Sign up for the FREE Tell Membership and receive benefits that include the digital edition of Tell Magazine sent straight to your inbox, product giveaways, coupons and much more!
Sign up for the FREE Tell Membership and receive benefits that include the digital edition of Tell Magazine sent straight to your inbox, product giveaways, coupons and much more!
Last year, I attended Derek Powazek’s panel on crowdsourcing. This year, he held a similarly-themed panel called Design for the Wisdom of Crowds, but it wasn’t simply a rehash. He had all new goodies. For example, he noted how peer-to-peer groups are becoming de facto recommendation engines. Want to know what the crowd thinks are the ten best U2 songs? Type U2 into Limewire and you’ll see a ranking of what most people own and download. Personally, I think we’re going to see "Most Illegally Downloaded" become a huge metric for corporations, even if it never gets published by Billboard.
This year, Powazek focused on answering the question – If the crowd is so wise, then why does the comments section on YouTube suck so much? He came up with a rather interesting answer, but it’ll take us a while to get there.
He started off by observing that, according to James Surowiecki’s The Wisdom of Crowds, in order for a crowd to be wise, it must exhibit the following characteristics:
- Diversity
- Independence
- Decentralization
- Aggregation
Which, in online terms, translates into:
- Small Simple Tasks
- Large Diverse Group
- Design for Selfishness
- Result Aggregation
That first part basically means ask simple questions like "Hot or not?" to get the best results.
Diversity goes beyond the usual demographics into every dimension you can imagine, and a good way to engender it is to lower the barrier of entry to participation on your site over time instead of raising it. Raising the bar kills diversity.
Designing for selfishness basically means you’re answering the user question "Why should I bother?" Google does this without the user even knowing it because the user, in this case, isn’t even on Google. Bloggers and Web publishers link to other sites for their own selfish reasons. Google uses those links as votes to determine a site’s relevance and rank. Tagging is a similar proposition. You might tag a photo on Flickr for your own reasons, but collectively those tags create a powerful resource that adds value to Flickr, even though you didn’t do what you did to help Flickr.
An interesting problem that comes up with the final step, aggregation, is gaming. That is to say, when the process you’re using to aggregate stories (like a ranking system, for example) becomes a game that people try to "win." This alters the quality of the aggregation. For example, when Flickr introduced the interestingness tag, they displayed the results as a ranking and people tried to game the system to end up on top. Flickr changed the display so that the user sees a random sampling of what people consider interesting, taking the game out of it.
Other clever implementations include Amazon’s user reviews. They show the most popular positive and negative reviews side by side, disabusing people of the notion that most popular = most correct.
(By the way, even color can influence how people interact with a site. Blue backgrounds engender creativity, red attention to detail, so you should choose your content and advertising accordingly.)
We now come to why it’s important to take all of the above steps to optimize the wisdom of the crowd that is your audience. As it turns out, our brains really like to fill in gaps. Whether it’s persistence of vision or creating narratives to make sense of the world, the human mind abhors a vacuum.
In online communications we are missing tons of sensory input. We are given no tone of voice to work with. No facial expressions. Our minds have lots of gaps to fill in. Unfortunately, and here’s where we get back to YouTube, we often fill in those gaps with our own insecurities.
In an experiment, Group A was asked a series of questions and no matter how they answered they were told they were correct. Group B was asked the same set of questions and no matter how they answered they were told some answers were right and some wrong, but not based on whether the answers were actually right or wrong. Group B was called the "out of control" group.
Group A, after answering the questions, was shown pictures of static or clouds and asked if they saw anything in the image. They saw nothing. Just static. Just clouds.
Group B, after answering the questions, was shown the same photos and asked the same question. They saw things. Something in the static. Something in the clouds. Meaning in the chaos. Denied control, made insecure, they filled in the gaps.
Think of Group B as the majority of YouTube commenters.
The wisdom-of-crowds-inspired design strategies Powazek prescribes aim to provide as much feedback as possible to reduce the number of gaps a user has to fill in and thus reduce the number of opportunities for that user to fill in those gaps with offense.
"How," Powazek asked, "can we give people in-control experiences?"
By the way, when Group B was asked to talk abut something they were passionate about and was then reshown the pictures, they no longer saw meaning in the chaos.
Here, by the way, are the slides from this talk.
Related Posts