Web Social Architecture

Motivating Community Participation

We had a huge success in the usabililty lab a couple weeks ago around motivating and rewarding participation in an online community. We were testing paper prototypes of a large new site. Our initial plan not only completely failed to motivate people to participate, it actually caused them to want NOT to participate.

Let me back up. Our client asked us for a proposal for an initiative to produce user-generated content. Great. We came back with something really REALLY cool (I'll say more when I can). The feeling was that people would want participate, just because the place is going to be cool (really REALLY cool), and to hedge our bets and ensure we'd see some very high-quality content, we suggested running a contest. The grand prize was going to be big, even on the order of a tricked-out vintage car.

We thought this was reasonable. Of course participation has its own rewards, but let's go ahead and make sure at least some people will invest some serious energy.

And we were wrong. In such a good way.

Not only did people not respond positively to the contest idea, the handful who acknowledged it expressed suspicion... and even revulsion. The overall response was more or less, "A contest? What are they trying to get from me?"

Here's the reason this result was so positive: People were in fact motivated to participate--not by the contest but by the social rewards that surrounded participation.

We killed the contest and focused on increasing the social rewards of participation. A great example of user-centered design in a community context: Usability lab feedback redirected a major aspect of site strategy and, I'm pretty certain, saved us. (Is anyone else out there testing prototypes of community sites with users? I'd love to hear about it.)

I just saw a blog post from Best Engaging Communities about best practices for motivating community participants. The post describes a KM project which motivated knowledge-holders to "leave a legacy." It's a great idea, because it reaches just the right people (folks nearing retirement) with just the right message: Your time here was meaningful. I believe someone in the leadership development world wrote a paper about career stages, and the mentorship stage is a critical late-career bit.

And the post was right also about not paying participants. Likewise, we saw very clearly that the answer isn't to run a contest. Big prizes are probably even less motivating than that $20 Starbucks card.

Extrinsic motivators don't work for community participation any more than they do for anything else. What does work? Social rewards. So if you want to motivate people to participate in your community, focus on building a rich social experience around participation. You could even say the richer the social experience, the more motivated participants will be.

Here are some ways to do that. I'm sure there are many more:

• Provide a flexible public profile space people can customize to present themselves as they want to be known. Do better than photo, location, bio. 
• Provide a data-rich private dashboard for people to monitor the activity they care about, especially information about how other people are interacting with their contributions. Show views, downloads, # of comments or replies, times favorited, overall rank, etc. 
• Flag new stuff. Make it easy to see the activity that matters to the individual. 
• Add thumbnails where people add content. Humanize the text. 
• Promote a social contract and give the community the ability to rally around it. Let users comment on your terms of use. 
• "Introduce" people with things in common.
• Let people choose whom to trust how much.

There are a million examples. The real revolution, from the perspective of understandably-nervous launchers of new community sites, is to focus energy on developing sociability, and to say no to the temptation to hedge your bets.

Community of Multiple Practices (Better Yet, an Open Source Secret Society)

Hee-hee. I started a secret society at work. I can't tell you much about it. Charter members sat down yesterday and talked a little bit about our secret society, but talked mostly about our divergent and convergent notions about the future directions of our different practices.

The energy and excitement were palpable. We could have talked for hours. I left with a dozen ideas for blog posts.

There are rules. It's a combination of highly-structured and totally unstructured interaction. There are no lurkers allowed.

I shouldn't say more.

And, I highly recommend that everyone start planning their secret societies and takke action soon. I already can't wait for next time.

Trying Out Windows Live Writer

At the suggestion of Sean O'Driscoll, whose blog I ran across at Lee Lefever's suggestion, I'm trying out Windows Live Writer. Wow, I'm pretty happy with it, particularly in comparison to Typepad's epic interface. And, I love that I'm typing this on the airplane, and not in Word, which has a nasty habit, albeit with workarounds, of pasting HTML with embedded styles.

I have a client who runs a group corporate blog with multiple authors, many of whom paste from Word and in the process override the global style sheet for the blog, producing posts with a mishmash of fonts and formatting. A headache to clean up.

Here's a picture of what I'm looking at right now:

And here's a picture in preview mode. Sweet! (Note the controls at the bottom of the screen for adding keywords, trackbacks, etc.)

Pretty nice. Thanks Sean. After a thorough trial I might suggest this to the client with the formatting woes. Anyone have an even-more-favorite tool or combination of tools?

The JetBlue Apology

Here's a great example of how to use the Web to manage a crisis. JetBlue had a problem. They issued an apology through YouTube. It worked:

• Because the apology comes across as genuine
• Direct to you from the CEO
• Who seems like a normal human, not a PR guy

And poof! The choice of YouTube as a distribution channel ensures:

• They get street cred for being Web-savvy
• They don't come across as issuing an ad, as they might in a branded space
• They get coverage in the blogosphere.
• Customers can react publicly and engage in the conversation.

Read through the comment thread on the YouTube video. The first commenter addresses the CEO by first name. That alone should be an indication of the power of choosing the right venue and the right medium to connect with customers.

More on Feevy

David Silver has more to say about Feevy and what's so new and interesting about it. David's point is  that by turning the standard blog roll into a collection of live excerpts, Feevy does more than make the blogroll more interesting. It changes the blog into a portal, with a collective authorship and new possibilities for reading in a distributed way. It shifts the point of view slightly toward the "we" side of the I / we continuum.

This strikes me as important, because it makes the distributed conversation of the blogosphere one bit more transparent. You can see right on my blog not just who I read but what they're saying today. My blog becomes a portal into my slice of the conversation.

I've added Feevy to my blog, as you can see in the margin here, and I'm pretty pleased. And, as you can see, the layout and presentation aren't perfect. I'm using the "liquid layout," because the default fixed-width layout doesn't fit in my margin. It looks a little funky--but I'm ok with it, to tell the truth. The content is what matters, right?

I initially had a quibble with the information design of the Feevy widget, because it looks a little strange to me to use the title of the post as the header and the name of the blog / author as a wee little link at the bottom of the excerpt. Now that I think about it a little more, though, I'm starting to see the presentation of the content as primary and author as secondary as an interesting shift away from viewing the blogosphere as a collection of identities and toward viewing it as a collection of voices.

In other words, a standard blogroll uses the author as the organizing principle, and the content is subordinate (and a click away). Feevy uses the post title as the organizing principle, and the author is subordinate.

Those headers still look weird to me, though. This points to a real challenge forthcoming in the design of a multivocal web: when lots of people contribute equivalent objects, like the headers in Feevy, they don't tend to make sense as a group without the right kind of consistent framing. They lack equivalency. The author / blog link in Feevy should provide that structural context, but it's a little too small to work well.

UPDATE: Feevy added a new layout option that works much better on my blog. See the new version at right, and David's comment below.

The lesson for me here is that collections of user-generated content objects require a backdrop of context, framing, and containment. The free-for-all idealist in me feels a little nervous typing those words!

There's a new information architecture on the horizon. Stay tuned.

User Feedback on a Community Design: What a Concept!

This week is the second round of prototyping for a large new community site ZAAZ is developing for a major (and favorite) client. The site is focused around sharing of user-generated content objects as a product value-ad, and it's also meant to reshape the way people think of the brand.

The first round of testing was based on paper prototypes. We used the RITE method. We made huge advances, in particular in understanding the motivations behind various levels of user participation. In a nutshell, we learned (and were able to prove, and show) that people are much more motivated to participate in a community site by the richness of the social rewards surrounding participation than they are by extrinsic motivators like a contest. That might be old news for community types, but for companies investing in major community initiatives, the fear of throwing the lame party nobody comes to is significant and understandable. Paper prototyping is a great way to prove the community concept.

I'll be saying this another thousand times: There is no substitute for watching users interact with a site while it's still in design. This week, testing with a functional prototype, we're learning:

• Stuff we suspected but weren't sure about
• Some stuff we worried about is working fine
• Some stuff we weren't worrying about isn't working well
• Users understand structure in ways we'd never have imagined
• In two months of design, we didn't think of everything. Some of these things are important and seem obvious in retrospect.

This site is going to rock, and we'd never have gotten there by relying on our own judgment and expertise.

Here's a cameraphone shot of the ZAAZ usability lab in action.

Debate Results: There Is No Debate

Tuesday's debate with Gary was fun, as expected. Along with the general explanation of folksonomy, I offered a number of digs, some personal, some professional, and generally tried to win the debate by:

1. Changing the subject to unrelated topics, such as Gary's shoes, sweater, and age.
2. Portraying Gary as an extremist.

It worked. Folks laughed, including Gary, and a number of the students were excited enough about being part of the future of the Internet that they peppered me with inquiries about whether my employer might hire an intern.

Here are a couple of the key (non-irrelevant) slides:

 

My point here was inarguable (because I had the mic). And if you accept the first point then the subsequent seals the argument:

Yes, that's marketshare. In the past year. (Data is from PEW.)

Good fun!

Feevy: Dynamic Blogroll Widget

David Silver points to a great new blog tool called Feevy. It uses RSS (see Lee Lefever's great post RSS in Plain English) to grab recent posts from blogs of your choice, and it displays them on your page. You tell Feevy what you want to watch, and it spits out HTML for you to paste into your blog. The result is a dynamic blogroll, showing not only who you're linking to but what they're saying.

Awesome. This strikes me as a clear step forward for blog tools.

For one thing, this tool has huge potential to raise your blog's search engine visibility, because it effectively adds frequency-of-publication to your blog, along with keywords dynamically updated from across the blogosphere. From an SEO perspective, if I'm understanding it accurately, it makes you instantly current in your area of interest.

I'm also fascinated by the metadata Feevy is collecting. They're capturing a giant, time-dimensioned map of the blogosphere. What on earth are they going to do with it?

Feevy's site isn't forthcoming about their revenue model. Watch here for updates. I'll try out Feevy on this blog when I get a few minutes to set it up.

UPDATE: I added the Feevy widget. It works fine. I had to select the "liquid layout" option to fit it in my sidebar. Looks just like part of my blog! I was momentarily confused when not all the blogs I added were showing up--it's because the feeds aren't current. They'll show up as blogs are updated. Aspects of the interaction design on Feevy's site are a little confusing as well, but I got through it.  Check it out, over to your right:

Viral Ice

Vanilla Ice is back! Vanilla Ice is back! Vanilla Ice is back! The Church of the Customer Blog has found something really special here.

I grew up listening to not much but hip hop, so Vanilla Ice has a special place in my heart. He takes me right back to 8th grade--but as it turns out, Vanilla, like me myself and I, has graduated from adolescent yearnings expressed to the beat y'all to the practical adult world of tax software and scraping together a living. He looks great--I hope he's doing really well.

Yes, I feel nothing but kinship with the fellow white boy. Here we are all these years removed from our b-boy prime, and we find ourselves on opposite ends of the viral advertising world. If things had been different, I could have been in his shoes...

Tax rap? What a dumb idea. I love it, and I love this guy:

Live! 1pm Tuesday! Social Tagging Smackdown

Tomorrow I'm participating in a debate in Mike Crandall's course in Organization of Information Resources at the iSchool at University of Washington. I'll be representing for social tagging and folksonomy, and Gary Carlson, Chief Taxonomist of Schemalogic, unfortunate soul, will be arguing the side of structured taxonomy and controlled vocabulary.

Now, Gary's a good friend of mine. He's an expert in his field, with 10 years of deep thinking about taxonomies, vocabularies, thesauri, ontologies, and other foofaraw under his belt. He's an articulate salesman for information management in the enterprise. It wouldn't be at all a stretch to say he's the brains behind Schemalogic's success.

He must be NUTS, to want to debate me. Doesn't he realize I'm a consultant?

I've already got a lot to say about folksonomy and tagging, because it's a topic clients ask about fairly frequently. The basic gist of the argument in favor of bottom-up approaches to metadata is that they're cheaper and faster to implement, more accommodating of cultural differences, and quicker to adapt to changing mores. Plus, they're cool right now.

In other words, taxonomies are for corporate drones who have no desire to keep pace with change on the Web. They're for cubicle-bound, dusty-shouldered fuddy-duddies desperate to keep the seething masses at bay, lest someone notice their shoes, out of date since 1982. They want control--not even for the exercise of whatever arcane power they might gain, but for fear of change. For fear of chaos. For fear of obsolescence. You listening to me, Gary?

It's all good-natured, of course. We'll reach some kind of grand hybrid conclusion about appropriateness of different approaches in different circumstances. Along these lines:

Folksonomies are good when:

• Content objects need to be tagged rapidly and at a low cost,
• Duplication, ambiguity, and inconsistency serve to strengthen the overall information environment, promoting discoverability,
• Relevance is more important than comprehensiveness,
• And, users have incentive to tag, meaning usually either promoting their own content or making things easier to re-find.

Taxonomies are good when:

• You need to manage distributed information assets within a discrete set of communities, such as within a corporation,
• Accuracy is critical, such as in an inventory,
• You have a staff of library-science geeks you need to keep busy,
• If you're more interested in structure and control than in reality. You listening to me, Gary?