paper.li | Shatterbox

Twitter for Lawyers: Not All Tweets Are Created Equal

October 13, 2011

There is no lack of tools that help you decide who to follow and unfollow on Twitter by examining the activity of accounts in your Twitter ecosystem.

One of the current “buzz” Twitter clients, Twit Cleaner, provides users a very interesting detailed breakdown of how it identifies potential “garbage” accounts among people, organizations and bots you follow on Twitter.

In general, tweets that demonstrate active engagement (e.g. “@” messages, retweets and links) are favored, while automated tweets (e.g. paper.li distributions, Foursquare check-ins, RSS bot feeds) count against you if used extensively.

After you request an analysis of your Twitter followers, Twit Cleaner sends you a Twitter direct message containing a link to your customized results. In addition to the statistical breakdown by “type,” the analysis includes thumbnail avatars of all the Tweeps in each category, which yield individual details as you mouse over each with your cursor.

To cull out the undesirables, just click on the corresponding avatars.

Here’s how the categories break down:

Potentially Dodgy Behavior

App spam – More than 50 percent of tweets are auto-generated messages, aka “app spam” (e.g. paper.li, Foursquare, blip.fm)
Uses advertising networks
Nothing but links – Few retweets, no “@” messages
Repeating the same URLs – Duplicates the same link more than 25 percent of the time
Posting identical tweets

Other Dodgy Behavior, Now Absent

No Activity in Over a Month

Not Much Interaction – Fewer than 10 tweets

Not active yet
Don’t interact with anyone – No “@” messages or retweets
Bots – M0re than 90 percent of tweets pumped out from an RSS feed
Hardly follow anyone – People who follow back fewer than 10 percent of those who follow them

All Talk, All the Time – Averages more than 24 tweets a day (excluding @ replies and direct messages)

Little Original Content – Retweets are 70 percent or more of total output