Notes from the Seattle Online Community Meetup

These are notes from the March 12, 2008 Online Community Meetup sponsored by Forum One, organized by Bill Johnston, and hosted by Robert Rebholz of Microsoft, in Redmond, Washington. Usual live note taking caveats apply: I did not capture everything, and nothing that I said. I did not generally capture who said what. No one asked for soft NDA and I explicitly said I was posting these notes. Spelling and grammar don’t count. Added later: see Theresa’s take aways and twitters.

Online Community Meetup Gathering at Microsoft

• An hour of conversation, beer, wine and food
• Introductions
• Open space with 6 speaker slots of up to 15 minutes. “Speaker” – meaning you can do a formal presentation, or just ask a question to stimulate conversation.
• We moved chairs from theatre style to a round and I was the first up. I offered my question but did not take notes during this round. …the gist of the question was “how do we encourage sharing of content and knowledge ACROSS communities and organizations. Currently in many sectors the habit is to do this internally. In international NGOs and NPOs, this doesn’t really help advance the cause…I told some stories about work I’m doing, people asked me specifically about adoption patterns of wiki use and behaviors, etc.
• Facilitate personal experience, creating meaning
• Informal networks
• How do we build up to the organization
• Percolating
• Metaphor for collaboration systems = messy – shift to a metaphor about relationship
• And behavior
• TED – Example of people driven

Kim Malek –start up around learning about a health condition from other people like you. Relevance and actionability of information. Have a credible base of initial people. Started with an invitation only network. Opinion leaders, organizational partners. How does that scale? How much friction do you bring into the picture when you also want viral growth. Closed or open? Invitation or not? How to grow viral and maintain base of credibility.

What about anonymity? Can’t provide a story w/o a persona but you don’t need to use a real name.

Bob – it is very rare in nature to find collections of peers. Becoming skeptical of pure peer groups. Rare to see successful groupings of collaborators that are peer based, a group of equals. Unusual from a cellular level. Even effective online groups, it tends not to consist of groups of not like individuals, but around common shared activity.

Peers by Bob’s definition means people are very much the same. Isn’t that what FB is all about. Hooking up.

Have you looked at dating sites. There needs tro be a different bar or barrier in terms of who you want to validate. Don’t want to be another web MD – a place for consuming content. You want a place for connection. How you profile u p and how far the bar is raised?

How long does it take to sign up, describe health interest. Have you thought about doing some of the sign up up front in person? Yes, have started to do some of that, but how viral and scalable is that? What about peer review? Looking for a give/get.

What about seeding it to be the way you want it to be? Create a space that encourages the kind of people you want. Cancer Survivor’s network was totally open. They totally please themselves. They come in and do things and before organization responded the community were on the people.

What was your invitation process? Light weight. We are reaching out aggressively to influencers in the space. No credit card. IT is free. Just want to validate you are a person, have a health interest and have written something intelligent. Other health sites are more closed. You can see anything on Trusera, can’t post until join.

Look at invite only as an acquisition tactic, not a barrier. Get self selected people who really want to get in there. ShareYour Story as an example.

Group cohesiveness is higher the higher the barriers are to entry.

The way you construct the registration process is indicative about how you behave. The MOD community looks pretty, smells pretty.

User generated health content – author and content rated.

Thinking about why you need to grow may be an important question to answer. Growth can kill it. What is the ideal size.

What kinds of grouping do your tools to support. Forums provide groups of certain sizes to interact. Incentive structures to engage groups of what size? 3 is different than 300 or 3000. Those rules are built into your system that govern behavior and interaction. They are represented by the way the product works.

Frank Jurden – VML/Wonderman network – agency that helps MSFT.

2 minutes of experience at TED conference. First time, attended sister event in Aspen. Talk about the challenges of growing an event and keeping it special. Started in mid 80’s. Next year transitioning to a global franchise model. Pleased how they approached and levered online what we do in the offline. Exchange biz card, etc. No conference I have been to – the way they handled your profile and ability to meet people online. You can chose your visibility, your contact notification. Everyone who registered has a profile. If they choose to make it visible. See what they are interested in. Gave you a set of predefined personality tags. You name tag/badge has your big picture from your profile, so can connect online and offline. Great way to connect. Wonderful example to leverage online and offline in conference setting.

Any social graph representations? Any way to connect nodes. Doesn’t look like intronetworks.

People sent emails to each other prior. Not much SMS or mobile apps.

Some conversation about Twitter a SXSW. No one central hub. Text you with session info. Neat to find your own channels. lots of third party and sub groups using their own tools/networks.

TED has some basic compulsory registration. In another setting a wiki is used, use the same space every year and build on that. Do you use it in between the conference? Some yes, some no. Interest in building continuity between F2F but the don’t work.

You connect with people the way you connect w/ people offline. Marc and I met at NTEN and we kept our relationship going outside of the NTEN website. The relationships transcend the events.

Web community forum – thought about doing a wiki for our conf on FB, but ended up using FB. People friended one another. Forms longer term connections. Build vs join.

How do you measure that? Repeat participants. Participants become organizers in subsequent years. You can spend a lot of money on 50 attendees, and then it is one off useful. Or use an external tool or existing website.

TED site is open to use beyond TED participants. Brand. Marketing expense. For metrics, what is the timeframe and the scale. Networks of participants – that would be your return. How much is it worth sustaining. Customer lifetime value, relationship. How do you measure the network effect and is a community part of that. Will people be expecting a conversational experience or not. It is possible to measure the network effect if you own all aspects of the medium.

New social forms are arising. When we talk about having to connect w/ people after the event, it used to mean chat on the phone tomorrow. Moving past how we connect in the physical world today. Creates new form.

TED tells you to put your laptops away and turn off your phone and live in the meet space. If you are a K scale blogger segregated area. Not just about meeting up after, but the F2F is the continuation of preexisting online relationships.

Depends on the context. Was the experience better with more or less of the back channel. Some people feel more connected as a part of the connection in that social web. An initial leap of faith that you can back out value.

Speakers rated based on how much twitter traffic they generate

Rachel Klein – Works in education, trying to improve HS graduation and college attendance for low in come. Millions of kids dropping out, many dropping out even in middle school. Crisis. We work with leaders of charter schools, principles, superintendents, state policy orgs, advocacy. Not so much teachers in the classroom. Talk to these people all the time. I want to know what so and so is doing, what they do about this, how that. Concrete tactical things. We have a conference, this is great, I want to see that from you later. So we are creating a website to do this. Question about level of sophistication of users, how much of this they can sustain/partake in. What they tell us they want to do and what they do is not the same thing. We provide resources online, we send them resources and they don’t use them.

How do you get people to tell us what they are going to do and actually do it.

Second thing – traditional group of people who like to organizing things in hierarchies and taxonomies, but not sure they work anymore. (Librarians have had heart attacks). If you take a problem or category it doesn’t roll up in one way. How do you represent that. And do you. Is our community ready for tagging. Or is there some visual or dynamic way of representing information.

What will users actually do? A hard question. One of the hardest ways, clearest dead ends is asking them. Lots of evidence in cognitive psych, if you ask what they want, what will make them happy = they can’t tell you. No better than tossing a coin if they are actually satisfied with what results. Weird wrinkle to that. If you come after the fact, they will often say they are happy. Often don’t know what is possible so can’t ask for it. Behavior based inferences are better than verbal inferences. By giving choices, and don’t let them talk about it, watch what people choose and infer their value systems.

Recent Harvard biz review – chorus product development. 2 phase. Decrease time to failure with small studies, then increase probability at launch. Fact finding/trust finding rather than loyalty to preconceived notions. Also you have to go back to users. People focus on the majority and their need states, maybe you need a MVP, or a notion of a group of select individuals who help you deal with the problem of articulated needs.

Bootstrap it.

Can’t expect it to happen on its own. Have to facilitate it over and over again. You have to be engaged. Actual peers, engaged as a peer in the space and be jabbing people behind the curtains. Ambassadors – not just spark, but basic education. Structured programs that can help scale. Reward and EMOWER passionate folks. It can scale if you are part of the community at the start.

Starts w/ grantee community , then opens up to wider.

Don’t chase too many features. Don’t chase tech early adopters. Organizational, professional reluctance to get within anything with social networking.

Do the community elements enhance the experience and how. If they are simple. Usable and things that help them get their job done. Things w/ purpose/meaning. Lots of social network stuff is just cool shiny factor. Start simple and watch behavior. Figure out a way that whatever action you are trying to get them to take can fit into preexisting workflow. Insert into that.

Example. When people add a forum post, now we ask them to tag. Then you can get RSS feed, etc. The tagging is an instance of a new behavior that is added to an old behavior. Hotwired into that flow.

Contact list – pulls it in easily into the tool. I hate doing that, but it works. Don’t know what the workflow is. Watch for the usage patters for whatever they do now to solve problems. Not send them to new site outside of their normal workflow. How much are you talking about creating a new model for people to do the existing things they do, or actually creating new thing to do.

People already interact on listservs and email. What is the next hook in.

Basic community building – Saul Alinsky talked about trying to organize an neighborhood to get a stop sign. Community did not get visibility. They picked a big boulder and dropped it right there. That shock, that notion got the people trying to organized engaged. Gave them a sense of a small win. That was a catalyst for subsequent action. Rules for Radicals. Example of community organizing. They looked at their own meager resources. People and rocks. One of the principles in Made to Stick. What capacity do they have and how do you tap into that?

Then how do you punctuate that

Part is an organic approach. Unobtrusive invite in. Add something to existing workflow. Can you afford to wait long enough for that to stick and what is the cost, vs the revolutionary rock dropping.

Search on Google for IT failures. Dropping the tech rock.

Tagging and social bookmarking –

Will you be able to surface content in multiple locations through the standard canonical navigation. Can an article show up in multiple parts of the site. Should our site HAVE sections. There are people who think we have sections. Lots of different topics. Do we put those out. Do we spend the content mgmt. Let search take over?

2 things. There can be many subjects. Not sure that passions are the same demarcation. People will group and activate around things that give them some passion. Do you categorize around areas of passion as opposed around subject areas.

The way we are handling – we have both subject areas, navigationally organized. Our people don’t user search. They are librarians. We get emails when things aren’t organized the way they want them. Virtual cards.

We allow people to form groups which is surfaceable/findable through same navigation scheme. Function similarly to subject sections. That’s where communications happen. Trying to pull together traditional web CMS w/ social networking tools to quickly get people to resources, answer questions, professional dev and make informal connections with people with similar connections.

Brian Hsi– on flip side everyone searches. Same end space, but different behaviors. 60% inbound traffic is still direct Google searches. They have short time in site.

Library staff – unique breed because they are knowledge navigators for other individuals. They are not end users of the content, thus adherence to external hierarchy in deference to end user has to repeat the navigation. You need to know the context of how it got there. They know it. It is their code, their culture.

Flickr – different.

Working with MSFT learning group. They have this same problem. Hierarchies or tagging? Tom down or bottom up?

I have to stick up for librarians. With all this user generated content, is there ever a point where you have to play big brother. Form synopsis. Hierarchies. I hate search myself. I don’t know how to navigate to content I want.

“Web 3.0”

Where is the gap between this user content and structure. Then start the cycle again.

Semantic web.

People have different preferences that impact how you create the infrastructure to support the content. Flickr invests in the tagging, not the up front organization.

Librarians already have an existing body of knowledge.

What about the Library of Congress Flickr photo project. There is hierarchy and a folksonomy.

Difficult to tell users to use search and they decide where things go.

Want to integrate existing content into social network — what we intend to do is relaunch our content on a platform that does a couple of things. 1) allows us to distribute editorial authority to whoever and as many as we like. Take user contributed content model as far as we want. Can’t do that now. 2)_wrapping existing content and intent to distribute authority to users with the social tools. We have message boards and a blog now, but later people will have deep profiles, can friend around a content item. Wrapped around this rich content collection we hope to grow through user contributions as opposed to building or getting from organizations.

What’s the core, if you have to pick one problem you are trying to solve. Trying to change content model? Lower cost of content? Develop new way to consume content? Adoption of new social tools and ways to make connections and share? First is find stuff to answer their questions and solve their problems.

My job is to get them here, her job is to keep them! (Laughter)

Discoverability is a huge issue. Deep content base built over 5 years. Have gutted every line of code, but migrating the content. As much as we downplay search, we are investing in search and integrating it into the social aspects. Third pillar – online training. Essentially content elements we charge for. We have always been a deep content repository. We are now want to increase findability and add social element.

IBM dog ear now part of Lotus connections.

If Rachel has archive of several thousand items, with good search and folksonomy, can she get by without a taxonomy.

(I said something about buckets, Jim called it Taxonomy light)

If they aren’t going to adopt folksonomy then you have to do tax

They will do an implicit folksonomy whether you want them to or not based on what they choose to consume. That information that they consume represents a folksonomy. There is in fact a personal selection that transcends a taxonomy.

Incentive systems may be an issue.

It is not either /or. Popular and recent represent a taxonomical classification.

(Frank read a quote about folksonomies that I disagree with. )

It matters how you present things. Find buckets then tag it for their own and build from the ground up.

Brian Hsi -What you just described is one of the key success points of communities. Multiple entry point. What do people think are the key success factors of any community.

• Conversation
• People
• Storytelling
• Goals

Brian thinks there are some core things that transcend.
Context is king to any evaluation

• What do you got?
  Whoever the sponsor is (corp perspective) is present and active in community
  Participants are aware of their options of how they participate
  Can easily discover resources from sponsor and community members
  Not only can they easily discover, but contribute and connect
  Identity
  Healthy and vibrant – sort of fuzzy catch all
  People motivated, passion, energy, impetus for participation. May be diverse reasons

When we look at integrating community in consumer support, we know 70 are only consuming content, but that is critical to success of community. What they search on and consume. You want to create a vibrant community no matter how large or small because value is in the lurkers.

How do you build in implicit or explicit reward for those contributing. Some do it for expertise recognition, MVP status. How do you activate that 2%.

How do you take people’s predisposed behaviors. Hard to create new behaviors.

What is it about a community, transitory or not, that makes it successful.

Curiosity?

Connection?

Quoting Charlene Li – 80% of first action on social networks is seeing what their friends are doing. Stalking? The business value of stalking. There is a great blog post there.

In a support community a lot of it goes to how personally identifiable and built on personal social connections. Built of people you have social connections with. On a support forum – much different social graph, different value propositions – personal and professional. And very specific behaviors you want people to do .

You have to be careful how you define contributions. Contributing content? Rating content? Those are all contributions.

Bob – I think a lot of these discussions have centered around incentive structures. (Bob pulls up a slide deck). We facilitate conversation. I’m using PowerPoint and anyone who knows me this is antithetical forced linearity. We do a variety of work outlining categories of incentives might be. Different from social systems. I’ve looked at all the things that people represent as incentives and try to distill what they are. The real question is for what activities do you pursue those incentives. And the other thing is the impact of the “Friends graph” – in the graphing social patterns conference I pretended I was a speaker and sat at the speakers table. Got a network tap and spoke to all the presenters. The smart folks knew there was more to life than the social graph as defined by people called friends and recognize there is a commercial opportunity to address those18 million

When we think through social incentive systems, what kind of things come to mid. Widget are cool, any contribution is a good contribution, Rirro and score reading and tagging, … wait a minute, reading passive contribution is a critical contribution to my community, completely ignore them. (I add, CoPs really look at periphery.) When someone reads what we put, that is a contribution. We should reward and incent that behavior. What rewards read activity. I use the word mirror. We could capture your read activities and give you credit for having accumulated knowledge on a subject. Since widgets are cool, I could give you a widget on FB that Joe is increasing his knowledge. Don’t care if they game it. (Tagged it, not just page views or dwell time. ) If you tagged or rated it, those are different activities than reading. TO say this person has paid attention to this sort of thing. What is displayed. Implicit and explicit things discoverable, give options Robert Rebholz on delicious (too many tagged social software)

How is this different than last FM’s plugin on FB. Not necc different. Conceptually Joe has decided to express his interest in subject A. LastFM is entertainment. Bookshelf applications another example.

Do you have any research on who would feel this is an incentive.

Even looking at consumption of reading as valuable activity, you can only activate that if they have created activity and joined, even in a semi false way. Not real until they identify with the system.

It bothers me a lot that my entire experience of what I can express is in what they have identified themselves to me of what they’ve read. I don’t want to incent people just to read resources that I provided. Prefer if they embellish their online reputation by accessing any resource they accessed across the web. Can’t address that in my current position.

Micro incentivizing micro behaviors. Look at Amazon. Exact model. Best seller lists are the same thing. Or people who looked at this also looked at this. Value in reviewing, but so are views and less intensive rating activities. Sort of anonymous.

What mybloglog is all about. You still have to sign up for it and opt into the service.

Some don’t want it. On the other hand more recognize if 2 people show up to get a job, one has a cover letter, resume, and one adds a website that points to a blog and evidence of things they have expertise in, they have a dramatic advantage over the first. Contribute to my FB profile or express something on my LinkedIn. Can do this with my internal community as well. Have engineers in MSFT who will make the claim they are a customer focused engineer. Anyone can say that. But if I can give him a widget or link to show that X has collected a bunch of points through his contribution to my community and becomes part of his linked In profile. Now realizes his participation is not only nice, internally useful but helps him get his next job.

You can game the snot out of that.

The people that end up gaming it care about it.

Write a bot and have it work while you sleep.

But the truth will out. People will ask you a question and …

In actual practice it would be fairly straight forward.

Oscar is already in the system. How do you … we have as support site that gets 650 million hits a year. They don’t stay. When you create a participation model to get people to contribute, who may not think of themselves as contributors, you want to get them to consume. The tipping point is getting them to join.

I don’t think so. It is one step before that. Is getting them to care that your widget has some value. That their online persona has value to them in whatever context they are engaged in, they will jump through hoops of some size to embellish their personal persona.

What will drive them to join?

Why should they have to do many things to fill out a profile. Why should they have to fill things in. Just ask them for a name.

People like to compare themselves to others and congregate with people who are like, or just below them.

(Note taking fatigue and have to pee. Gonna stop now)

P.S. Passport is Steve’s fault

Thanks to Bob for hosting. Bill (redplasticmonk) says it is the nicest place we’ve ever met.