Lessons learned from Collaborating Minds


What is Cminds

Collaborating Minds (Cminds) experimented in scaling and sustaining virtually a high-performing teams’ capabilities and success problem solving . Our efforts combine:

  • Assembling diverse individuals with a set of common beliefs conducive to creating a high-performing team
  • Strategically mixing asynchronous and synchronous software processes designed to :
    • Facilitate team member interactions
    • Structure problem solving activities (e.g. sourcing, processing, sharing and digesting structured and unstructured data),
    • Enable formal and informal communications
  • Actively managing and nurturing familiarity among members, (loosely called “community”) and actively managing the problem solving process.

Cminds’ distinct freestanding system combines content management, workflow , community interaction and a problem-solving process.

The benefits of integrating functional tasks and social capabilities mirror the natural fluidity and impromptu interaction typically present in high performing teams . We wanted to avoid both an overly structured methodology that would force people to quickly respond rather than think and reflect. At the same time, we wanted to add to the rich profiles and interactions that many online communities and enterprise social networks offer but providing structure to collect and process output in order to advance work-related tasks.

Most recently, some tools have begun to integrate community and workflow by organizing collaboration around a specific artifact in an existing workflow. For example, in Chatter/salesforce.com, participants can post comments around the artifact of the documented customer relationship. Other tools let people comment on evolving designs or engineering drawings, with participants’ comments attaching directly to the specific item. Cminds, however, invites freer responses, fosters open-ended aspirations and allows members to generate artifacts together as well as comment and engage with whatever others have started.

We wanted to explore if enabling both problem-solving structure and community would prove mutually supportive — that the establishment of the team would make problem-solving more effective, and conversely, that the opportunity to work together in problem solving would strengthen the community.

To create our community, we deliberately sought out people who shared a set of values that we thought would increase the chance of the team succeeding. These values included:

  1. An interest in problem-solving and getting better at it,
  2. An interest in relationship building and getting better at it, and
  3. A desire to help and be helped

Our Journey: Progress to date

What we’ve done so far

Diversity of skills, experience and background are highly desirable qualities associated with high performance teams. We wanted to test if it was possible to scale the diversity without sacrificing the performance.

We recruited 250 interested people over 12 months. We began with personal appeals to the organizer’s contact networks , followed shortly with directed appeals to “likely suspects,” or existing organizations, established linked communities (e.g. Linked-In groups, association list serves). The indirect appeal included a 2 ½ minute animated video we created explicitly for recruiting.

Recruits were asked to review our community principles and then complete an extensive profile, or “member information” form. The form deliberately created a high barrier to entry that proved effective in pre- screening people based on willingness to help and be helped. The form proved daunting for many, and we nudged, reminded, and offered assistance in a variety of formats such as video conference informal “coffees” and access to other member’s video responses to the questions. By the time we kicked off problem solving , about 120 people had fully completed the “member information” and were eligible to participate on the full community platform in Cminds.

We achieved our diversity objective; our membership included people from a variety of disciplines and globally. We wanted diversity of knowledge and perspective too in order to maximize the problem-solving creativity of the group.

We appreciated how diversity increases short-term communication costs. In order to communicate with people outside one’s own field, one must learn to state clearly and explicitly frameworks that one has long internalized but that strangers to the field don’t take for granted (and are usually even unfamiliar with). We also knew from our own experience that once people know each other a bit, the ongoing opportunities are much larger than the early costs. Once members completed their profiles, they were given access to the community platform where we offered participants opportunities to engage with each other virtually.

The problem-solving prompts consciously encouraged clarification, and the plan was that with frequency of working together, interpreting and understanding each other would get easier. We hoped that members would also make the investment in getting to know each other, and therefore our team would increase its communicative capabilities.

The group worked together on one structured problem, i.e., the use of data to improve decision-making in American public schools. The problem solving structure we designed included seven rounds of iterated responses that would then produce a recommendation to our client. After three rounds of idea generation over two months, participation declined and we stopped when we realized that the interaction design needed more work.

Members also interacted in the community space in several ways:

  1. Members posted and responded to “requests for help” and “members other projects” in the community setting.
  2. Members engaged in online asynchronous discussion boards on topics other members had posted
  3. Members read each others’ profiles and communicated with them
  4. Members were interviewed by videoconference and had edited videos posted; they watched others’ videos
  5. Members attended video-chat “office hours” and “coffees” that were designed to let people see and hear each other

The results :

  1. New Insights for the client – the questions and clarification offered new perspectives and clarifications that helped expand his understanding of the issues and situation in spite of years of experience working in/on the problem.
  2. Strong community ties—Many weak connections increased and new connections were made. There was no way to really measure the community tie.
  3. Good experience for (many) participants—many acknowledged surprise and were pleased to see their own fears and doubts vanish over a virtual coffee, or seeing the synthesis and contrasting contributions to their own thoughts and perceptions.
Team vs. Community vs. network vs. bilateral relationships vs. value network

Although we talk about the Cminds community, our actual effort sought to strengthen bilateral connections, and to create a team and ultimately a value network.

Teams can have a performance target. Networks, communities, dyads and value networks usually don’t have performance targets, and it is harder to think about how to measure them. Our premise is that teams perform better when individuals on the team commit to both the team’s success and the success and development of other individuals on the team. (See Katzenbach, Wisdom of Teams). We shared Katzenbach’s belief that the performance of the best teams depends on the quality of the bilateral relationships among its members.

Cminds had components aimed at trying to enhance bilateral connections. For example, we expected creating rich profiles would make it easier for people to take an interest in a stranger. We also enacted a series of public “community” meet-ups that furthered bilateral opportunities. We provided opportunities for members to help each other, both in a community setting and bilaterally (publicly or privately); we believe that if and when Member A helped Member B on a matter of concern to Member A, the tie between Member A and Member B would grow stronger, and would also to a small degree make the community stronger.

We were focused both on creating more network ties among members (a denser network) and higher quality ties. Our focus on the quality of the ties reflected our interest in creating a “value network” in which members could provide value to each other, on matters handled within Cminds and on matters outside Cminds’ purview. Our hope was that as members began to interact, richer bilateral connections would emerge. Specifically the casual interactions required for Cminds’ activity would lead to richer, more meaningful inter-relationships. In some cases, members did discover one another and explored and developed mutually beneficial activities outside of Cminds. Cminds was to serve as a substrate and catalyst for other interactions in which members would discover — in pairs, small groups, or whatever — additional beneficial opportunities . We intentionally sought to create a stimulating, virtual environment (the community angle) where people willingly give to others (according to the initial values on which members joined) and also accept rewards coming from others (paying it forward, karma).

In our development of the environment and the task and as we began to attract and engage with members, more of these concepts became evident, present in the system, and conceivably available to analyze; and in turn they served as levers to affect activity, interactions, etc., as summarized below:

Result Locus

Activity focus





Value network


Bilateral results (increased knowledge of other individual and commitment to their success and development)

Tom read Mike’s profile and contacts him

On Cminds hosted multi-person videoconference Bob meets Sally and they decided to follow up privately

Olivia’s public comment on Oscar’s “request for help” strengthens connection between Olivia and Oscar

A member acting as “broker” introduces Mike and Max

Providing unilateral , easy to provide help to a member you already know strengthens bilateral ties

Tom reaches out to Mike to ask that he contribute to the group activity


Community result (increased sense of belonging and identification)

Video Interview standard forma and questions ushers and welcomes you into the community.

Interview standard format and questions ushers and welcomes you into the community.

Water cooler conversation about problem solving leads to community reflection about its own strong capabilities

Public dialogue on cultural differences –increases awareness around communications beyond Cminds

Network result (new nodes or edges in the network) Any new bilateral connection is a new edge in the network Any new bilateral connection is a new edge in the network A member acting as “broker” introduces Mike and Max    
Value network result (incremental increased value at nodes > increased costs) Same as bilateral results Same as bilateral results Same as bilateral results Same as bilateral results  

Team result

(better results for team)

Same as bilateral results Same as bilateral results Same as bilateral results   Adding new members with particular skills or capabilities (we didn’t do this but we could have);

More Reflections on the results

Connectional intelligence

The group did generate some examples of connectional intelligence. Members made small contributions and others built on them. Different stimuli provided opportunities for members to advance ideas which other members furthered using discussion threads. Within the structured problem flow, we provided multiple entry choices and members built on each others’ ideas; that included teasing out further meaning with examples or stimulating new thoughts. Some contributions came either in direct comments made by the client or by others –and introduced ideas that would not have been linked to this topic without the benefit of previous comments that had been loosely offered within the freer input structure we provided.

Diverse inputs

The diversity of the inputs was remarkable. Members’ suggestions to the test problem showed how the original narrow frame limited the solution set, and offered the problem owner new alternatives, increased opportunities and paths worthy of further investigation.

Member benefits

We created an environment and tone conducive to learning by doing. Problem solving activities offered members a chance to analyze, improvise and use their intuition, and thus demonstrate, hone and extend their skills. Members also learned by seeing what others said and did. And if they posed “requests for help,” posted about “other projects,” to group, then many got help on their own problems.

Members also shared and explored topics of interest by opening new discussion threads, or posting references, or suggestions for book discussions.

Network/ community/ value network/ team building

Members did make new acquaintances, but more importantly gained access to a willing group of reliable, like -minded individuals who share an appreciation for helpful exploratory discussions. Members, especially core team members, actively connected members to one another via “matchmaking” in response to a particular issue, need or simply because someone thought two people would like to know each other. Some members were introduced to new contacts outside of Cminds through their connections on Cminds. Cminds’ membership proved a sufficient endorsement when participants did not know each other well (similar to the prescreening that a common alumni affiliation may confer in other settings). Sometimes the willingness to connect to a third party outside came as a result of exposure within Cminds.

Building Cminds, the software and the organization: what it took and what it still needs



What we built and planned

What happened

What we should change or add

Place where members could post “request for help” and “other projects”

Some posts and some assistance

· People don’t like “help”; maybe something like “input”

· Having both “request for help” and “other projects” was confusing – probably should just have one

Watercooler and other community groups

Some good discussions


Edited interview videos (usually about 5 minutes)

· Good insight into members

· People like doing the videos but not too many people watched them

· Doing videos bound people to the community

· Took a lot of time and energy

· Share videos with members to use elsewhere

· Shorter videos perhaps

Extensive profiles

· Served as barrier to entry (as planned)

· Many people found questions challenging; some found process enlightening

Need a way for people to search profiles to find experts, contacts, or common interests

Need a way to display who is who in various categories, cf. guy in Santa Barbara who has this approach

Office hours video conferences – Twice per week the core team was available on a video conference call

Sporadic attendance but very good discussions (4 -8 people) when they occurred

Remove the academic association and rename them as something like the Weekly Preview or Weekly recap or something more engaging, and also send more select invitations to people to help them feel special. Feature a topic etc.

Online “coffees” – Videoconferences with some topic to stimulate discussions

Some led to great discussions (6 -12 participants). Others were duds

This holds great promise but email blasts and time zones were insufficient. Again, incite interests by inviting members to suggest a topic and trade off the facilitation.

Community management by core team

Core team did a lot of community management work and network weaving. If there’s a threshold at which this becomes a spontaneous activity of the community we didn’t reach it

Provide more roles for members

Slow and difficult ramp-up

It was challenging maintaining the small community as we recruited new members. The recruitment effort could have been accelerated with more resources upfront and the use of additional automation support, which we acquired and developed over time, e.g. don’t start recruiting until you have all the lists together, use the tracking and email tools

Have a stronger core before a launch, e.g., 20 people committed to the community and willing to be active in it

Other new possibilities


Other structures or approaches to make bilateral relationship building easier??

Problem –solving

What we built and planned

What happened

What we should change or add

Good problem-solving starter questions and a model/idea to continue to adjust (positive ratchet effect)

Worked well. The redefinition of the problem was not put to use so there are some logical disconnects that need attention.

Figure out ways to display results more clearly (UI issue). Use the redefinitions either as the basis of new strands, or have the problem owner review and then clarify the problem continued exploration.

Structure of multiple rounds, with information only visible to members after a round was done

Members wanted to see what was happening during the round, e.g., what were other people saying

· Increase transparency -make inputs visible sooner

· Make rounds shorter — depends on getting synthesis to happen faster, and in any case only shorten rounds to the point where slower people still have time to contribute

· Provide members with statistics on how many people have contributed in round so far and other things of interest to them

Sandwich meetings to allow synchronous

Did not use

Actually use them!

Back office synthesis – by humans

turns out to be hard due to large volume of unstructured data. Took a lot of time and energy

· Support the human synthesizer with automated tools for tagging documents

· Automatically (or not) suggest splitting paragraphs for users before they save so users can identify the best tag for one thought and also identify role in argument*[1]

· Synthesis needs to be seen as a “fair process” — people need to object to the fact that their pet idea was included [JM1] and know that their objection has been heard — then they can move on

· Could also do more widespread, shared synthesizing following model of Chris Corrigan

Large group problem solving alienates some members

Some members or prospective members feared being disconnected in a large group and preferred a way to participate in a smaller group

· Build in smaller teams for certain problems and subteams for large problems. Following style of Chris Corrigan (above) you can move ideas through different groups for responses and elaborations

We anticipated the volume of information would overwhelm any single individual and so we intentionally planned to break into smaller chunks manually, and then display it in an accordion style format

Individuals were instructed to only respond to a subset of the information, and also given freedom to review and respond to more subsets or the total.

· Need a way to contextualize the information without compromising their unique objective perspective Otherwise they view the task as overwhelming ; perhaps UI/UX corrections will suffice to alleviate this issue.

Group communications used to motivate member participation in different rounds

Transitions between rounds was slower than expected, as back office process changes and UI fixes were made to support next round. Though beneficial it also caused us to lose momentum and involvement of participants along the way

· Build better synthesis, process and back office tools.

· Provide more resources for problem-solving management


What we built and planned

What happened

What we should change or add

We were intentional about our communications, and prepared many in advance such as the process flow and step wise trainings. We actively tried to provide routine updates in a standard format (e.g. our newsletter and process round announcements).

Lots of good communication:

· Newsletter tailored to both prospects and then to members.

· Support for problem solving, e.g. the training

· wider communications via the public blog

More resources

We chose a particular way to represent the problem (as one must) and provided both structured and unstructured ways for members to respond

Many members did not respond right away; hypothesis is that they were overwhelmed both by volume of information and multiple ways of responding

Make a clearer separation in early rounds between redefining the problem and responding to the problem as posed

Force client to make more decisions — and team to appeal decisions explicitly or live with those decisions after appeal/ a “fair process”.

Video conferences supplemented with text discussion that engaged and helped people get to know each other

Worked well

Get some members to invite others; expand the chain of involved people

Training — the rules of the road — on screen access to instructions. Videos on how to do things

People were still confused…in part we suspect due to UI/UX, absence of predetermined representation (the prompts were intentionally open-ended); and because the overall process was complex.

Explore what could be done better although we had a good start.

Email updates

Email updates were driven by on site activity but were hard to read. The system did not permit comments or update via email; but required users to log in and be onsite.

Use rich text format, alternative APIs increasing responsiveness via email or other direct options (e.g. voice, text (SMS), Twitter etc.)


Rewards and motivation/ gamification

What we built and planned

What happened

What we should change or add

No rewards for participation in the problem solving or for content quality

Some participated and some didn’t

Need some mechanism for giving credit/ compensation

Points for contribution to the community

Some people were motivated by it (e.g., contacted us when they thought it wasn’t recording them correctly) and others appeared indifferent

Unlike StackOverflow and single-issue communities, reputation in this community may not prove necessarily motivating


What we built and planned

What happened

What we should change or add

Moderated discussion forums

Activity improved as we had more members cross over into the community space. Moderation was not needed to check quality of answers but rather to ensure that people got responses

Initialization of activity could have improved with some additional back channel contact and connection.

Mailchimp managed targeted list mailings e.g. newsletters, and phase kickoffs, coffee announcements etc.

Open rates were consistently high, tracking helped us manage further communications

Improve the regularity of various communications

Gmail/Streak for tailored/custom mailings and tracking

Used in recruiting and later in assigning different users tasks in the problem-solving phase


Statistics to track contributions, and site visit activity


Make some of these visible to members at large

Better stats to drive further contributions: e.g. ,

• member interactions

• Top contributors listed

• Synthesis contributors


Substantial investment in administration and communication

Substantial investment required to hit point where community is self-sustaining (or substantially so) — we didn’t get there


Intentionally staged recruitment process that tracked repeated inquiry, helped agreeable parties to begin the completion of membership , scheduling video interviews and coordinating their formal entry onto the full community platform

Supplemental staff with exclusive responsibilities and coordination of status notifications were established across a variety of systems

–CRM, mailchimp, streak excel spreadsheet and Cminds site profiles

Consolidation of systems improve process and focus on communications over tracking.


UI and UX challenges

UX issues
  1. Problem-solving needs to have steady rhythm (presentation, comment , synthesis, comment, decision, closure)
  2. Keeping track of what happened and making it accessible is important – the system provides a historical documentation
  3. People want to know what’s happening along the way — there’s a tension between the round structure and the ability to see what’s going on.
  4. Need only one place to do anything in particular
  5. We need to design for the moderately experienced user and provide help for the beginner. (OR SHOULD IT BE INTUITIVE EVEN IF LESS FLEXIBLE)
  6. Need search capability
UI issues
  1. Some specific issues
    1. Needs to be mobile
    2. Pop up box/ Modal frame was OK idea but didn’t work on mobile or most tablets
    3. Ajax would be an alternative solution for desktop
    4. Voice would be an alternative solution for phone.
  2. Do mobile first design which will force simplification
  3. Get professional help!! on UI and mobile design

[1]Role in argument refers to Dialogue Mapping approach that says that each statement is either an idea, a question, an argument in favor of an idea, or an argument against an idea. The software could be designed to capture the parent of the statement and ask the participant to describe what their statement is


2 Responses to Lessons learned from Collaborating Minds

  1. Bill Hass January 19, 2016 at 3:27 pm #

    Nice Summary …. Big investment of time …
    People need immediate feedback and incentives…

  2. Bill Hass January 19, 2016 at 3:28 pm #

    Nice summary … Big investment in time!
    Learning is expensive!

Powered by WordPress. Designed by Woo Themes