and the follow spam continues (increasing massively, again, this week).

I would like a report spam button just like in gmail. then twitter could block the spammer for me and then find out if he actually is a spammer, by counting the times he's been spam-blocked.

yes, a simple report spam link or button on the profile for community-wide flagging a la Craigslist would be great.

Please.

Yes vote for "spam" button on profiles. Alternately (may be suggested already, haven't read/searched posts), admins should be automatically alerted when any user is blocked repeatedly, likely indicating a spammer, so as to take action before user is specifically reported as a spammer.

I would like a way to pull the user's blocklist via the API so that external sites can manage spam reports in this manner.

I looked into if there was a way to pull the blocklist via API, and I also found that twitter does watch block counts to determine if someone is a spammer, so that feature you all keep asking for already exists.

Also, I found a site, My Tweeple, whose goal it is is to have you manage your following there, so that when you block someone it hits the front page as a 'ding'. However, since I don't think this is super popular (some of the huge spammers, even the ones already deleted, only have 20 or so dings), it is probably not very effective as it's not well known.

"...that feature you all keep asking for already exists." Well that's nifty. What would be even niftier is if Twitter posted official responses to these complaints/requests in an easily viewable form, i.e. updated their FAQ to include this information so we don't have to waste our time filling out forms to report spam because the FAQ tells us to and we don't know how else it will get reported.

Maybe have a button similar to the block one under the block button called "Report Violations" and then have a set of options for different TOS violations, SPAM being one and then have it display on a screen and/or intranet at Twitter HQ.

Maybe have a button similar to the block one under the block button called "Report Violations" and then have a set of options for different TOS violations, SPAM being one and then have it display on a screen and/or intranet at Twitter HQ.

I'd favor this be tempered with internal metrics at Twitter. I've describe the idea here http://gsfn.us/t/khp but having a feedback loop for items that fall outside the normal automation process is critical. So, I'm in favor of a "report" function of some manifestation.

Crystal started this discussion 4 months ago offering a solution many of us have supported: simply put a "Report as spam" button on the profile page like Gmail has. That hasn't happened. Why? I suspect because Twitter doesn't have the staff to deal with even more people reporting spam!

For this reason, I'll suggest something slightly different...

How about an automated system whereby once the # of people Blocking an account exceeds the # of followers of that account, it is automatically reported as spam?

The issue I have is that it's relatively easy to block a spam account, but it takes at least 2x the time & effort to go copy the account name, hit the Help button, click on Submit a Request, choose spam request, paste the account name into the field, and scroll down & press the Send button. This is OBVIOUSLY a pain in the butt, and it likely hasn't been changed because it's not in Twitter's favor to overburden their staff by making it easy for us to report spam. I do it for every damned spammer, but that's cuz spammers suck & I'm not gonna let 'em get away with it. Still, some days (like today) I spend easily 20-30 minutes doing all the steps above for every spam account that follows my accounts!

So... IF WE'RE NOT GOING TO GET A "REPORT AS SPAM" BUTTON, CAN WE AT LEAST GET SOMETHING BETTER than the current process???

The advantage of the idea I suggest above is that it wouldn't require any of us having to report an account as spam if we don't want to. Eventually, a true spam account will have more people blocking it than following it. A particularly annoying spammer should be able to reach this quite quickly. To be on the safe side, a bottom limit could be set before this process would be triggered: for example, only when an account is blocked by at least 100 different twitter accounts would this process come into play. This would protect an individual who is not a spammer but might upset some people at some point. The second protection, of course, is that once the # of blockers exceeds the # of followers it only REPORTS that account as likely spam. This would not automatically set any account as spam, but would simply put into play the same action as if any one of us were to go submit a 'spam request' in the convoluted process mentioned above.

Long enough. But I'm sick & tired of those damned 6figureteam, beta.goofy2.com, & petpals of the world wasting my time on Twitter. Thanks for listening. (is this thing on?)

Cheers,

@ggroovin

Too many of these suggestions pay no attention to how spam actually works.

You can't ban IPs, registrations almost certainly come from compromised user machines, of which there are literally millions to choose from.

Twitter just posted a new blogpost regarding their efforts to combat spam.

See "Turning up the heat on spam" on the Twitter blog

Also, there's a new Suspended Accounts help page (had not seen that one before today) that lists the various reasons why an account may be suspended, spam being one of them.

---- start of excerpt ----

What is Spam on Twitter?

Commercial or promotional use of Twitter is allowed. There are many companies who create valuable, opt-in relationships with users on Twitter who want to keep up to date with them.

However, if you are following other accounts in order to gain attention to your account or links therein, you may be considered spam.

If you are creating a series of accounts in order to promote the same thing, you may be considered spam.

If you are sending large numbers of @reply messages that are not genuine replies, you may be considered spam.

If you are creating updates in order to show up in search results, you may be considered spam.

If you are disguising links (i.e., writing about one thing but linking to another), you may be considered spam.

If a large number of users have blocked you (relative to those following you), you may be considered spam.

---- end of excerpt ----

I like the @spam method, but it feels that a "Flag as Spam" button would be easier.

Yikes! This is a long thread. It is a good one though.

My suggestion is along the lines of many others here, but the details may be different:

Add the "Flag as" link to the profile pages. I suggest a generic "abuse" label, where flagging it would bring the end user to a form asking to select the type of abuse and allowing an explanation/comment. These would be tallied and the results sent to an employee tasked with investigating. The employee would work on the highest tally first and would have the option of clearing the abuse flags, suspending the account, or taking other appropriate action.

Looks like we can now officially report spammers by sending an @reply to the @spam account on Twitter.

At least, that's what a recent update from Crystal in another thread says.

The recommended syntax seems to be:
@spam @spammerusername OptionalMsg

A button would be great. Also love the idea of just sending a Tweet - @spam @spamusername Optional Message. But the button is good enough.

'Report Spam' button. If enough people press it, it should automatically suspend the account.

Okay, chill out Tom.

It could be a relative measure, i.e. the percentage of followers who mark as spam.

We can now also report spammers by sending a Direct Message to Twitter's @spam account.

To do this, you'll first need to follow http://twitter.com/spam. You can then send a direct message to @spam either by:

1. Typing a direct message from the "What are you doing?" box
Ex. d spam @spammy1, @spammy2 are spambots!

2. Sending a direct message from your phone
Ex. d spam @spammy3 is yet another spambot

3. Composing a direct message at http://twitter.com/direct_messages

There's still no SIMPLE way to report spam accounts? Seriously? This discussion has been going on for SIX MONTHS. Ridiculous. And yeah I know about the twitter spam account but really there needs to be a one click option to report an account.

i like the report spam button

it should be on the profile page

I occasionally get direct messages to the Twitter account I manage for my company. They're almost always spam of the "thanks for adding me" variety -- we have auto-follow-back. I was thinking the other day that it would be nice to have an e-mail address to forward these e-mailed direct messages to, kind of like the phishing@ email addresses for ebay/paypal.

A quick 'report this' button would probably be the best answer.

A "Report Spam" button would be super. I think this button should do only two things automatically: add one's report to others about the same user, and block that user from following he-or-she-making-the-report. I think it would be a huge mistake to have someone banned or deleted automatically based on such reports. As folks have said, too open to abuse or simply differences in defining "spam."

I'd like to see a "Report as spammer" link in the list of Actions on an individual's page. There are people who tweet incessantly about topics I don't care about; that's fine, I'll just not follow them (or I'll block them). I'm more concerned about accounts that are purely links to Viagra or porn sites.

I think the best solution for those is a link on the individual's page itself, to report them to Twitter HQ as a spammer.

I couldn't care less about twitter spam, its just one of those irritants that goes along with using the internet, whether you use Ebay, amazon whatever, you are gonna get spam, way to much real stuff in life to care about than spam.

Add a spam icon to twits

Well, the best thing to do is, as mentioned, add a "report spam" button that takes into account both how many people flagged it as spam versus how many people didn't. It should also factor in frequency of posting, frequency of non spam posting and a few more aspects like that. Then after a buildup, the account is put on notice and allowed to "redeem" itself by not getting flagged. thus its like adding water to a glass bit by bit and very slowly allowing for it to go up and down until it overflows, if it does at all. Make sense?

Obviously, whatever system is there now, it isn't working properly - my account affordwealth was suspended and no reason is given.

In my experience a lot of people would report as spam or block messages that include political statements they don't like or in general, opinions they don't like.

On the top of that, as it seems, Twitter suspends accounts, acting on the assumption "guilty until proven innocent", which is essentially undemocratic.

I created my present account only to be able to complain about the suspension of "affordwealth", because I can't post otherwise.

Very disappointing - just started with Twitter and already have problems.

i think a block is all you need. I follow people I don't know all the time. Not because I want to spam them but because I want to hear what they have to say. This is what twitter is all about, collective voices. If you don't want to hear someone then do not follow them. Simple as that. I think you can do a nice remove follow button that sends a friendly message like "I'm grateful you wanted to follow me but I have too many followers at the moment" and then it removes them. I also want to say that spam is subjective. twitter is not a niche site. It has varying topics and to one person a twit could be spam but another might welcome the information. As stated above political or religious topics can be considered spam. I have been on twitter since it was new and have not had a problem with spam at all. I just don't follow unnecessarily.

modemlooper: But what about the ones that are obviously spammers? In the past 4 days this same boob-job surgery clinic has started following me on 3 different accounts - the second one started when the 1st was suspended and just now the third one since the second has been suspended. I;m a guy - I don't want a boob job.

When you say that "spam is subjective", I really wonder if you have thought it through? Spam is defined not by the content but by the method - sending the unsolicited message to multitudes of people, most of whom couldn't care less about the product or service.

Do you really think that it is OK to annoy the vast majority just to catch the eye of that tiny fraction of a percent who might want to buy what you are selling?

As for discrestion in whom you following, that only stops you from getting getting spam-tweets they send. But spam-bots adding you to the list of who they are following, you get the email saying somebody new is following you. You check out their twitter home page but their nick is unobvious so to figure out who it is (in case it is a fried or acquaintance) you need to go to the website given on their twitter page. Voila! They have just gotten you to see their spammy website!

Do you really think it''s OK for spammers to set up twitter-bots that do that?

No offense intended, but your position comes across as a little naïve.

I'd definitely like to see a "report spam" button coupled with a backend statistical algorithm that auto-removes an account when enough people report it as a spammer. This will remove spammers from the system faster and truly allows users to truly be the police of the system

I like the idea of a spam button to push as someone else suggested.

Maybe a simple command of spam followed by the persons user name so you can report through text.

I think it's just that simple. "Flag potential abuse" or "report spam" would work just fine for me, and that should automatically block that user for me as well.

A simple report spam button would be great!

A one touch button to report Spam collect the spammer's info and to block/remove them from my lists all in one shot would be fabulous!

Twitter is 100% opt in, so, it self polices SPAM. I can follow, unfollow someone at will. If you spam, you loose me. Does that make sense?

A Report spam button would be perfect

A report spam button would be ideal. However, it should lead to a brief question regarding why this user / post should be considered spam. For example a brief multiple choice with the top four spam categories plus a "fill in the blank" line.

Question: Should spam be reported on a per tweet basis or reported for a particular user overall?

May I suggest being able to report (and self-