1. This site uses cookies. By continuing to use this site, you are agreeing to our use of cookies. Learn More.

Testing aggressive following &/or follower churn, evading suspension triggers.

Discussion in 'Twitter' started by StealthLeads, Oct 24, 2016.

  1. StealthLeads

    StealthLeads Newbie

    Joined:
    Aug 16, 2016
    Messages:
    38
    Likes Received:
    11
    Gender:
    Male
    Hey all,

    So my bot has excellent content flow and engagement. It's growing by itself but I'm ready to speed it up a little with some churn. I'm about ready to tune my follower churn bot and I thought it might be productive to post here with the information I gather. I have hopes that others will add information if this is found useful to them. Let's save each other a few banned accounts?

    Here is the information I've gathered pre-testing. It seems like there are three automatic banhammer flags we will need to avoid.

    1. Automation Rules
    • You may not use or develop any application that allows for the following or unfollowing of user accounts in a bulk or automated manner.
      • So first of all to do this we will have to mimic manual following / unfollowing. Since they cannot see our front end with personal code, all they can do is check the bots activity. This means we have to be as human-like as possible. No bulk processing at fast speeds. My theory on this is to have multiple work sessions per day, every few hours or so on a semi-random interval. During each work session follow a small group of users with short but randomized timing between each follow request. Something along those lines..
    2. Aggressive Following​
    • We monitor all accounts for aggressive following and follow churn (repeatedly following and unfollowing large numbers of other users). You can read more about these below, but if you don’t follow or unfollow hundreds of users in a single day, and you aren’t using automated methods of following users, you should be fine.
    • Aggressive following is defined as indiscriminately following hundreds of accounts just to garner attention. However, following a few users if their accounts seem interesting is normal and is not considered aggressive.
      • Key thing to note here is they mentioned "Hundreds" twice. Meaning it's likely that their flags will not trigger for accounts that follow less than 100 users per day. I will probably start with somewhere around 75-90 per day every day for safe sustainability. It's worth noting that you can probably get away with much more if it isn't sustained. You can get away with following under 1,000 people in a day, but probably not doing it daily. My goal is to boost growth, but keep it subtle enough that I can get away with it 24/7.
    3. Aggressive Follow Churn
    • Aggressive follow churn is when an account repeatedly follows and then unfollows a large number of users.
      • This one might be the trickiest. The simplest fix is just don't churn. The problem is that relies on natural growth and a good follow ratio. Personally, I plan to start by monitoring my follower ratio and only following when my ratio allows it. This is the safest and most sustainable way of continuously building up an authority account. If the account does not grow enough naturally some unfollowing may be required. In previous tests I have managed to unfollow 300-500 people once every 3-5 days without issue. I plan to test spreading this out as gradual daily unfollowing in small human-like amounts (like following). I also think it would be wise to mix in unfollowing some followers with all the unfollowers we are removing.
    That's about all I've got for you guys. Hopefully this post helps out a few fellow botters.

    Let me know if you guys find anything worth mentioning, specifically any follow/churn rates that are known to trigger suspensions. If you get banned from churn, it's very helpful to let us know how fast you were churning. Similarly your follow rates for non-banned but actively churning accounts.

    Good luck friends. :cool:
     
    • Thanks Thanks x 1
  2. girihanmandla89

    girihanmandla89 Newbie

    Joined:
    Apr 16, 2016
    Messages:
    1
    Likes Received:
    0
    Awesome i do follow u from today ...
     
  3. WanderingSoul

    WanderingSoul Newbie

    Joined:
    Oct 24, 2016
    Messages:
    5
    Likes Received:
    1
    Gender:
    Female
    These are great thoughts! Would love to hear about your results!
     
  4. karupoiss

    karupoiss Elite Member

    Joined:
    May 16, 2012
    Messages:
    1,869
    Likes Received:
    594
    Gender:
    Male
    Location:
    Europe
    How did the tests go?
     
  5. BufferOver

    BufferOver Power Member

    Joined:
    Mar 4, 2009
    Messages:
    513
    Likes Received:
    40
    Will keep my eyes on this one.. good luck!
     
  6. Olovarno

    Olovarno Registered Member

    Joined:
    Jan 19, 2010
    Messages:
    90
    Likes Received:
    12
    I've used CrowdFireApp to follow/unfollow users in 8 old accounts (2+ years old, 10K+ followers). What I found out until now:
    1. I can follow up to 120 people in a blink (less than 200 milisseconds between two follows!), but later I need wait for 60 seconds or more before continuing;
    2. I can follow up to 1K people daily, but I don't consider it safe - I prefer to follow 700-850 in each account;
    3. I can unfollow thousands, so I guess they aren't considering tough constraints about unfollowing.
     
    • Thanks Thanks x 1
  7. StealthLeads

    StealthLeads Newbie

    Joined:
    Aug 16, 2016
    Messages:
    38
    Likes Received:
    11
    Gender:
    Male
    Forgot I made this thread lol.

    I made a couple mistakes in testing this. Here's the results I have so far.

    So basically I setup auto following to start and let that trickle in. I didn't hide timing pattern very well, just did like 1 follow every couple hours. (mistake 1)

    I was also scraping competitors follows to add to a to-follow list from the same app / account that was doing the following. (mistake 2)

    I was feeling confident and decided to test it on my 2010 automated content account that is getting ~25,000 impressions/day with only a few hundred followers. (Mistake 3)

    So when twitter saw the scaping &or following going on I hit an automatic flag. One of those ones where you use your email or phone to unlock it. Problem it, this was the second lock and Twitter would only let me use my phone number. Now.. naturally.. my phone number is banned from twitter after my first account banwave.. (RIP 11 replybots..). So i had used a temporary phone number to verify the accounts apps... I got very lucky and support let me unlock it with my email address with a blunt warning about having "accidentally triggered an automatic ban". I think the 2010 creation date + thousands of content only tweets helped.

    Since then I've been focusing on getting impressions without needing followers so.. I haven't cracked aggressive following yet. The auto account locking they implemented has taken things up a notch. I do plan to setup a group of new test accounts eventually. Will post again if I get anything down the road.
     
    • Thanks Thanks x 1
  8. StealthLeads

    StealthLeads Newbie

    Joined:
    Aug 16, 2016
    Messages:
    38
    Likes Received:
    11
    Gender:
    Male
    Very useful info, thanks for posting this. I will have to play with these rates when tinkering on my bots.
     
  9. SharpConversions

    SharpConversions Junior Member

    Joined:
    Jun 30, 2015
    Messages:
    109
    Likes Received:
    8
    Gender:
    Male
    No offense but automatic following is a complete waste of time anyway, all you're going to be doing is following bot accounts. And you can't scrape usernames and follow people that way because that will just trigger a ban.
     
  10. StealthLeads

    StealthLeads Newbie

    Joined:
    Aug 16, 2016
    Messages:
    38
    Likes Received:
    11
    Gender:
    Male
    No offense taken! I've been starting to agree with you. I'm having more success ignoring the followers entirely and just focusing on showing up in keyword searches. I'm getting curious how much follower count really matters.

    I do however think you could probably successfully automate follower churn if you can figure out how to scrape for user targets successfully.
     
    • Thanks Thanks x 1
  11. Turnip832

    Turnip832 Newbie

    Joined:
    Nov 29, 2014
    Messages:
    35
    Likes Received:
    7
    I think it's fairly simple, find a competitor's followers page, follow 1000 people from that. Wait a couple of days and unfollow all that didn't follow back. Rinse and repeat.
    In January last year I had about 9000 impressions. In January so far this year (i.e. 9-10 days) I've got 400k already! :)
     
    • Thanks Thanks x 1
  12. StealthLeads

    StealthLeads Newbie

    Joined:
    Aug 16, 2016
    Messages:
    38
    Likes Received:
    11
    Gender:
    Male

    Any chance I can get your follower count?

    I have:

    Followers: 640

    Impressions this year: 245K
    Tweets this year: 13.1K
    Profile Visits This Year: 782
    New Followers This year: 30
     
  13. Turnip832

    Turnip832 Newbie

    Joined:
    Nov 29, 2014
    Messages:
    35
    Likes Received:
    7
    Sure, 23.2k at present.

    Profile visits Last year ~162k
    Generally couple of thousand new followers a month apart from the last 3 months or so when I haven't been as active.
    Normally around 200 tweets per month (+/- 100).
     
    • Thanks Thanks x 1
  14. StealthLeads

    StealthLeads Newbie

    Joined:
    Aug 16, 2016
    Messages:
    38
    Likes Received:
    11
    Gender:
    Male

    This is actually pretty interesting, though hard to compare.

    I'm posting 20,000% more tweets than you. (seriously)
    I'm getting around 61% of the impressions that you get.
    I have about 2.5% of the followers you do.
     
  15. Turnip832

    Turnip832 Newbie

    Joined:
    Nov 29, 2014
    Messages:
    35
    Likes Received:
    7
    I'm not sure as well, the reach I described was the last 2 weeks or so whereas those other stats were over the last 12 months. I'm not sure whether your stats meant last 12 months or not. I think my reach the past 12 months would be in the region of maybe 8 million.

     
  16. StealthLeads

    StealthLeads Newbie

    Joined:
    Aug 16, 2016
    Messages:
    38
    Likes Received:
    11
    Gender:
    Male
    I was comparing to your stats from this year so far. Your 400k vs my 245k. And I post about 40,000 tweets per month.
     
  17. fullwall3t

    fullwall3t Power Member

    Joined:
    Apr 22, 2015
    Messages:
    713
    Likes Received:
    77
    How many different links and post in those 40k?
     
  18. StealthLeads

    StealthLeads Newbie

    Joined:
    Aug 16, 2016
    Messages:
    38
    Likes Received:
    11
    Gender:
    Male
    Currently 0 links pure content tweets 100% image posts with 1 tag. Still tuning it for impressions and growth, so I haven't monetized it yet.

    However, in the past I got away with 1 link + mention every 4th tweet at the same posting rate.
     
  19. Turnip832

    Turnip832 Newbie

    Joined:
    Nov 29, 2014
    Messages:
    35
    Likes Received:
    7
    Haha that's crazy! 40000 per month! I'm at a tiny fraction of that lol. Is it all on one account?
     
  20. Peachpies

    Peachpies Newbie

    Joined:
    Apr 11, 2015
    Messages:
    21
    Likes Received:
    21
    Trying to figure out some limits myself.
    Currently increasing the follows by 5 everyday, until it hits 100 per day. Accounts are all aged (2 months~).
    Have a test account in which I favorite, retweet, post, post with link and follow over 100 per day along with a DM. So far it is not banned. Will try incrementally increasing until I find a sweet spot.
    Quick question: Am I right to assume that my link will get flagged if one of my accounts gets suspended?