Community Startup Plan
Mailbox #40
April 19th, 2008
JonathanGCohen@gmail.com
News and Views for Organizations
A Web Community Proposal for the School of Information

Introduction
Communities need something to talk about. If people are offered appealing content, a non-threatening place for discussion, and sufficient rewards for contribution, they’ll probably show up and participate.
User contributions at social news sites like Digg– such as links to content at other sites – are easy to create. By casting votes on these outgoing links, members filter out high-quality information.
For some communities, motivating and rewarding interaction may yield greater or equal benefits than rewarding valuable information. When community members interact, organizational stakeholders with a financial interest benefit from an increase in member investment, retention, and support. Members benefit from networking, developing relationships, and a sense of camaraderie.

Purpose and Intended Audience
My community idea is a link submission and discussion platform for organizations. Members would submit outgoing links to original content; all links would be stored and could be browsed by other members. Whereas social news voting sites only direct members to their most popular outgoing links, this platform would highlight the most popular discussions surrounding outgoing links. People in an organization may be just as interested in what most people are currently talking about as valuable information. Comments and outgoing links would receive equal visual prominence on the landing page, appearing in the same main content field. The purpose is to encourage interaction among members of an organization, increase knowledge-sharing, valuable social connections, and opportunities to be recognized for meaningful contributions.
I will explore the utility of this community idea for the School of Information (“SI”) at the University of Michigan, though it may have potential use for other organizations. This academic program has an enrollment of approximately 300 masters and 50 PhD students. They share an interest in information technology and interact in courses while pursuing a graduate degree in nine specializations, such as Human-Computer Interaction. My intended audience is all SI students. The primary form of online interaction that spans their enrollment lifecycle is a mailing list.
But mailing lists aren’t perfect. Subscribers may receive messages at inconvenient times, especially if they need to check their email often. They’ll be less likely to read carefully, act on information, and participate in discussions. Maybe another email commands their attention or something else is on their mind. If they use email management software such as Outlook, pop-up alerts might notify them each time they receive a new message. These people often can’t control when messages first seize their attention.
For busy mailing lists, niche items receive too much publicity and newsworthy items don’t receive their fair share. The subject line of an event announcement takes up the same screen estate as the subject line of someone’s pet interest. (Though some email programs let senders denote importance with tags– this convention isn’t standardized across different email management applications). An important unread message is less likely to be discovered in a pile of new correspondence. Further, “importance” is subjective – different topics appeal to different subscribers.
As the membership of a mailing list grows, senders and responders risk more backlash for conversation. Unless subscribers are interested a discussion, they may feel bothered. Considerate members may only may only engage in discussions they perceive having universal appeal. In sum, these problems harm the potential for social connections between members as well as individual utility.
Masters students at SI belong to three mailing lists, which yielded 763 messages in the Winter 2008 semester. si-students-masters and si-all are primarily used for announcements, while si.all.open is reserved for more general discussion. Yet messages on si.all.open are unlikely to start a conversation, and those that do are short-lived.
Mailing List Messages # of Replies % of total replies
si.all.open@umich.edu 446 126 28%
si-students-masters@listserver.itd.umich.edu 91 1 1%
si-all@listserver.itd.umich.edu 226 13 6%

A note on methodology: I searched for messages sent to each mailing list during 1/1/08 – 4/21/08. I then searched for “Re:” in the subject line and removed bad hits. (i.e., a message with re: in the subject title that wasn’t a reply)
Of course, mailing lists have benefits: senders know who will receive their message, wide distribution is instant and simple, and membership can be controlled or open. Mailing lists are great for getting the word out, but big and busy organizations need a different platform for cultivating online discussion. I will discuss how the SI mailing list could coexist with a link discussion community later in this proposal.
The SI student population comes from a wide range of academic, professional, and technical backgrounds. Also, students are less likely to share courses with people outside of their own specialization. Even with sporadic events such as workshops, there are few ongoing ways SI students have to get to know each other better and share what interests them.

Value Proposition

Participants will gain value from the community once it has succeeded from a sense of belonging, self-esteem, and professional development. A network effect exists - the presence of other people is critical for utility – but there are no competing communities exclusively for SI students. Increased interaction among organizational members may lead to more trust and improved team relations, as well as the potential for collaborative partnerships. The restriction of site accounts to organizational members would encourage group identity by defining a social category. (Ren et al.) Bonds between members could be encouraged by making social interaction easy (e.g., employing canned feedback), sharing personal knowledge (e.g., showing which members are currently online), and playing up similarities (e.g., displaying a list of members that awarded canned feedback to the same submission)
People who participate will need to enjoy sharing outgoing links with other members of their organization. Also, they will need to actively find or passively encounter content that compels to other members of their organization, as well as be able to recognize its appeal.
In organizations with sub-groups, the need for universal information that poses a potential threat. Members in neglected sub-groups may feel put out if they can’t find something that relates to them. Their group identity would suffer. One technological feature – tagging – would help people filter out conversations that matter to them. For example, School of Information students could tag outgoing links with specializations (HCI, etc. or All) and/or courses (SI684, etc.)
Members would also want feedback about the value of their link submissions and comments. I would make use of canned feedback to let people easily communicate with each other. Whereas comments allow people to express their thoughts with prose, “canned feedback” is a preexisting term – often a link constructed by community designers - that members can click to communicate a thought, rating, or feeling to another member. The broad interpretation of positive canned feedback enhances intrinsic motivation to contribute because people are free to choose why they were recognized and how well their contribution stacks up. I’ll discuss the design of canned feedback in the next section.
To reach “critical mass”, a core group of frequent contributors will need to contribute a “minimum volume of message traffic to draw and retain members” (Arguello et al). I will use the social news site Digg.com as a base for realistic expectations of community engagement. Digg receives approximately 25 million unique visits per month and has 2.5 million registered user accounts. Let’s assume the average Digg user visits ten times per month. For the sake of this argument, I will assume 50% potential engagement – so 175 members. If this population visits ten times per month, 80% freeride, 15% comment and provide positive canned feedback, and 5% contribute outgoing links to original content, then one can expect approximately 8-9 people contributing outgoing links. Since these contributors would also be visiting only ten times per month, the total number of new outgoing links per day would total about 3, which wouldn’t sustain successful interaction. Growth would be an uphill climb, but I think promotion and incentives would lead to a successful outcome.
Metric Sources:
http://www.quantcast.com/digg.com
http://www.techcrunch.com/2008/01/29/digg-nearly-triples-registered-user...
Growth Plan

Barriers should be kept low to encourage interaction and the contribution of outgoing links. Creating original content such as blog posts is hard, but the barrier for submitting links to original content is easy. Thus members will have more content and opportunities to interact with each other. Since members of an organization can generally be identified, registration could be automatically extended to all members. A site admin could maintain a list of active individuals and create / remove access and submission privileges when members enter and leave the organization. The School of Information could facilitate access through the Kerberos CoSign authentication system. Members would already be familiar with their username and password.
A username tied to a real identity would have the additional benefit of helping to regulate behavior. Removing anonymity reduces the likelihood of flame wars because bad conduct online could be conceivably punished in the real world.
In the early stages of the community’s growth, I would reduce expectations for content by limiting the initial audience to a group of “beta testers”. Members in a small community are more likely to contribute because their activity is more visible than in a crowd. By gradually opening the gates new batches of members would be socialized with the expectation that most people actively participate. The “beta tester” group could be expanded until enough messages are submitted each day to maintain successful interaction. In the SI community, I expect 10-15 active members per day would be needed before a full launch.
With the support of organizational leaders, a mailing list could be leveraged to promote this platform. A collection of spotlight links and conversations could be emailed to all members at a regular interval (once a week) and at a considerate time. Increased salience of community activity might encourage participation, or provide benefit to those members who don’t participate (keeping them up to date on popular discussions). In the School of Information, event descriptions for the upcoming week are sent out late each Sunday night.
The site should allow many ways to access content: RSS feeds for outgoing link submissions and comments, an optional daily email digest (all content or filtered by tag), and optional email/RSS notifications that notify a member when their links or comments receive a reply or positive canned feedback.
Positive canned feedback at 43 Things takes the form of a “Cheer” – which members can click to “show support for each other”. (43Things Community Guidelines) Whereas 43Things developers want to encourage inspiration, I would like to encourage interaction. Positive canned feedback in the SI community might be “Funny”, “Interesting”, or “Useful”. Good contributions could get people talking by making people laugh, showing them something new, or showing them something they can apply to their coursework. The amount of canned feedback each member can give should be restricted to mark unique contributions and status.
To make people want to stay in the early stages of the community’s growth, an earned reputation system would allow members to measure their impact over time. I believe this would motivate contribution because reputation marks are a public display of goal achievement. Whereas the earned reputation system at newsvine.com represents goal achievements with six leaves on a vine, I would customize achievement markers for each organization. In the School of Information, each achievement marker could be graphically represented in the form of a “credit” (a nice looking dot). Members that earn each “credit” would receive a “diploma” displayed next to their username.

Achievements for Contribution: (in order of easiest to hardest)
1) Post a Story
2) Post a Comment
3) Give each type of canned feedback
4) Post 25 comments
5) Post 50 canned feedbacks
6) Average 1 contribution a day over three months. (comment, canned feedback, outgoing link)

Achievements for Recognition (in order of easiest to hardest)
1) A single story receives a canned feedback
2) A single story receives 5 comments
3) A single story receives 10 canned feedbacks
4) All stories receive 50 comments
5) All stories receive 100 canned feedbacks
6) Outstanding contribution (recognized by a moderator)

Summary
(Word Count: 1958 Words)
A link discussion community could encourage interaction among members of an organization with 1) low barriers for member entry, content submission, and communication 2) high barriers for non-members, 3) customized technological features that align with organizational norms and structure, 4) appealing links for discussion, 5) motivating the submission of appealing links by constructing positive canned feedback that rewards valuable contributions with increased self-esteem, 6) getting people to stay long-term with a measure of earned reputation, and 7) support and promotion from organizational leaders.
Potential Roadmap (RoR)

May 5st – set team developers (3?), revise roadmap
May 12th – community analysis, personas
May 19th - prioritize features, select development framework, needs analysis
May 26th – create wireframes, contact SI faculty with resource requests (send wireframes)
Jun 2nd – revise roadmap, create MVC for links and comments
Jun 9th – MVC for accounts
June 16th – MVC for tags
June 23 – MVC for canned feedback
June 30 – MVC for earned reputation
July 7th – set foreign keys
July 14th – graphic design mockups
July 21st – alpha version, start testing
July 28th – Kerberos
August – testing, launch during 1st year orientation