Netflix & The Value of Crowdsourcing (if Such a Thing Even Exists)
October 5, 2009
Contesting the Crowdsourcing in Netflix’s Competition
Recently, Netflix engineered a “crowdsourcing” competition, awarding $1 million to the contestant who could make the company’s existing movie recommendation engine 10% more accurate.
There’s also been a lot of hullabaloo over the term “crowdsourcing.” Netflix didn’t crowdsource; running a generative idea contest within the software development community, if PR-catchy, doesn’t constitute the outsourcing of one task to a large number of potential contributors.
Likewise, crowdSPRING (left) brands itself as a crowdsourced “marketplace for creative services,” but ultimately allows for the selection of one option, sprung from a non-collaborative world of ideas.
(Not to get down on Netflix– we love what they, and other, shared service providers like Zipcar are doing to encourage “the new sharing economy.”)
What’s the Value of Crowdsourcing, if Such a Thing Even Exists?
Recently, I came across a snarky Forbes article compiled by a crowdsourcing near-nihilist, Dan Woods, entitled “The Myth of Crowdsourcing.” With respect to Netflix, he was right in that any late-stage pooling between teams seemed more interdisciplinary collaboration than crowdsourcing (crowdsourcing by the company was well out the window); innovations appeared, more or less, to be “aggregations of the inventions of individual virtuosos.”
But what we consider to be truths (and virtues) of crowdsourcing, the author refused to give due credence across scenarios:
1. “General” crowds can provide more compelling results than specially talented individuals.
“Whatever term we use, let’s not call it crowdsourcing and pretend that 10,000 average Joes invent better products than Steve Jobs.”
Dan Woods for Forbes, “The Myth of Crowdsourcing.”
We see the answer to this in the aggregation and layering of “average” contributions. Per Yochai Benkler, the NASA Clickworker’s program had 85,000 users voluntarily visit a Web site where they could mark or classify craters on Mars, the aggregate results of which effectively created a heat map “virtually indistinguishable from the inputs of a geologist with years of experience in identifying Mars craters.”
Truly “general” crowdsourcing works with repetition in mind; it harnesses the wisdom of many (or, a few minutes from 85,000 average Joes to replicate the long labor of a highly skilled NASA worker). More popularly, Wikipedia creates a [less visually trended] refinement of content through the work of thousands of contributors, self-selected from the general Internet populace.
2. “Crowds” can be composed of many talented individuals, rather than reliant on one “individual virtuoso.”
“There is no crowd in crowdsourcing. There are only virtuosos, usually uniquely talented, highly trained people who have worked for decades in a field [...]. Yes, there are large teams of developers on open-source projects, but without the virtuoso contribution at the outset, they would achieve nothing.”
Dan Woods for Forbes, “The Myth of Crowdsourcing.”
Woods’ philosophy of collaborative communities and crowdsourcing acknowledges two types of individuals–individual virtuosos and average Joes.
Why can’t there be an organically-formed crowd of virtuosos? Or niche topic “savants”? The most successfully crowdsourced initiatives allow for individuals to volunteer for tasks, and to self-determine the size and nature of their contributions within the confines of the project.
For tasks where monetary rewards aren’t prominent (or for individuals not much motivated by monetary rewards), participation in a crowdsourced activity is largely mediated by intrinsic (i.e. “just because it’s interesting”) and social-psychological rewards (perception by others, personal satisfaction)–so it’s no wonder that “crowds,” as they’re popularly required for crowdsourcing, tend toward “leader” communities.
Yochai Benkler & Resources on Crowdsourcing
For more on crowdsourcing, the open-source movement, and the dynamics of collaborative communities, check out Yochai Benkler‘s TED talk below on “The New Open-Source Economics.” The presentation just brushes on topics covered in his free, 79-page PDF, “Coase’s Penguin” (a selection from the Yale Law Journal, written in 2002 and eerily on target). Benkler also wrote The Wealth of Networks, published in 2007.
Given the increasing semantic flexibility of the word “crowdsourcing” (and the occasional denial of its existence), we still think it’s an impactful utility and a curious innovative, social phenomenon.
Header image courtesy of editor’s flickr, (cc) some rights reserved.
Generally speaking, I find that most often a few “virtuosi” are enabled by a larger organization—and enable the organization. This person could be a true genius like John Resing, Linus Torvalds, and yes, even Steve Jobs). These key individuals often do more than coalesce a group. Their presence in the community makes participation worthwhile.
You could even contrast virtuosi with figureheads who had a great idea at the right time like Craig Newmark or Bill Gates. Their presence is tolerated and often respected, but doesn’t draw the same enthusiasm from the greater whole.