[Air-L] Who to UnFollow

Dave Karpf davekarpf at gmail.com
Wed Jul 1 10:55:51 PDT 2009

Hi folks,

I'm a recent addition to the list... couple thoughts to add into the mix:

First, Barry's 500 followers seem entirely sensible, given his status as a
known author.  I would postulate that there are four main avenues for
picking up followers on twitter.

(1) address book search.  If I recall correctly, this is integrated into the
system, such that when a new user joins twitter, it offers to search their
address book and give them the option of following any matching profiles.
That set of users will approximate one's social network and/or facebook
friends list (perhaps only loosely so).

(2) famous names.  Extrapolating from personal experience, I know when I
joined twitter I looked over the "following" list of a close colleague with
overlapping research interests to see who I should add.  Any author whose
book or article I had found particularly interesting was immediately added.
This set of users are where I derive much of the added value from twitter,
personally.  Without them, twitter isn't *much* different from facebook
status updates.  With them, I can learn about interesting articles and
emerging topics beyond my network of close colleagues.

(3) retweets/in-tweet mentions.  I pick up a few additional followers after
every tech conference, and they are often people who aren't attending the
conference.  This is because I'll be mentioned or retweeted by a colleague
with a much larger following.  The retweet mechanism is probably the most
important for modeling/understanding the development of twitter norms, I
think.  Those who solely tweet about their lunch will virtually never be
retweeted.  Those who offer humor, valuable links, or 140-character insight
will be.  Humor and insight, in particular, likely encourage additional
followers.  Institutions like #followfriday are an outgrowth from this

(4) offline mentions.  If Barry includes his twitter handle on the first
slide of a conference presentation, he likely sees a few immediate adds.
When CNN, Oprah, etc announce their twitter handle to a mass audience, that
both spurs an increase in twitter accounts and a surge in followers.  I
personally haven't partaken in these sites, and have little to no sense of
what to make of it.

There's a fifth category, which includes spambots, advertisers, etc, but I
think we can safely set that aside as noise for the moment.

Another dynamic to consider in any modeling exercise (and one which will be
a HUGE pain for any empirical studies) is variance in platform.  I use the
iPhone app "tweety."  Barry uses tweetdeck.  It sounds like these platforms
have different functionalities, and the growth of twitter will only
encourage further code-based innovations of this sort.  Compare those to
twitter-via-laptop or -desktop and you have a completely different user
experience.  I never used twitter until I downloaded tweetie, because I
found it a less-appealing distraction than facebook and a couple of favored
blogs and discussion boards.  With the mobile client, I now check twitter
while waiting in line for coffee or sitting at a red light.  Mobility and
platform have a huge impact on how I'm experiencing the medium, and that in
turn shapes my normative opinions about how people should and should not use
the medium.

Very interesting topic, thanks for raising it!
-Dave Karpf

Dave Karpf
PhD, Political Science
University of Pennsylvania

Postdoctoral Research Associate
Taubman Center for Public Policy
Brown University

davekarpf at gmail.com

More information about the Air-L mailing list