4 ms·

I saw this a few days ago, tried it out, and found it pretty useless. (I follow ~100 people). Unfollowing someone because they don't follow you is stupid. In f

I saw this a few days ago, tried it out, and found it pretty useless. (I follow ~100 people). Unfollowing someone because they don't follow you is stupid. In fact, many of the most insightful tweets I see come from accounts that don't follow me, mostly people who are to some degree high profile (@wycats (Merb/now Rails guy), @defunkt (GitHub founder), etc.) If someone's talkative, it's pretty obvious, and I'll either get annoyed by their tweet volume and unfollow them after a time of dense tweets, or they're not bothering me so it's fine. Quiet and inactive are a horrible reasons to unfollow someone, since if they're quiet the cost of following them is basically 0.

zemaj 17y ago
I think it becomes most useful if you have an auto-follow back on your account and 200+ following. It starts to become difficult to get much utility from your twitter stream at that point. I'm interested - what metrics would you want for choosing who to unfollow?
- axod 17y ago
  If they say things that interest you keep following, if they spam/post crap unfollow them? Maybe I have just got completely the wrong idea about this follow/unfollow lark though. I think it's dishonest/useless to follow people if you're not genuinely interested in things they say.
apike 17y ago
I absolutely agree. This tool doesn't help you increase your signal to noise ratio, which should be the real goal. That's why I recently built this: http://www.unladenfollow.com/ http://www.unladenfollow.com/ Unladen Follow suggests people to unfollow based on how noisy they are, not whether they follow you back. It gives you metrics about things like excessive hashtags, links, and so on.
- TobiasCassell 17y ago
  "This is equivalent to following Robert Scoble 21 times." Thank you, this made me laugh out loud..
- jackowayed 17y ago
  That's really, really awesome. One idea, add .@ detection. If they ever use .@'s, I'd automatically give them a score of infinity. (Not really, but it sure is annoying and would be a cool thing to factor in)
  - apike 17y ago
    I considered that - right now they're just penalized as normal @mentions. I'm not sure if ". @steve Going to the grocery store now!" is much worse than "Going to the grocery store now, @steve!" I've also considered extra-penalizing manual "RT @steve", but will wait until all the main clients support it (Tweetie for Mac, for example.)
- johns 17y ago
  I love this tool. A couple points of feedback. I don't mind @mentions if they're sent to people I'm not following since I don't see them. Also, it would be nice if you included the table headers every 50 rows because I can't remember what each column is once they're out of view. Great tool though!
  - apike 17y ago
    Hey John, sorry I missed this comment when you first posted it. @mentions are not counted if they are @replies for the reason you suggest. The repetition of table headers is a good idea, though, thanks!

zemaj 17y ago

I think it becomes most useful if you have an auto-follow back on your account and 200+ following. It starts to become difficult to get much utility from your twitter stream at that point. I'm interested - what metrics would you want for choosing who to unfollow?

axod 17y ago
If they say things that interest you keep following, if they spam/post crap unfollow them? Maybe I have just got completely the wrong idea about this follow/unfollow lark though. I think it's dishonest/useless to follow people if you're not genuinely interested in things they say.

apike 17y ago

I absolutely agree. This tool doesn't help you increase your signal to noise ratio, which should be the real goal. That's why I recently built this: http://www.unladenfollow.com/ http://www.unladenfollow.com/ Unladen Follow suggests people to unfollow based on how noisy they are, not whether they follow you back. It gives you metrics about things like excessive hashtags, links, and so on.

TobiasCassell 17y ago
"This is equivalent to following Robert Scoble 21 times." Thank you, this made me laugh out loud..
jackowayed 17y ago
That's really, really awesome. One idea, add .@ detection. If they ever use .@'s, I'd automatically give them a score of infinity. (Not really, but it sure is annoying and would be a cool thing to factor in)
- apike 17y ago
  I considered that - right now they're just penalized as normal @mentions. I'm not sure if ". @steve Going to the grocery store now!" is much worse than "Going to the grocery store now, @steve!" I've also considered extra-penalizing manual "RT @steve", but will wait until all the main clients support it (Tweetie for Mac, for example.)
johns 17y ago
I love this tool. A couple points of feedback. I don't mind @mentions if they're sent to people I'm not following since I don't see them. Also, it would be nice if you included the table headers every 50 rows because I can't remember what each column is once they're out of view. Great tool though!
- apike 17y ago
  Hey John, sorry I missed this comment when you first posted it. @mentions are not counted if they are @replies for the reason you suggest. The repetition of table headers is a good idea, though, thanks!