Spotting Spam on Twitter
Posted under: This post was featured at @Twitter_Tips
Twitter is full of Spammers. In the last hour I received up to 21 notifications telling me I had new followers. 17 of those were spammers, 3 of those legitimate accounts and the last one was @SookieBontemps that mentioned my blog post regarding True Blood.
(If you are new to this blog I will tell you one important thing: I like to have a clean followers list. I have declared a war on spammers since my account was wrongfully suspended. I also call them out on Twitter before blocking them using the #nomercy hashtag. @ShellyKramer is doing it as well. If you don’t follow her, you should)
That said here is a crash course on how to identify the new Spam accounts showing up on Twitter.
Meet the Ladies
Spam/Fake Accounts
At first these accounts look legitimate but, if you put them all together can you spot a pattern? I can! (And I love patterns)
The pattern here is that all of this accounts’s user names end with 2 numbers.
Users sometimes put their age or year of birth at the end of their username but can you believe julibo12 is 12 years old? That aval was born in 1955? What kind of name is Abigaiirrh any way?
But let’s take this further shall we?
Following / Followers
Meet britneylo00 (notice the 2 numbers again at the end?)
1,088 followers and 143 following. Better yet, of those 143, 60% are Spam or Fake Accounts.
The same was valid for the other accounts that I checked: A big discrepancy in the Following/Followers ratio.
Ok I know what you are thinking: these are accounts new to Twitter. They are just starting their adventure in the TwitterVerse, they are hungry for information and they just want to make new friends. Right? Wrong!
Newbies that are Geeks
Newbies using API
Remember how I wrote that I love patterns? I found another with one Fake/Spam accounts: They get their Tweets™ mostly from News Feeds (Daily News and Celebrity Gossip seem to be the winners in what comes to delivering content to Spam accounts).
But what is even more interesting? They don’t Tweet™ from the web like a real newbie does, or from DestroyTwitter, HootSuite, Tweetdeck.
They Tweet from the API.
For this you need geek skills. High Level Geek skills. Like my mate @LeoLobato. But Leo usually Tweets™ using TweetDeck.
Doesn’t that make you think?
Conclusion
Accounts that end with numbers + Following/Followers discrepancy + API Tweets™ + (a girl-next-door-photo) = Spam or Fake Accounts = Don’t Follow
Follow me on Twitter and tell me what you thought about this post. Or leave a comment please.
If you are wondering why I am using the ™ everytime I write Tweet™ check this post here
Great post! Very informative. I am curious on how YOUR account got suspended. How did Twitter decide YOU were a spammer? Some of my current followers obviously are not real people yet pass as them every time Twitter sweeps up spammers.
Hey, I think you might be able to use this API against the spammers. What if you wrote a application to check for the signs you describe?
Lets check the API if it’s possible…
There is a function for followers, which returns their screen name, their followers and friends count. so check for 2 numbers at the end and check if (friends / followers) is more than X, finished!
To be even more sure you can fetch a few of their Tweets(TM) and check where they came from, if 90% = API proceed to blocking them, which of course is also a function of the API.
References:
http://apiwiki.twitter.com/Twitter-REST-API-Method%3A-statuses-user_timeline
http://apiwiki.twitter.com/Twitter-REST-API-Method%3A-statuses%C2%A0followers
http://apiwiki.twitter.com/Twitter-REST-API-Method%3A-blocks%C2%A0create
Hello Sugah,
I am also finding a trend in names ending in a lot of numbers … like 1111 or something similar. This is yet another reason I enjoy using socialtoo.com (although its numbers are unreliable, at best). BUT it captures for me, daily (usually) a list of new followers and I can see via a quick glance new followers, their names AND their follow/follower counts. Like you, I am all about patterns (plus, am smart as a veritable whip ) and once I figured that out, it has given me special joy to immediately check out the Spam Suckah, BLOCK them with finality AND fire off a #nomercy Tweet, publicly announcing their loser-ish-ness to the world. These spammers do NOT give us enough credit, Fernando – they really do not. And we shall continue our #nomercy quest and drive them away – maybe all to Facebook!!
I’m nearly finished!
Check my blog for updates on this.
I’m even trying to get Askimet working, but it’s approving my spam since I can’t supply an user-agent ad ip.
[...] This post was Twitted by mita56 [...]
Thanks for this – I got rid of almost 120 from my two Twitter accounts reading this and following your advice
[...] This post was Twitted by arevoredo [...]
[...] This post was Twitted by nonasmanis [...]
this is really a great post,thanks for sharing it,
lately i too notice that the spam are getting really sneaky,i had 10 of new spam followers who had started to tweet at same date on july,28.. just like in your post with girl next door avatar,and today i found out that the are all doing the #followfriday tweet,which they’re list are included another spam and they mixed it with a real ppl,all the 10 just #FFing each other,LOL..so i just blocked them all.
hope this will help us to be aware of them.
thank you
btw,i’m now following you
The hint to the API really is the nail in their coffin. I usually also block them.
[...] on this post I started working on a anti-spam app to block these spammers. My app checks your followers [...]
Re: “They Tweet from the API. For this you need geek skills.” Some of them could simply be using TweetLater. I used that recently and noticed it indicated that my posts were coming from API versus the web. I’m not even clear on what the heck API is… but I know I’m not spamming anyone.
Hey Sugah, this is so crucial and every legit Twitter user needs to pay attention. Spammers follow you to flood you with garbage but more importantly raid your followers and spam them. When you see one, report them and block them immediately. Don’t open any link they provide as it likely traps your IP address and anything else it can find. We’ll also see some pretty severe payloads (bots and other nasties) coming from these “people” in the near future.
This is a war we must win if twitter is going to reach its potential.
Also, you should meet @basebot who is also an antispam warrior!
[...] This post was Twitted by advieira [...]
You are so right. I have spotted these, too. Some copy Tweets from other people, so I would also add that if the Tweeter writes in several different styles (i.e. in good English in one Tweet and in appalling English in another) and sometimes in different languages, they may be a spammer. (Though take note: there may be multilingual people doing the latter.)
I recently spotted a bunch with similar names, all allegedly based in Indonesia, with no avatar, which I also blocked. Those I neglected to wound up Tweeting about some teeth-whitening formula (and were blocked a little later).
While I have not yet suffered a ban (knock on wood) I have been actively deleting these non-human accounts. You just never know when they might cause you trouble.
[...] This post was Twitted by honorman [...]
I know I have a lot of these bastids following me. I need to get in there and purge like a madwoman!
Thank you for this great post.
One recent giveaway sign that I spotted was that a few spam followers were tweeting in HTML! What more can I say but #Fail. This is one end of the spectrum.
On the other end of the spectrum, today, I ‘met’ a bot who retweeted me. When I went to its profile page, I saw that the font of its tweets was in white… on a white background! Apart from links & @usernames which were in black, everything else was completely white. There was no way I could check its tweeting style, its number of followers and so on. Needless to say, I blocked it.
My concern is, whilst some of us may understand some of their trickery, there are still many real people who have absolutely no idea and who would be happy to follow anyone who #ff’ed them or retweeted them, be it animal, mineral or vegetable… or bot.
How to Spot Spam on Twitter | via @fjfonseca | http://bit.ly/1dGMXo
@Briansafetyguy Here you go: How to Spot Spam on Twitter via @fjfonseca http://bit.ly/1dGMXo
RT @ShellyKramer: How to Spot Spam on Twitter | via @fjfonseca | http://bit.ly/1dGMXo
RT @ShellyKramer: How to Spot Spam on Twitter | via @fjfonseca | http://bit.ly/1dGMXo
RT @ShellyKramer How to Spot Spam on Twitter | via @fjfonseca | http://bit.ly/1dGMXo
RT @ShellyKramer How to Spot Spam on Twitter | via @fjfonseca | http://bit.ly/1dGMXo
RT @LoriMoreno: RT @ShellyKramer How to Spot Spam on Twitter | via @fjfonseca | http://bit.ly/1dGMXo
How to Spot Spam on Twitter | via @fjfonseca | http://bit.ly/1dGMXo (cc @anacik ) #nomercy
RT @ShellyKramer: How to Spot Spam on Twitter | via @fjfonseca | http://bit.ly/1dGMXo (cc @anacik ) #nomercy
Spotting Spam on Twitter – http://bit.ly/23VXP1
WOW!
And you can actually make this work? That would be awesomeness!
If that is the case, like users using @YoTwits, TweetLater hould change the message to “from TweetLater” when interacting with Twitter.
Like I say on the last lines of my post, the ultimate factor is to read the timeline and see what that specific user is all about.
I have been getting reports of Spam/Fake accounts that #FF eachother. The war is on to fight spam but the best weapon is our individual judgment.
Thank you for your comment, Bonnie.