Netflix successfully crowdsources its R&D. Now what?

Almost three years ago, Netflix announced it would award $1 million to anyone who could improve its movie-recommendation algorithm by at least 10%. (That's the computer programming that tells you that if you like Eternal Sunshine of the Spotless Mind, chances are you'll also like Kurosawa films—odd as that may sound.) Well, for the price of $1 million, Netflix now has itself a nifty new recommendations algorithm. Techcrunch details the last, hectic moments of the contest:

Who knew statistical computing competitions could be so cut throat? Since we reported on the contest last night, two teams in the Netflix Prize have spent the last few hours jumping back and forth on the Netflix leaderboard as the three-year-long competition ticked into its final moments, with last minute sniping submissions coming from both sides. Finally, the results are in: The Ensemble has managed to come from behind to upset BellKor's Pragmatic Chaos with a top submission of 10.10% — an improvement of .01% — only 4 minutes before the contest closed.

It's a big win for the idea of crowdsourcing—not a new concept, by any stretch of the imagination—but one that has captured a lot of attention in recent years. What's the take-away?

Well, if you come up with an interesting enough problem and put a chunk of money on the table, the people in the world you most want solving your problem will come to you. The computer engineers who toiled away in the service of better movie recommendations hailed from around the world, and from some pretty top-notch institutions—places like the University of California-Berkeley and AT&T Research.

How widely that process can be replicated, though, is still a question-mark. In this recent paper, researchers at MIT summarize their results of studying 250 different examples of "collective intelligence"—from Wikipedia to the minor-league baseball team the Schaumburg Flyers. The conclusion: many, many circumstances have to line up in order to get these sorts of projects to work. Everything from sources of motivation (monetary and otherwise) to how the task does or doesn't break down into smaller parts.

Bounce around the Internet and you'll quickly see that the sorts of things Netflix customers care most about are improving the company's online streaming technology and getting its full roster of movies-by-mail onto its Web site. Those two tasks—which seem to be 1) a fairly run-of-the-mill programming issue; and 2) a problem for copyright lawyers—might not be the stuff of Netflix Prize II. It's a rare business need that's flashy enough to warrant three years and $1 million.

Barbara!