James Governor is right to highlight Asymmetrical Follow as a core social software design pattern.
Asymmetric Follow is a core pattern for Web 2.0, in which a social
network user can have many people following them without a need for
reciprocity. Asymmetric Follow is unlike email for example, which tends
to be within small groups, with all users knowing each other
(newsletters are a clear exception here). If you see a social network
where someone has 5000 followers and only follows 150 back - that’s
Asymmetric Follow.
His post explores the concept I've described as pull attention management vs. push (email), and gets enterprisey with @jobsworth:
...The technical approach that is most appropriate to support
Asymmetrical Follow is well known in the world of high scale enterprise
messaging- its called Publish And Subscribe.
But Publish and Subscribe also talks to the social aspects of Asymmetrical Follow. As @Jobsworth puts it:
“I think we would be far better off considering Twitter
as neither Pull nor Push, but instead as Pub-Sub, as Publish-Subscribe.
The first and most beautiful thing about Twitter, as far as I am
concerned, is that I only see the tweets of people I follow, people
whose tweets I subscribe to. It is up to me to decide how many people I
can follow. For some people this may be a Dunbar number, stabilising
around 150, perhaps finding a Twitter adjustment to that number and
raising it. Others may be Scoblesque in their reach, dissatisfied
unless they push the 5000 limit (as in Facebook; I must admit I have no
idea what the Twitter limit is).
So one way of avoiding increasing noise levels is to avoid
increasing the network beyond one’s capacity. I can choose to “follow”
(or subscribe to tweets from) just as many people as I am able to cope
with. This is not something you can do easily with BlackBerry or with
e-mail in general.”
James explored further in his next post, getting Matt Biddulph to comment on if this pattern arose in Dopplr.

This graph shows the distribution of how many people
share with more people than share with them, or vice versa. The hump is
slightly higher on the right, indicating that there’s a bias towards
sharing with more people than share back.
The peak is centred on zero, indicating that the average traveller
shares with about the same number of people that share with them.
I think the hump is on the right because Dopplr is private by default, while Twitter is public by default. Defaults matter for social dynamics than any other designer intent.
With a default-private network, you share as a gesture, perhaps in
hopes of getting sharing back to form a confirmed tie. With a
default-public network, you share by default, and you follow as a
commitment of attention. With default-public, the hump is on the left,
in the form of a power law. Both networks follow the Asymmetric Follow
pattern, but they have different defaults reflecting the nature of the
objects shared and how people socialized around them.
Socialtext also employs Asymmetric Follow, where people follow other people's Update Feeds to subscribe to events like edits, posts, comments and tags. I actually think Asymmetric Follow emerged as a pattern designed for adoption, when you consider the alternative -- requiring a confirmed tie before sharing information.
One thing I've found fascinating about how this pattern works out in an organization, with its pre-existing social network and context, is what you can learn about someone from how attention is spent by them and to them.
To the right is how this feature shows up on a Profile in Socialtext People. Unlike Twitter, we needed to make it visually easy to discover who is following who (Adina would describe the design decisions better than me). What I can say is one of the greatest values is on-boarding new employees. Being a new employee is joining a new social network and the greatest challenge is understanding how information actually flows and how work actually gets done. By having a default-public network like this in place, on-boarding is accelerated and the strength of weak ties is leveraged.
Side note: yes, this should have been two blog posts
Recent Comments