Friday, April 3, 2009

Crowdsourcing: When the Crowd is Wrong

Crowdsourcing evangelists are pretty breathless about its benefits. Jeff Howe, author of Crowdsourcing: How the Power of the Crowd is Driving the Future of Business, says "the crowd is more than wise -- it's talented, creative and stunningly productive." In his book, Wikinomics, Don Tapscott contends, "Billions of connected individuals can now actively participate in innovation, wealth creation, and social development in ways we once only dreamed of."

It all sounds so wonderful, but when did it become fashionable to be associated with a crowd? We used to want to "stand out from the crowd." There was a time we hated to "get lost in the crowd." And when we said "Two's company; three's a crowd," it wasn't because three people was the desired state.

Of course, Crowdsourcing--the act of organizing large groups of people outside an organization to collaborate on tasks that might otherwise be done within it--holds terrific potential, but it also can be dangerous. To effectively tap the "wisdom of the crowds," it is vital the task be carefully chosen and defined, all possible outcomes foreseen, and appropriate limits and guidelines implemented.

One danger of crowds is that they're prone to be affected by passionate minorities. Stephen Colbert has proven this several times already. In 2006, Colbert asked the viewers of "The Colbert Report" to vote for him in an online poll to select a name for a new bridge in Hungary; within two weeks he had earned 17 million votes--7 million more than the population of the country of Hungary. (The bridge was eventually named Megyeri Bridge, a name that didn't even make the second round of polling.)

Colbert is back in the news because he urged his viewers to vote for him in a NASA poll to name a new space station module. NASA offered four names of its own but permitted write-ins, which created an opening for Colbert's highly engaged fans. When voting ended, "Colbert" beat the most popular NASA name, "Serenity," by 40,000 votes. As noted in the Associated Press article, "NASA's mistake was allowing write-ins".

You might think that NASA covered it's bases since it "reserves the right to choose an appropriate name." In this case they're probably safe, since although Colbert and his fans might howl in protest when NASA passes over the comedian's name, it's still all in good fun. But what if a different set of passionate people had hijacked this exercise in crowdsourcing?

As a hypothetical example, what if instead of Comedy Central fans, it was a vocal set of opponents of California's Proposition 8 that had targeted the NASA voting campaign? Such a group might have mobilized folks from across the country to have the space module named "Milk," after the gay politician who was the subject of the recent award-winning film, "Milk." In this case, even with the stated right to choose an appropriate name, NASA would've been caught between a rock and a hard place--on the one hand conservatives might have objected to the politicizing of the space station, while on the other hand activists would've accused the space organization of homophobia and bias if NASA overruled the popular vote.

This example demonstrates several considerations for organizations looking to mount a crowdsourcing effort:
  • Consider every possible outcome and ensure the worst-case scenario is one that can be accepted.
  • Setting the rules so that the organization can make the ultimate decision regardless of the group's wishes does not mitigate every risk. Consider again the worst-case scenario in order to determine if the control retained by the organization is real or imagined. Recognize the danger inherent in publicly nullifying the will of the group after having asked that group for their time, consideration, and opinion.
  • In many cases, open-ended tasks may furnish an opportunity for a passionate minority to overwhelm the majority. It may be better and safer to allow participants to choose from a set selection of options or to otherwise constrain the thinking and activity.
  • Put limits in place that prevent or reduce ballot stuffing. While it may be difficult to obstruct all manipulation of a crowdsourced program, many sensitive situations can be avoided by setting appropriate rules and implementing technical restrictions such as requiring registration and limiting participation based on email address and/or IP address.

Crowdsourcing furnishes tremendously exciting ways to gather knowledge from stakeholders, increase loyalty among customers, mitigate risks with fresh thinking, and provide a unique perspective from outside the organization. If you execute and manage a crowdsourcing program with the appropriate foresight and care, you'll find yourself agreeing with P.T. Barnum, who said "Every crowd has a silver lining."

No comments: