The Netflix Prize: Got an Information Systems Problem? Maybe This is Your Answer

According to the Netflix Prize Leaderboard, as I write this some 8,352 teams have volunteered to try and improve the online movie rental company's matching engine in the hopes of garnering a $1M prize.

That is a lot of brainpower being applied to a single business problem--far more than $1M would have bought Netflix had it gone through the normal requests for proposals and contracting process--and without the guaranteed results. (If Netflix doesn't get at least a 10% improvement by its own measures, it doesn't have to pay a dime.)

It's Information Systems Meets The X Prize, and I like the concept.  I especially like the groundrules. As best as I can tell with my never-been-to-law-school eyes, Netflix isn't making claims on the intellectual property of any of the entries. Even the winner, it seems, need only give Netflix a non-exclusive license to the solution in order to collect the money.

That's brilliant marketing, as Netflix probably realized early on that if it tried to take ownership of all the entries or to lock up the IP of the winner, it would keep a lot of people from even applying.  (Ever seen those photo contests where "All photos become property of the contest provider"?) This way, they get the IP edge first, get it cheap, have it built around their own test date, and get the leap on any competition that might by the solution down the road. When this prizewinner gets a little dusty, just fire up the contest engine again. Lather. Rinse. Repeat.

Of course, this technique can't work for every situation at every company. The solution has to be high-value enough for the contest creator to be able to offer enough incentive to make it worth the efforts of wanna-be winners. Anyone with a video camera, some free time and a bit of software can enter the Colbert Report's Greenscreen Challenge with nothing more than momentary Comedy Central glory on the line. Finding people willing to have a go at your supply chain management problems might take a bit more of a lure--possibly even more than Netflix's cool million.

Then there's the problem of sensitive information. Netflix gives entrants anonymized rating data as a sample from which to work. Most corporations, however, might be a bit more concerned about providing, say, gross margin data from their supply chain systems so that folks could tinker with it.

Still, it's an interesting approach to the problem aided and abetted once again by the Power of the Internet.

By the way. We have a serious spam problem with our blog comment engine. I've got $50 and a Salesforce.com Mr. Potato Head for the first team to solve it for me. Takers?

Join the discussion
Be the first to comment on this article. Our Commenting Policies