5 ms·

To add to LifeShield's shady practices, it looks like most of their Twitter followers are spam accounts that they've created to boost their followers: https://t

To add to LifeShield's shady practices, it looks like most of their Twitter followers are spam accounts that they've created to boost their followers: https://twitter.com/#!/lshomesecurity/followers https://twitter.com/#!/lshomesecurity/followers Every follower has no bio, around 10-15 "follows" and about 1,500-1,800 followers. All generic photos, generic usernames, names, etc. What a fucking joke this company is.

epoxyhockey 14y ago
You are correct, sir: http://i.imgur.com/Ytf9b.png http://i.imgur.com/Ytf9b.png EDIT: The flat lines are the fake followers..
- nicksergeant 14y ago
  Ha, awesome.
- dakotasmith 14y ago
  Mind sharing the code that generated this? I am not the only person on HN who might like to have a way to examine real/presumedfake follower ratios for any number of reasons.
  - epoxyhockey 14y ago
    I would setup a webservice for it, but Twitter's API only allows 300 queries per hour. Just get the list of follower IDs via the API, sort them, then plot the deltas between each ID. Most Twitter follower networks registered thousands of Twitter accounts in a quick, automated fashion, making their entire list of Twitter ID's almost sequential. The plot shows the deltas of the Twitter IDs from 0 to the highest ID. What's even more amusing is that, once you've identified a network, you can deduce the entire list of their clients by what users the network already follows. Part of me wonders why Twitter doesn't just nuke the follower networks, since they are so easy to identify. But, on the other hand, it's almost comical looking at no-name Twitter users with tens of thousands of followers. There is no way to force a Twitter network to unfollow them, so they are stuck with the fake followers and all of the embarrassment that comes along with it. Though, the scary part is that one can attack an adversary by buying up 25K Twitter followers for them for a pretty insignificant price (maybe $100).
    - zedadex 14y ago
      > Though, the scary part is that one can attack an adversary by buying up 25K Twitter followers for them for a pretty insignificant price (maybe $100). If that kind of attack became well-known, the next logical step would be that a smart company buys some fake followers, then protests against it as though to imply that one of their competitors did so. Slandering a rival while reaping additional followers (though fake...) seems like a win-win.
RenierZA 14y ago
Using Google's search by image I found every follower's picture I tested to be fake. One of them is even Miss World from 1994: https://twitter.com/#!/Darbyko3 https://twitter.com/#!/Darbyko3 https://twitter.com/#!/aishwaryarai https://twitter.com/#!/aishwaryarai
droithomme 14y ago
Wow, thanks Nick, that is an amazing note about the fake Twitter accounts. The fast majority of them have profiles photos of attractive young women and teens and are otherwise indistinguishable as you point out. Except there is a high incidence of very unusual names, such as if the fake accounts were created using a seeding list of all names without distinction to actual usage frequency. Also lots of ethnicity mismatches between surnames and photos. Another tell of course is its usually older people who buy home security packages, and most followers are teen girls and young adult women. This is so shady, and so interesting that it is a "security company" doing this, one has to wonder what else they are doing.
mparlane 14y ago
So basically, Lifeshield are paying someone to do really dodgy SEO and twitEO. They did their Google SEO then got fu*ked by it, so they started a dmca take down of all the left over sites that had fake link backs. Now they just need their twitter follower network shutdown.

epoxyhockey 14y ago

You are correct, sir: http://i.imgur.com/Ytf9b.png http://i.imgur.com/Ytf9b.png EDIT: The flat lines are the fake followers..

nicksergeant 14y ago
Ha, awesome.
dakotasmith 14y ago
Mind sharing the code that generated this? I am not the only person on HN who might like to have a way to examine real/presumedfake follower ratios for any number of reasons.
- epoxyhockey 14y ago
  I would setup a webservice for it, but Twitter's API only allows 300 queries per hour. Just get the list of follower IDs via the API, sort them, then plot the deltas between each ID. Most Twitter follower networks registered thousands of Twitter accounts in a quick, automated fashion, making their entire list of Twitter ID's almost sequential. The plot shows the deltas of the Twitter IDs from 0 to the highest ID. What's even more amusing is that, once you've identified a network, you can deduce the entire list of their clients by what users the network already follows. Part of me wonders why Twitter doesn't just nuke the follower networks, since they are so easy to identify. But, on the other hand, it's almost comical looking at no-name Twitter users with tens of thousands of followers. There is no way to force a Twitter network to unfollow them, so they are stuck with the fake followers and all of the embarrassment that comes along with it. Though, the scary part is that one can attack an adversary by buying up 25K Twitter followers for them for a pretty insignificant price (maybe $100).
  - zedadex 14y ago
    > Though, the scary part is that one can attack an adversary by buying up 25K Twitter followers for them for a pretty insignificant price (maybe $100). If that kind of attack became well-known, the next logical step would be that a smart company buys some fake followers, then protests against it as though to imply that one of their competitors did so. Slandering a rival while reaping additional followers (though fake...) seems like a win-win.

RenierZA 14y ago

Using Google's search by image I found every follower's picture I tested to be fake. One of them is even Miss World from 1994: https://twitter.com/#!/Darbyko3 https://twitter.com/#!/Darbyko3 https://twitter.com/#!/aishwaryarai https://twitter.com/#!/aishwaryarai

droithomme 14y ago

Wow, thanks Nick, that is an amazing note about the fake Twitter accounts. The fast majority of them have profiles photos of attractive young women and teens and are otherwise indistinguishable as you point out. Except there is a high incidence of very unusual names, such as if the fake accounts were created using a seeding list of all names without distinction to actual usage frequency. Also lots of ethnicity mismatches between surnames and photos. Another tell of course is its usually older people who buy home security packages, and most followers are teen girls and young adult women. This is so shady, and so interesting that it is a "security company" doing this, one has to wonder what else they are doing.

mparlane 14y ago

So basically, Lifeshield are paying someone to do really dodgy SEO and twitEO. They did their Google SEO then got fu*ked by it, so they started a dmca take down of all the left over sites that had fake link backs. Now they just need their twitter follower network shutdown.