Transcript Starting New Online Communities

Starting new online communities, Paul Resnick,
Joseph Konstan, Yan Chen, and Robert E. Kraut, 2012
Starting New Online Communities
KSE 652 Social Computing Systems:
Design and Analysis
Uichin Lee
Oct. 8, 2013
Contents
• Carving out a niche
– Modeling interaction opportunities
– Dealing with communities with multiple spaces
• Competing for a niche
• Getting to critical mass
– Bootstrapping by leveraging early members
– Attracting early members
• Increasing early stage value
• Providing early adaptor benefits
• Setting expectation for success
Starting a New Community
• Most online communities never get off the ground
– SourceForge: only 10.3% have more than 3 members
– Smokefree.gov: so few people used for testing..
– And many other instances..
• Why failure?
–
–
–
–
It’s not clear whether it offers services/experiences that users want
People want, but they never find it
It loses in a competition
It attracts a stream of users (one at a time), but fails to hold them
• Three major challenges in starting a new community
– Carve out a useful niche
– Defend that niche in the ecology of competing communities and
alternative ways that the potential members can spend their time
– Get to critical mass
1. Carving out a niche
• Three main design decisions
– Scope of the community
• Breadth of topics, kinds of people, types of activities,
purposes
– Extent of compatibility/integration with other
sites
• Borrowing of features and user interface elements,
sharing of user identifiers, importing/exporting
content/people
– Internal organization of content, people, and
activities within the community
1. Carving out a niche
Modeling interaction opportunities (Pull vs. Push)
• Cost-benefit analysis of visiting a space
– Model each space as a collection of interaction opportunities
– match_value: expected utility of examining opportunity and
possibly engaging with it
– collection_size: # of new opportunities since last visit (e.g., # of
messages in a forum, # of status updates)
– navigation_cost: cost of getting to the info space and waiting for
its contents to display
– interruption_cost: synchronous interaction via push notification
in which a user is notified each time a new interaction
opportunities becomes available (e.g., emails << pop-ups)
• Pull vs. push model
– Pull: match_value*collection_size – navigation_cost
– Push: match_value*colleciton_size-interruption_cost*colleciton_size
1. Carving out a niche
Modeling interaction opportunities
• Synchronous interaction:
– DC1: Lower volume and higher time-sensitivity of
interaction opportunities, and lower interruption costs
increase the benefits of push notification
• match_value reduced by:
– DC2: Mixed-topic scope
• Lumping several independent topics together as a mixed-topic
scope; resulting too many uninteresting articles..
– DC3: Ambiguous scope for an interaction space
• Ambiguity about its scope may result in a mixed-topic scope
1. Carving out a niche
Modeling interaction opportunities
• match_value improved by:
– DC4: Activities that bridge interests in different topics in
spaces with mixed-topic scope.
• Bridging interests: intersection of interests between two topical
scopes
– DC5: A transcendent or bridging topical identity in
communities with mixed-topic scope.
• May have an interest in or identify with a broader topic (e.g.,
naming such as NBA, World Cup)
• Connections between specific identities or topics can itself
become a new topic around which identities and a community can
form (bridging community of practice, Wenger 1999)
1. Carving out a niche
Dealing with communities with multiple spaces
• Advantages of combining multiple spaces into a single
online community:
– Amortize the fixed costs of development and branding over
more users and uses
– Reduce search costs for members (easily exploring other spaces)
– Create synergies among spaces within a community
• How to organize/display multiple spaces?
– Globally defined views (fully shared context) vs. personalized
views (less shared context)
• DC6: Personalized collections of “most related content”
enhance match_value but reduce shared context
1. Carving out a niche
Dealing with communities with multiple spaces
• How do divide spaces?
– DC7: Subdividing spaces after they become active creates more net
benefits for participants than having lots of inactive spaces.
• In communities with lots of interaction spaces
– DC8: Navigation aids that highlight more active spaces
– DC10: Recommender systems that help people navigate to spaces
• Less active synchronous spaces:
– DC9: A schedule of “expected active times” can coordinate visitors and
become a self-fulfilling expectation
• DC11: Ambiguity of scope: sometimes helpful for adjustment
and member ownership (e.g., by learning from members;
helping them to interact one another for meta-discussion)
2. Competing for a Niche
• Why conducting a competitive analysis?
– Help understand how potential members muddle through with their
existing technologies
– Give chances to understand future competitors if any
• Incorporating switching cost into the net benefit model:
– Switching cost: learning s/w + finding a match + learning norms +
building up social ties/status
• (learning s/w) DC13: Differentiated user interface elements in the
competitor community create startup costs
• (building up) DC14: Non-shared user IDs and profiles between
incumbent and competitor communities creates startup costs
• (building up) DC15: Content sharing between competing
communities raises awareness of the exporting community and the
value of posting there, but raises the value of consuming content in
the importing community (beware of tradeoff)
2. Competing for a Niche
• DC16: Conveying a succinct unique selling proposition will
attract members
– Important to clearly convey the benefits
– Many people rely on heuristic processing (e.g., relying on short
description/review, etc)
– In online community, it may be a topic/tool
• DC17: Making a focal point in a competition between
communities, by advertising and celebrity endorsements
that help to create awareness of a community
– “the aura of inevitability is a powerful weapon” (Shapiro and
Varian 1999)
– Cultivating public awareness to create such an aura
3. Getting to Critical Mass
• Network externalities or network effects
– One person’s value from using a product/system increases with
the number of participants
• Yet:
– below critical mass, adding a member may not create enough
benefits to outweigh the cost => no participation
– Joining crates a positive externality (benefit for all the other
members), but benefits to the newcomer may not outweigh the
costs => no participation
• How to attack this problem?
– Effective use of the early members to attract more members
(called bootstrapping)
– Making communities more attractive to early joiners
3. Getting to Critical Mass
How to improve bootstrapping?
• DC18/19: provide incentives for early
members to generate primary content (e.g.,
Epinions)
– Note that user-generated metadata (e.g., tags,
comments) is less useful; it would be useful when
critical mass has achieved
• DC20: enable displays of membership that are
visible to non-members (e.g., subscribed
membership info in a user’s profile page)
3. Getting to Critical Mass
How to improve bootstrapping?
• DC21: make members' actions in the community visible to
their acquaintances outside the community
3. Getting to Critical Mass
How to improve bootstrapping?
• DC22: allow members to forward content from the community to
their acquaintances outside the community
• DC23: allow members to invite acquaintances outside the
community to join
• DC24: pay-for-referral and revenue-sharing from referrals increase
bootstrapping
3. Getting to Critical Mass
How to attract early members?
• When to join? early stage (stage 1) vs. later stage
(stage 2); utility modeling
– Stage1:
• Start up cost:
• Participation net benefit at stage 1
– Benefits from participating at stage 1 - switching cost
• Early adopter benefit = additional benefit for the early
members (e.g., status in the community)
• Success probability of this community
– Stage 2:
• Participation benefit at stage 2 = an expected benefit from
the second stage
3. Getting to Critical Mass
How to attract early members?
• When to join? early stage (stage 1) vs. later stage (stage
2); utility modeling
– Utility(join now) = part_ben_s1 – startup_cost +
success_probability(part_ben_s2+early_adopter_ben)
– Utility(join later) = success_probability(part_ben_s2startup_cost)
– early_adopter_ben : additional benefit for the early
members (e.g., status in the community)
• Attractive if utility(join now) - utility(join later) > 0
 part_ben_s1 – startup_cost(1-success_probability) +
early_adopter_benefit*success_probability > 0
3. Getting to Critical Mass
How to attract early members?
• Attractive if utility(join now) - utility(join later) > 0
 part_ben_s1 – startup_cost(1-success_probability) +
early_adopter_benefit*success_probability > 0
– Benefits of early stage participation
– Providing early adopter benefits
– Setting expectation for success
3. Getting to Critical Mass
How to attract? benefits of early stage participation
• DC25: providing single-user and small-group
productivity, entertainment, or commerce tools (e.g.,
bookmarking, notes, Yahoo! Groups, Amazon’s social
recommendation)
• DC26: providing access to professionally generated
content (e.g., NYTimes’ discussion in each article)
• DC27: providing access to syndicated data
– if the syndicated data is not otherwise easily accessible
– if it is presented in a novel way that adds value.
3. Getting to Critical Mass
How to attract? benefits of early stage participation
• DC28: staff members’ positive sides (e.g., MSDN Q&A w/ staff
helped encourage people to answer)
• DC30: if staff members act as contributors of last resort, they will be
needed less and less as the community achieves critical mass.
– Trick: wait until volunteers perform tasks and take on only those tasks
that volunteers do not
• DC29: starting with a limited scope and expanding later allows
focusing of staff resources toward getting to critical mass in the
limited scope (e.g., Craigslist, Citysearch)
• DC31: Bots that simulate other participants can help attract people
to an online space before the community features are successful.
3. Getting to Critical Mass
How to attract? providing early adopter benefits
• DC32: Promising permanent discounts
• DC33: Promoting the status or readiness benefits of being
early to an online community (e.g., wikipedia, online
games, Slashdot’s user ID)
• DC34: Promoting a site as cool but undiscovered (e.g.,
exclusivity of gmail invitation)
• DC35: Creating scarce, claimable resources (e.g., user
namespace)
• DC36: Contribution minima for maintaining scarce status
can lead to greater contribution by status-holding members
• Recall: early adoption as sunk cost, but its effect plays an
important role for decision making
3. Getting to Critical Mass
How to attract? setting expectations for success
• What are design choices that convey signals of affecting
expectation of success?
• Signals of Convener Quality and Commitment
– DC37: Professional site design increases expectations about the
probability of success
• E.g., Fogg et al.’s work on credibility of website: design look
(+pleasure), info design/structure.
– DC38: visible resource expenditures can be a credible signal of
commitment to future investment in a community
• Theory of credible signaling (Spence 1973): community conveners as
high quality actors of providing signals
• Use of immediate resource expenditure as a credible signal of greater
commitment to future investment
• Online examples: 1) new design, software tools; 2) new feature
updates; or update frequency
3. Getting to Critical Mass
How to attract? setting expectations for success
• Signals of Positive Member Response (+activity)
– DC39: Images of members (say, accompanied w/ user
generated content)
– DC40: Prominent display of user-contributed content if there is
any (to signal “activity” within community)
– DC41: Indicators of participation levels (e.g., who’s online now?
And how many?)
– DC42: Indicators of membership/content growth
– DC43: Small size + slow growing: acking each new
member/contribution
– DC44: Small size + fast growing: displaying growth rate
– DC45: After reaching critical mass: displaying absolute #
– DC46: Conditional participation commitments (e.g., if 250
people buy... Groupon..)
3. Getting to Critical Mass
How to attract? setting expectations for success
• External Signals
– DC47: Drawing analogies to successful communities can
raise expectations that a new community will be similarly
successful (e.g., mamapedia, “the wisdom of moms”)
– DC48: Drawing attention to external publicity and
endorsements can raise expectations about future success
Summary
• Carving out a niche
– Opportunities model: e.g., (push) synchronous interaction design vs.
pull: match_value (topic scope, ambiguity, bridging interests, topical
identity)
– Communities with multiple spaces: match_value (personalized view
vs. shared context, how to divide?, navigation/recommender, size
control)
• Competing for a niche: e.g., unique selling proposition, interface
design, being a focal point
• Getting to critical mass
– Bootstrapping by leveraging early members: e.g., content generation,
exporting/forwarding content, inviting people, referral
– Attracting early members
• Increasing early stage value: e.g., tools, content access, staff support, value
• Early adaptor benefits: e.g., discounts, exclusivity, scarce resource,
• Setting expectation for success: e.g., professional design, visible resource
invest, participation level, drawing analogies