Scott Porad

Ben recently asked this question:

Consider Hollywood, where millions of struggling amateur writers are creating new shows and some get made, is that crowdsourcing? If not, what about American Idol?

I’ve addressed this question in some of the talks I’ve given on user-generated content.

The notion of crowdsourcing typically relies on the notion of receiving many individual inputs from the crowd with the average or sum of those inputs resulting, more often than not, in an accurate result.

In other words, when you break it down there are (at least) two key elements to crowdsourcing: 1) many inputs from the crowd, and 2) some sort of averaging or sum of inputs to yield a result.  With that in mind, I would answer Ben’s questions this way:

Hollywood is not crowdsourced.  There are many inputs (writers submitting scripts to studios).  But, the second criteria, of some sort of averaging or sum, is not met because in Hollywood there is a relatively small cabal of individuals who decide to “green light” a movie.

On the other hand, I would argue that American Idol is crowdsourced, but not in the way you think.  American Idol is like Hollywood in that it accepts many inputs (contestants), but those contestants are filtered from many to a final few by a small set of judges.  Unlike Hollywood, however, American Idol crowdsources the selection of the winner: many individuals (viewers) vote, and those votes are tallied to produce a winner.

When I talk about UGC, I make the point that American Idol is one instance of an interseting transformation in user-generated content which I call “user-driven” or “crowd-filtered” content.  In these new cases, the model has moved from relying on the crowd to source the content toward relying on the crowd to filter the content.

This is exceptionally valuable because user-generated content facilitators, like Cheezburger, can filter down large volumes of content to the submissions that are the most accurate (in this case, where accurate is what is satisfying to the audience).  In other words, crowd-filtering provides an averaging or summing mechanism for the types of things that cannot be averaged or summed.