The #FollowAugust experiment explained.
When my son turned 13, he got a Twitter account. When he asked me who I follow, I said, “Everybody who follows me and then some.” But when I looked more closely at my Twitter profile, I realized that my follower list was not appropriate for a 13-year-old’s eyes. By virtue of following back those who followed me, I ended up following bots, spammers, and/or porn accounts.
I felt dirty.
On July 1, 2009, I had about 27,000 followers on Twitter and was following 24,000. That day, I launched the #FollowAugust experiment where I started unfollowing people at the rate of 1000 per day. My goal was not only to get rid of the junk but also to only follow those accounts that I actually follow. Today, nearly months after starting the experiment, I follow 263 and have about 15,000 followers.
During the experiment, I locked my account (to prevent new followers) and greatly scaled back my tweeting. Most of my tweets were Google Reader shared items published via FriendFeed. I have a <href=”https://www.giantpeople.com/1090.html”>drawing that explains my social networking setup.
Here are my Twitter stats during the #FollowAugust experiment.
Date | Following | Followers | Tweets | DMs | Follow Requests | @grader Score | @twitterank Score |
07/01/09 | 24000 | 26978 | 5234 | 8973 | 27 | 100.00 | |
07/02/09 | 23000 | 26530 | 5275 | 8975 | 76 | 99.98 | |
07/03/09 | 22000 | 26131 | 5318 | 8976 | 125 | 99.96 | |
07/04/09 | 21000 | 25666 | 5353 | 8980 | 162 | 99.95 | |
07/05/09 | 20000 | 25235 | 5356 | 8980 | 187 | 99.80 | |
07/06/09 | 19000 | 24800 | 5364 | 8980 | 205 | 99.80 | |
07/07/09 | 18000 | 24332 | 5382 | 8977 | 225 | 99.90 | |
07/08/09 | 17000 | 23928 | 5409 | 8971 | 254 | 99.90 | |
07/09/09 | 16000 | 23555 | 5438 | 8980 | 272 | 99.80 | 99.77 |
07/10/09 | 15000 | 23157 | 5462 | 8974 | 300 | 99.97 | 99.72 |
07/11/09 | 14000 | 22724 | 5488 | 8974 | 325 | 99.95 | 99.72 |
07/12/09 | 13000 | 22366 | 5497 | 8981 | 351 | 99.97 | 99.63 |
07/13/09 | 12000 | 21966 | 5512 | 8982 | 364 | 99.80 | 99.38 |
07/14/09 | 11000 | 21566 | 5530 | 8982 | 402 | 99.80 | 98.79 |
07/15/09 | 10000 | 21163 | 5548 | 8985 | 412 | 99.80 | 98.79 |
07/16/09 | 9000 | 20741 | 5565 | 8985 | 441 | 99.80 | 98.39 |
07/17/09 | 8000 | 20358 | 5571 | 8985 | 467 | 99.80 | 98.39 |
07/18/09 | 7000 | 19921 | 5581 | 8984 | 494 | 99.80 | 98.28 |
07/19/09 | 6000 | 19448 | 5582 | 8984 | 521 | 99.80 | 96.92 |
07/20/09 | 5000 | 19120 | 5589 | 8984 | 533 | 99.80 | 96.92 |
07/21/09 | 4000 | 18738 | 5603 | 8983 | 559 | 99.90 | 95.08 |
07/22/09 | 3000 | 18401 | 5616 | 8984 | 574 | 99.80 | 95.08 |
07/23/09 | 2000 | 18065 | 5633 | 8984 | 594 | 99.60 | 92.32 |
07/24/09 | 1000 | 16935 | 5643 | 8984 | 612 | 99.60 | 92.32 |
08/09/09 | 1015 | 15643 | 5650 | 8987 | 796 | 99.80 | 70.41 |
08/10/09 | 5 | 15602 | 5655 | 8985 | 801 | 99.80 | 70.41 |
08/10/09 | 0 | 15523 | 5656 | 8985 | 802 | 99.80 | 70.41 |
08/10/09 | 243 | 16224 | 5657 | 8985 | 0 | 99.80 | 70.41 |
08/27/09 | 263 | 15316 | 5771 | 8986 | 0 | 99.60 | 93.46 |
Change | -23737, -99% | -11,662, -43% | +537, 9.25/day | +13, 0.2/day | +802, 30/day | -0.40 | -29.36 |
A few things jump out at me. Both @grader and @twitterank value high following counts. While I was on vacation, I did not tweet, and it’s clear that @twitterank also values constant activity. It also appears that most of the people who queued up to follow me (while my account was locked) promptly unfollowed me after I failed to follow back.
The bottom line: after unfollowing everybody and refollowing a select few, I lost 43% of my followers. But I kept 57%, so the glass is (more than) half full.
True, I have lost some of my Twitter “authority,” but I also feel like I have regained my Twitter virginity. At the very least, I can show the list of folks I’m following to my kids without fear of embarrassment.
Related Articles
- 2009-07-04 – How I Tweet: Just the FAQs (interview with Guy Kawasaki)
- 2009-07-03 – Should Twitter Crack Down on Pornography?
- 2009-07-03 – 8 Talking Points On Twitter Follower-Building Tools
- 2009-07-02 – Marketers Call For Twitter To Filter Content For Pornography, Spam
- 2009-07-01 – Want More Twitter Followers? Break Out Your Wallet
- 2009-06-30 – Twitter Follower-Building Services – Gain Numbers, Lose Respect?
- 2009-06-30 – http://tweetspinner.com is a tool for Twitter spammers.
- 2009-06-30 – Twitter Revamps Following and Followers Pages
- 2009-06-30 – FriendOrFollow.com Helps Find Snobs And Spammers: These Are Mine [NSFW]
- 2009-06-25 – The Twitter Cycle: Curiosity, Abandonment, Addiction. Global Visitors Hit 37 Million.
- 2009-06-25 – 8 Ways To Find Relevant Followers On Twitter
- 2009-06-19 – Dear Wanna-bes, Your Twitter Stardom is Coming to an End
- 2009-06-01 – New Twitter Research: Men Follow Men and Nobody Tweets
- 2008-10-23 – How ‘Follower Spam’ Infiltrated Twitter – and How to Stop It
Erik J. Heels writes about technology, law, baseball, and rock ‘n’ roll. You should follow @ErikJHeels on Twitter.
Good heavens. That’s some clicking you must have done!
Ian
Greetings Ian,
Yes, I unfollowed 24,000 people on Twitter manually. At the rate of 1000/day. Took some time to unscramble the eggs. But I think I’m in a better position now.
Regards,
Erik
Interesting experiment Erik. I’ve considered doing something similar as I have all sorts of “unsuitables” following me.
How did you do the unfollow – manually?
ian
OFF TOPIC
Erik,
I’m a 1L. I can’t seem to find a good law student “chat room” Any recommendations?
You’re an interesting guy!