Crowdsourcing creative content: a case study

Over the past few years, a lot has been made of “crowdsourcing” trends. It seem like everything — from graphic design and logos to funding — can be made better, faster, or cheaper thanks to crowd. Stephen and Sunil decided to test whether crowdsourcing creative content, specifically an entire web series, could be accomplished at a reasonable cost (with similar quality).

Unknown Sender: “Traditionally” Produced

During the Writer’s Guild Strike of 2008, when SAG and its sister unions allowed their members to defer salaries for the only remaining game in town, original web content, Steven produced a web series, entitled Unknown Sender. Unknown Sender was conceived as a suspense anthology akin to “Tales of the Crypt” or “Alfred Hitchcock Presents”, with macabre morality plays lasting six minutes or less. Steven and the Unknown Sender team were lucky enough to attract such talent as Timothy Dalton (License to Kill, The Tourist), Joanne Whalley (The Borgias, Willow), and comedy legend Stan Freberg. Their behind the scenes talent had credits ranging from Lost to The Incredibles and Hellboy. Although all the cast and most of the crew deferred salaries, as a product (and possibly brainwashed victim) of the studio system, Steven did not scrimp on equipment, safety personnel, and, most critically, catering.

Here are two sample episodes:

Sample 1: Slippery Slope

Sample 2: If You’ve Seen This Tape

Cost per minute to produce on his own, keeping costs lean = $ 1,000/minute.

Crowdsourcing: De Souza, Scripped and Talenthouse

Unknown Sender got considerable critical acclaim, not only from the blogosphere but from the mainstream media as well, and was a triple honoree in the 2009 Webby Awards for Best Series, Best Writing, and Best Individual Performance (for Mr. Dalton). Wanting to continue Unknown Sender but now re-immersed in his conventional media projects stalled by the strike, Steven turned to Sunil and Scripped to crowdsource new scripts for Unknown Sender, and Talenthouse to crowdsource videos based on those scripts. They received over 200 submissions for the script content (in one month’s time), and averaged roughly 10 video submissions/per script once the video portion of the contest began (duration of two months). An incentive of $ 200 per winning script was allotted, and $ 500 per winning video. All winners received a contract for half of the future net profits. Entries came from not only the United States, but from the U.K., Chile, Spain, and Russia, and the results were impressive:

Crowdsourced video example 1: Car Alarmed

Crowdsourced video example 2 (NSFW): Red Light Lisa

Cost per minute to produce using crowdsourcing = $ 140/minute.

Takeaways/Lessons Learned

Stephen and Sunil were impressed by the quality of the crowdsourced screenplays and crowdsourced videos. That said, the cost per minute shown does not account for their time to vet the scripts, time to vet the videos and, most valuable for all involved, time to give each filmmaker one-on-one criticism.

In the end, there were definitely a few major takeaways from the whole exercise:

The crowd needs management before, during and after production. This was an area they fell down in: After the scripts were approved, the filmmakers were left to their own devices. A surprising number lost track of the ground rules, i.e., that all entries were to appear to be “found footage”, and diverged from that concept in post, if not during production itself. Had they had a layer of screening and interaction with the contestants in the gap between delivery of script and delivery of finished cut, and had viewed dailies and assemblies (easy enough in a wired world), these filmmakers could have been saved from disqualifying themselves by veering off course. Here, the model for the vaunted “new media” is clearly the almost century-old studio system.
The quality was much better than expected, approaching independent film or broadcast quality in some respects, such as acting, production values, or direction, but not yet in all cases across the board.
Filtering the artists before they submit might be a more viable solution, or “curated” crowdsourcing. This solution may not be as attractive for other types of projects, but for video production, is a must.

Stephen and Sunil’s overall takeaway was that crowdsourcing will ultimately challenge the standard model of independent productions, but only in certain unique circumstances such as anthologies like their examples, or possibly picaresque, episodic features akin to Dazed & Confused or the more recent “Life in a Day”.

Crowdsourcing will also fundamentally change the way brands think about film production — small businesses can use crowdsourcing effectively to put together ad campaigns — or big companies looking to cut costs. That said, crowdsourcing video alone is not good enough, because there are too many undesired outcomes. Filmmakers need scripts to work from in order to make video crowdsourcing effective, and produce desired outcomes.

With credits including Die Hard 1 & 2, 48 Hours, Commando, and Tomb Raider, Steven de Souza is one of a handful of American screenwriters whose movies have grossed over $ 2B in the worldwide box office. He has also produced over a hundred hours of network television.

Sunil Rajaraman is the co-founder and CEO of Scripped.com, a community site for nearly 100,000 screenwriters. Scripped recently launched a subsidiary, Scripted.com, to enable brands to purchase content directly from its writers.

Image courtesy of Flickr user woodleywonderworks.

Related research and analysis from GigaOM Pro:
Subscriber content. Sign up for a free trial.