Tuesday, December 8, 2009

The day Twitter broke down (failure of the hive mind)


Legend has it that Dorothy Parker of the celebrated Algonquin Round Table was asked to use the word "horticulture" in a sentence. Without missing a beat she responded, "you can lead a horticulture, but you can't make her think."

Sharp wit requires fast response, and yesterday Twitter failed miserably at that task. It all began when Edward Boches, creative chief of the ad agency Mullen, suggested four of us test the limits of Twitter as a debate forum. The structure was simple: We'd wrestle with Edward, John Winsor (formerly of Crispin Porter + Bogusky, now head of Victors & Spoils) and Ben Malbon (of BBH Labs) for 24 minutes on four topics (Twitter as a connector, content-finder, marketing tool and crowdsourcing application). We were to launch at 4 p.m. Monday, mark tweets with #tw24 to track the discussion, and see who else jumped in.

Chaos reigned. Delays in input and responses caused Twitter threads to overlap, split, and suddenly five or six topics were being chased concurrently. Observers chimed in with questions and thoughts, but most often puzzled critiques. It looked like four ad guys had suddenly gone mad on Twitter. After an exhausting hour of trying to catch up with each other, we stopped and admitted we need a better tool such as Google Wave or Jordan Kretchmer's new LiveFyre.

Your missing channel: Collaborative communication

What did we learn from this failed test? Twitter, like almost every other communication tool, still is primarily a one-way messaging service similar to physical postal mail. Users, including us, tend to be egocentric. We send out our updates or links to share what we are thinking, and hope others will respond directly to us. Yes, we hope some messages will spread virally -- retweeted for a modicum of personal fame, or for a business, marketing meme dissemination -- but those are also one-way vectors out into the masses.

The problem with all of these tools is one-way transmission makes for lousy collaboration networks. Think of your own office and the crazy email streams that begin when one person asks a question, then more people answer, and suddenly you have 35 emails marked "Who Wants to Play Secret Santa at the Holiday Party" cluttering your in box. The old dynamics of one-way mail don't work for rapid consensus building. Now, think of the best collaborations you've had, perhaps over Thanksgiving dinner, where trains of thought bounce seamlessly around the table building momentum of ideas, wit, and solution. Communication at its best becomes unhinged from one individual into a form of hive mind, an evolving intelligence based on the contributions of the whole tribe. (We've argued before that artificial intelligence may not come from computers, but rather markets of humans in the aggregate that form hive minds of higher, even predictive, insight).

This is hard to do among humans, who have yet to learn to cooperate like bees, and even more difficult online, where typing and transmission and refresh rates create barriers in communication. Google has seen a market opportunity here and is pushing Google Wave as a new collaboration platform (think a constantly evolving Wikipedia-style rich-media page for any email conversation, and you get the idea). LiveFyre, the brainchild of Jordan Kretchmer of Twitter RFP fame, hopes to build yet another collaboration portal (it's still in beta, but we have high hopes).

It's ironic that in this age when we all must check voicemail, email, physical mail, Gmail, Facebook, Twitter, and Google chat for messaging inputs that we still crave one more way to communicate with others. But we do. The selfish, ego-at-the-center one-way vectors of messaging past won't get us to the future we need ... one where we give up some control in exchange for collaborative input from others. Sorry, Dorothy Parker. We're not with you yet.

10 comments:

edward boches said...

Interestingly, our experiment was about what Twitter is good for. Expanding your tribe, discovering new content, marketing (maybe) and crowdsourcing (i.e. finding people to help achieve something big). Guess it isn't good for chat, though I try again on behalf of #socialmedia at noon today. Will let you know if it works better with Tweetchat up.
Edward

Brian Morrissey said...

as someone who dipped into the conversation yesterday, i can confirm it was bewildering to follow. i soon gave up. we've had mixed success with tweetfreak twitter q+a's. i've found that shorter is better. we've kept them to five questions.

you're probably right, tho: everyone is asking too much of twitter. it would be nice to have that kind of conversation sectioned off, whether as a wave or with livefyre or oovoo or some other tool. twitter's role is probably best just to promote it.

Jason said...

Ben,
I kept an eye out yesterday (between conf. calls) and applaude your collective efforts here. The only way to figure this stuff out is to do and for that I believe you all passed. Why? As trusted resources for your clients, you can talk from experience of what Twitter, the tool, is good for and what it lacks. Your "test" yesterday saved a handful of clients $500,000 I'll bet.

I've been running a weekly chat (w/ partner @marc_meyer) for 37 weeks using Twitter as the medium. We started as a test, learned by trial and error, and today have dozens of chapters of quality content on our topics, all crowdsourced.

I encourage you to keep experimenting and let me know if I can help in any way! Rock on!
Jason Breed

MHB said...

brilliant post.

Bob Knorpp, @thebeancast said...

I think you're being a little hard on the effort. And I don't think Twitter failed as much as you all tried to make Twitter do what it's not built to do.

Essential you were running one of those old AOL, structured celebrity chats where we all tune in to hear you debate and then maybe throw in a few questions. And while that may work in a moderated environment, Twitter is not the place for it.

The conversations that work best on Twitter are the ones where everyone is on a level playing field, where anyone can ask a question and anyone can respond. Frequently the conversations fragment and side conversation start, but as a whole the conversation functions like a conference mixer. And as soon as you impose top-down structure to a medium like that, the medium always rebels.

I would recommend that future conversations follow the (non)format of #journchat or #carchat and just have the group of you agree to participate and see what happens. The less structure the better.

My 2 cents.

Bob Knorpp
Host of The BeanCast
Posts every Monday @ http://beancast.us

Nick said...

This is what Google Wave is meant to do, in part, isn't it? Though it's hard to see how it will get round the same issues of topics splitting apart, etc.

Techcrunch seem to have similar reservations here.

KR said...

Chat is a nice to have with people. It's scaleable. When it grows, we find new ways of dealing with it. The experiment was more fun to look back at after it was over, rather than participate in. So I guess that makes Twitter the wrong place for that one right now. Twitter is still very good at other things. Maybe Twitter will adapt. Maybe things will just stay the same.

@ScottyHendo said...

Ben:

I agree with the sentiment already expressed that this was the right debate topic and wrong forum.

I was on a social media for social good panel last spring that was conducted purely on Twitter. It was the equivalent of having an in-person audience randomly shout out questions and make comments.

Twitter is an excellent platform for disseminating ideas and making new direct connections. It's also great for 1:1 and small circle dialog, but not group conversations.

Might I suggest we try this again, but this time with something more old school. Perhaps memorandums typed in triplicate and passed along via a communication tree. It might be as equally frustrating.

Scott Henderson
www.rallythecause.com
@scottyhendo

Anonymous said...

Like Woodstock, many will claim to have been there. I'm proud to say I truly was present for this historic event. And the brown acids wasn't all that bad...

- @eriknorwood

edwardboches said...

Well, this was supposed to be our trial run on LiveFyre. However, because the subject was Four Virtues of Twitter we thought it made sense to try it on Twitter. It would have been a lot easier if it were done with Tweetchat which was down. In fact today I hosted another chat for #socialmedia from @Marc_Meyer and it went off without a hitch. Everyone got to participate. It was relatively easy to follow the three 20 minute question/answers (with a little overlap due to Twitter's API delays) and the feedback was all positive. That being said, agree that it was a bit of a disaster. But as I like to say "willing to fail, determined to succeed." So we blew it. But we'll be back.