There's been much ado regarding the recent Netflix announcement to pay $1M to whomever can crack the code to improve their collaborative filtering engine. While much banter has been tossed around regarding this topic, the PR splash, and other facets, I see two other stories going on behind the scenes.
Story #1 - Building a talent pool. (see netflixprize.com): As of this post, there are currently 8775 contestants on 7170 teams from 97 different countries engaged in The Contest. As David posts on his GenuineVC blog, the value of solving the code is worth far more than $1M as it drives renters down the long tail of their available titles and away from more expensive, first run titles. That said, Netflix now has a talent pool of ~9,000 softwre developers that it is building a relationship with, and able to evaluate their problem solving capabilities to boot. This is a very interesting, captive talent pool of smart, committed, and entrepreneurial people. What is the value of coalescing that many like-minded people (way more than $1M)? Never mind the fact that the publicity value alone has surely funded the $1M contest already. This story becomes far more interesting if Netflix has other plans in the works or needs to scale for some particular business reason. I wonder if other companies will launch similar problem solving exercises in the future?
Story #2 - It's not about the 1 million dollars. I was talking with my good friend Greg Boesel over at Swaptree who is one of the smartest guys I know, and he suggested this: If you read the rules of the contest, you will learn that you don't need to submit the solution algorithm until you've been qualified as the winner. So if the winning team is able to create a more successful collaborative filtering model, why would they settle for $1M? I can think of a host of companies that would pay more for the algorithm and thinking behind improving the suggestion engine of their website (um... why isn't Amazon.com better at suggesting things I really want?). Why wouldn't the noted winner, once confirmed that they have a superior solution, try a different approach to extract the most value out of their notorious solution?
This may be interesting to watch, provided a superior solution can be developed...
Comments